DOCUMENT RESUME

QED: 165 054 CG 013 123

AUTHOR . Dent, James K.; Furse, George A.

TITLE Exploring the Psycho- Social Therapies Through the
Personalities of. Effective Therapists.
INSTITUTION. Natioual-Inst. of Mental Health (DHE1) , Rockville,
Md.
REPORT NO DHEW-ADM-77-527
PUB DATE*" 78
'MOTH . 213p.
AVAILABLE, FROM Superintendent of Documents, D.S. Government Printing
Office, Washington, D.C. 20402 (017-024-00686-6)

EDRS PRICE MF-$0.83 HC-$11.37.Plus Postage.

DESCRIPTORS Behavior Problems; *Counselor Characteristics;
Delinquents; *Emotionally Disturbed; Individual,
Characteristics; Milieu Therapy; Neurosis;
*Personality; *Psychotherapy; Research Projects;
*Schizophrenia; *Therapists

ABSTRACT
Several specific research approaches are compared
with regard to cost-effectiveness, types of disorders to which 'they
best respond, general strategies, and therapist personality.
Replicated findings include:. (1) support for' both. the functional
reversal-and semantic reversal of the "A -B Scale;" (2)
Characterization of therapists who are effective with neurotics,
through use of a new scale involving rejeation of problem solving;
3) identification of therapists who, are effective With
schizophrenics through a scale determining level of.interest in the
patient; and (4) requirement of great tolerance and understanding
when working with schizophreniCs. Future research must give careful
Consideration to defining both client and therapist population.
(Author/CKJ)

*********************************************************w*************
* Reproductioni supplied by EDRS are the best'that can be made * .

;* .. from the al d6Cument. - --*

..i,

**'*********************************************************************
/
 Exploring the Psycho-Social Therapies
Through the Personalities of
Effective Therapists
y.
Drug-Free Psycho-Social Therapy
With Schizophrenics, DepYessives, Neurotics, and Juvenile Delinquents,
and Therapy Plus Drugs with Schizophrenics
.0

by
JAMES K. DENT, Ph.D.
Division of Biometiy
National Institute of Mental Health
find
it the Study of Juvenile Delinquents (Chapter V)
George A. Purse, Ph.D.
Clinical Coordinator
Montgomery County Emergency Service
Norristown, Pa.
U.S. DE PARTIMENT OF HEALTH.
EDUCATION I WELFARE
NATIONAL INSTITUTE OF*
EDUCATION
THIS,' DOCUMENT HAS BEEN REPRO..
OUCED EXACTLY AS RECEIVED FROM
THE PERSON OR ORGANIZATION ORIGIN-
ATING IT POINTS OF VIEW OR OPINIONS
STATED DO NOT NECESSARILY REPRE-
SENT OFFICIAL NATIONAL INSTITUTE OF
EDUCATION POSITION OR POLICY

U.S. DEPARTMENT OF HEALTH, EDUCATION, AND WELFARE;

Public Health Service
Alcohol, Drug Abuse, and 'Mental Health Administration
, ,
National Institute of Mental Health ,
V'
5600 Fishers Lane '
Rockville, Maryland 20857'
 . f - !
i.:.
The National Institute sof; Menial, Vealth.has obtained perniissiOn -from the copyright holders to reproduce certain quoted"
material which appearOni the pages indicated below. Further reproduction of these passages is prohibited, without:SpeCial
permission of the copyrightholder.All other material contained ifi this publication is in the publicdoMain and may be used and
'reprinted withoUt speCial permissipn. Citation as to source; however, is aPpreciated.

4-From The Collected Paperi of Sigmund- Freud, /ions by Harry Stack Sullivan, 1974. Reprinted by
"Volume 2, edited by Ernest. Jones, MD., authorized mission of W. W. Norton &
-translation and supervision of Joan RiViere, pp. p. 7-From -Psychotherapy: its Nature,' Assumptions,: .
354, 362-363. Sic Books, nt
blished by Basic I., New " Limitations by _Hans: Prinzhorn. Translated by
York; 'by 'arrangeiient with The Hogarth Press Ltd. .ar Eiloart. Georg Thieme Verlag, Stuttgart, .Gerinany,
and the Institute of Psycho-Analysis, London. and ionathon Cape, London, 1932.
p. 4-From Client-Centered Therapy by Carl R. Rogers. p. .8-From Menninger, Karl: Whit are the goals of.Psy-
° Copyright. a 1951 by Carl R. Rogers. Reprinted by chiatric education? Bulletin of the Henninger Clinic:
permission-of Houghton' Mifflin Co. 1952, 16, 153-158.
p. 4 and p:7-Frozia the Collected Works of C. G. Jung, 16-From American Journal of Psychiatry, 1)1; pp.
editPdby Herbert Read, Michael Fordham, Gerhard 321:331, 1954. Copyright -1954, the- American, Psy-
Adler; William McGuire; -translated' by R. F. C. Hull. chiatric Association-Reprinted by permission. ;
Bollitigen Series XX. Voluine Civilization'in, Tran,- 74-seethe first acknowledgment above,
P.
sitian p 1964, 1970 by Princeton University Press. 74-From Principle: of intensive"Psgchotherapy by
Selections from pp. 159-163, 164 reprinted by per- Frieda. Fromm-Reichmann. University of Shicago
..
nuision. Press, 1950.
6-From Persona/ Psychopathology;. Early Fo

-
. . i ,'
pp. 17, 22, 36-39, 48, '66-69, 75, 117.127, 144-147;162i-1qt Certain items are adapted and reprinted
from the Strong Vocational Interet Blank for-Men (Revised), -.Form M by Edward K.
Strong, Jr. with the permission of -tl' e publishers, Stanford University Press: Copyright © .
-1938, (renewed 1965), 1945 (renewed 973) 'by; tile, Board of Trustees of the Leland Stanford
. Junior University. 4t

L.
DHEW Publication No. (ADM) 77-527
.

Printed 1978

iFor sale by, the Superintendent of Documents. U.S. Government Printing Office.
. Washington; D. C. 20402
Stock Number 1)17-024-00688-6
 y,
.

the Na ioial Irfsti te ,of Mental a Ith. 14 been intereste 'in the ef-
7 fectivene s of vitii s..ways 0::4galint,wit:11.It 4:entai, emo 19 nal, ` and
behavioral disorcters. It never, there is a.C4ntiiinpit tr,Oblern:i4 e luation in
that it is sometimes difficult: to deeeribe acc*atelit:*114.,g....: ., on in therapy. I
have adopted an approach picineered iiy.,,whitiiirrk.,. .'*01 Iii`rty stu-,
dying t e personality ottheinterte*t,-,ive`migh 'e to cl:What is
being done, and thereby: .assella tlieeffeetiveness interven-
-The itud'vekare piesented'chron logical*, and s
the reader to s4e clearly where:the
went and where, indeed, they aiveltr to:it have, worl ed we
IJ "we" in chapter V; ithwever, is not editorial; that chapter ref*
tion'with Qeorge Purse in thestudy of juvenile delinqueney.'.:;:.
this .
was,: prepared .With two audiencesin "mind, Pei
.

7 primarily interested in practice willwant to read chapters I.

. ,
.
.
,
and` they may also want to' look at appendices 2 and 10. Technica eiiallia
been.kept to a. minimum in these portions of the report:
i.
the siiiiyeike-.04ian
, are reported in chapters III,iIV, and V. and research considefOiOna'lare-...
discussed in chnpter VII and in appendices 1 and .0 to 9 : '-'L!:': I'
A number of critical issues are considered, mime of them in,et edolOgio;
Others .substantive. One recurring question concerns the eXtentitb which. its;
may be possible to specify: good psycho-aocial therapyin general. Or must,-e, 'w,e:°!,5
find the psycho=social therapy which is appropriate to each type of human
problem? On this issue will .ultimately depend the optimal designs fOrjniental
health serviceS, for mental health research, and for mental health training...:
 Acknowledgments
I

The research reported

. here vroild not have been possible without the
tireless work of many people, most of 'whom received no monetary cpmpensa-
. tion for their 'work. First of all the e are the respondents who must remain
na eIess. But,I wish to mention the staff at the Loysville (Pa.)
,
Youth Develop-
me Centel. who not only responded to three questionnaires, they also filled
ou innumerable rating 'sheets and dia ostic forms, frequently' taking ,
these
ings home to work on them.
Vivian E. Dent, my wife, assisted in, the design, the data collection, and 1

the analyses, and reviewed- the man cript. Jonquil- Furse helped_;vith-.the
design of data collection at Loysville. Lynne Hottle assisted in some of the
analyses.,
Irvin L. Child made availab e to me all his questionnaires and a number of
unpublished papers,. The impor ce of his work is .discussed below (pages 21,
108)., But he' must not 'be held .r sponsible for what I have 'done with his
materials, as indeed is the case for others whose help is acknowledged here.
Daniel.Katz commented on this research at various stages in its develop-
ment and provided valuable criticisms of the manuscript. Victor Fields, Jill
Rierdan, and Mary Philipp alio contributed- valuable comments.
The interest that Barbara J.BOz took in this work is attested to by the -

frequent references 'below to ideas and hypotheses running far beyond her
published works: Joseph H. Stephens made available 'from the files of the
Henry Phipps Clinic large amounts of data that were-essential to the work that
is pursued here. . -., --- . .
The late John C. 'Whitehorn spent many hoUrs talking with me about this
.)
area of research, how it had developed, what pitfalls it presented, .and-'what
promise it held for the future.Aniong my deep regrets is that his work was not
sufficiently appreciated during his lifetime. I would have liked particularly for
him to have read this report. .

,, Y
#
Many perions assisted in the collection-of the -questionnOres. They are
nameless since .I do not wish to indicate whence the questionnaires came. In4td- 'Y
dition to the direct mailings, questionnaires were ,collected in a variety of
clinical settings. The cooperation of Victor Fields was essential at the early
- stages of data collection.
'Essential sta tistical and computer assistance was provided by John J.
Bartko, Berthold Brenner, Ray Danner, Kirk L. Dorn, Wayne E. Johnson, Nils
B. 1Vitttsson, Charles P. Pautler, Jr., Donald S. Rae, Robert R. Rawlings,
Markaret T..Rop'er, and David Vansant. Valuable comments on the manuscript
were received from Morris Parloff, Milton F. Shore, Karen. D. Pettigrew, and
Irving D. Goldberg.
MY coauthor in chapter V, George A. Furse, labored long and effectively
in the design and execution of the large study of juvenile delinquents. Though
only one chapter in this report, it is a capstone of the research structure.
 Table of Contents

Preface
Acknowledgments iv
Abbreviations and Coded Variables x

Overview 1

Introduction 3
Why Study the Personality of the Therapist? .4
The Good Therapist vs. the Appropriate Therapist A 7
The Differential Hypothesis 8
Why So Few Studies? 10
Theoretical Summary' . 11
Diagnosis 12
Is psychotherapy Effective? 13
Whitehorn and Betz: 14
he "A-B Scale" 17i,

i
The First StudyA Pilot Investigation
1
.
The Second StudyHospitalized Schizophrenics,
Depressives, and Neurotics / 6
Study Design ';' 26
Design of the Personal Teridencies Questionnaire 26
The Respondents , 28
Female Therapists , 29
Psycho-Social vs. Somatic. Orientations 30
What Is theMeaning of the "A-B Scale"? 31
Further Analyses of the Phipps Clinic Data 34
The Phipps Data Drug-Free Relationships , 35
PTQ Correlates of the Drug-Free Relationships 45
The Phipps Data Psychotherapy Plus Drugs for:-
Schizophrenics li 46
i
PTQ Correlates. When Drugs Are Prescribed for ,. ; ,
Schizophrenics
V. The Third StudyJuvenile Delinquents (with George A. ..,

'Furse, Ph.D.) . , 54
The Loysville Youth Development Center .... . .. . , . - q.
54
.. 55
Study Design
The Problem of,Therapeutic Influence in a
Milieu Setting , .56
Yorith Measures * 57
HelperMeasures 60
Helper,se, Persdnalities , /1
61
 CONTENTS

VI. Clinical Interpretation 78

Netirotics 73
Schizophrenics 75
Pbrsonality Disorders 76
The A-B Functional Reversal 78
The Functional Psychoses 79
The Depressives 79
The Generalist Approach and the Differential
Hypothesis 80.

VII.. Implications for Research 83

Research Strategies in Exploration 83
Limitations of the "Biomethodology ' 85
Specific Issues 90.

VIII. Summary 94

References 96

Appendices
-1 Statistical and Psychometric Considerations 102
griterion-Based Testa ,-
, 102
Hypothesis-Testing in Exploration , 103
Measures of Association Used in The'se Studies 105
Clustered Samples 106.
Construction of the PTQ 107
Social Desirability 108
Internal Consistency.of Scales 109
2. Possible Personality Correlates of the "A- 1' 111
General Characteristics of the A-B S e '-.
--,-,..
111
Possible Meanings of the A-B Subscales ., 142
Significance of the Issues 114
The Personal Tendencies Questionnaire 116
4. Personality Scales in the PTQ 128
- 5.. Supplemental Analyses of the Phipps Clinic Data 143
6.. PTQ Analyses Of the Phipps 'Clinic Therapists 150
.7. Additional Analyses Second and Third Studies 152
8. Somatic vs. Psycho-Social Orientations 161
SVIB Correlates of Drug Prescribing 162
Treatment Ideologies 1
9. Multivariate Analyses ...... , 16
10. Commonwealth of Pennsylvania Diagnostic Classifications ..,---

of Delinquents - , 169
Index of Names. 172
Subject Index , , 175
 e

-"' CONTENTS

List of Tables
;b) lipi 1. Strong Vocational Interest Blank Items'Which Differentiate
f, ',
1 ,
, .
A and B Therapists
2. Intercorrelatibn of Selected Scales
3. Correlation of tke,--A-B Predictor (18. items) With.Items in
Other Persons** gales
4. Correlation of theleore for 8 A-B Items *Which Imply
RejeCtion of Mechanical Interests With Iteins in Other
Personality Scales Among 31 College Men With High
Empathic Infeiest
5. Correlation. of the Score fdr 6 "Other" A-B = ms With -Itenis
in Other Personality Scales Among 3 e MeR With
-High Empathic Interest ,.

6. Scales Included in the Revised Personal Te c_ ies \

Questionnaire ....... i . .... .. . .. ... .. . ... . .... . . 27
7. Male Respondents to theiRevi ed P Tendencies
Questionnaire . _,- 28
8. PTQ Response Rates for Former Phipps Clinic Residents 31
9. Correlation Between the WB-22 A-B Predictor and PrrQ
Person4lity Sales 33
10. SVIB Items Correlated wits Therapists' Percent
Improvement in Patients Who Are Schizophrenic,
Depressive, or, Neurotic 36
11. SVIB Items Correlated With Therapists' Percent
Improvement in Patients Who Are Schizophrenic - 37
12., SVIB Items Correlated With Therapists' Percent
Improvement in Patients Who Are Depressives ) 38
13. SVIB Items Correlated With Theripists' -Percent
Improvement in Pa4. ollts Who Are Neurotic 39
14. The Criterion Variab 'rePercent Improved of
Schizophrenic, of Depressive, and of Neurotic Pitients -
Intercorrelated With 'the A-B Predictors. and the A-B
Clusters ., '40
15. Intercorrelation of the A-B Clusters u . , 44 '
16. Personality. Correlates of the A-B Predictors and . -

Clusters, Druk-Free Relatjonships

17. SVIBittems Correlated With Therapists' Percent
Improvement in Schizophrenic Patients for Whom They
Prescribed Drugs
18. Interrelation of Criteria 'and.A:-B Predictors (or
Therapists Who Prescribed Drugs
19. Personality Correlates of Two A-B Predictors for
Schizophrenics 52
20. PTQ Respondents, Hospital andlionhospital 53
21. Diagnosis, Improvement, and Severity of Youths Included
in Analyses - 59
22. Intercorrelation of ,Helper Improvement Scores for Six
Diagnostid Groups 61
trelation of sonality Scales With Helpers'
Improvement cores:fOr Six Diagnostic Groups 64

L,.
viii
Orrehition:of A-BPredictors With .HeOei.' Improvement
. Scorei for Six Diagnostic Gi,oiipa -.,..'./,,....: . : 7 .',. ,'. . ..... ..... . . 5/
. .''SVIB Iteriis Correlated With' HelPers Adjusted u :'i

Improvement Sbores for Yoitli Why Are .N.2.crotic -- - .- - , - P' 66

26. : AVIB Items COrrelated With- Helpers Adjusted , :1.. . ,E.
,
,
Improvement Scores ftir. Youth Who Are Mampulators . . . . i . .
27. SVIB Items .COrrelated With H4Peis'"2444juated Improvethent
Scores. for Youths WhO Are:.-Subcti.,Itiiia,1 Identifiers . . 4 . ''. ..-.,. . 69
28. ,SVIB Items Correlated' With, Ilelpi3ra' Adjiist:ed
Improvement Seores-fOrYoUths'WhOAre ,S,tv.ationciii . ,. .> 70/'
29. Quay Items Correlated:. With: Helpers'. Adjusted :

Improvement Seores.for.Youths/Who Are Neurotics

and Subcultut;c4s . ....:... ,r ...: . . . .:i : . . . : . .0: .. : ::. ;2 : . . : 71
30. Quay Items Correlated With Helpers' Adjusted :. .
Improvement SOores for Youths Who Are Mitnip*(40a:.
and Situationa/4(:. . . . . . , , . . . . . . . -... .. :. i. , .
1 4 ,
, : .. 2
1.
..

Appendix Tables

31. The Need,for Closure . . . . : . . . ... . . . .

/
. . 128
,
gr
32. Reession in O le Service of the EgO . . ., ./4 ., . .... 129-
33. EmPathic Interest ' I ,. .. . . 131,', ,

34. %The Work Ethic , .. .

.- . .:: . .*. . . ' 132
4. The Need for Order . . . . . ; ..... :. . .' ... . . -.. , 4 r . ..' . : 132
. .

36. Tolerance of thd UnreatiStic Experience . . . . .

37. Tolerance for Ambivalence. , .. ... . . ... : . . . k 9 .. ..... 134
38. Preference for. Decision Making .. -. . . . . . . . . ? ... .' 184
39. Personal Standards . . 135
40. Remorse . . . . ...... . . .. . ; . . .. . ...... . .. .. . . . . ...... . 135
41. Antidemocratic Scale ..,,. . . .,. ...... . ..... . . . . . .... . ...... . 136
42. Individualism .. . . . .. . . . . . . . . :.- ..... . : 138
43. Deference Anxiety .... .. . . : .. . .. . . . . . . . . . . . ..... . . . ....... . . 139
44. IsTurturance Anxiety . N, 139
45. Science and Moral Values 4 139'
46. Preference for IntelleCtual Challenge , 140
47. Extroversion 140
48. Fervor 141
49. Life Satisfaction ,f, 141
50. Work Activity ... , 142
51. SVIB Items Correlated With T,,herapists' Percent
Improvement in Patients Who. Are Schizophrenic, ,
Depressive, or Neurotic : . . . . ..... .. . . ,. . .
. . . . .. . ... . . . . . . 144
52. SVIB Items Correlated With Therapists' Percent
Improvement in Patients Who Are Schizophrenic . . ... . . . . 145
53. SVIB Items Correlated With Therapists' Percent
Improirement in Patients Who Are Depressives' 146
SVIB Items Correlated With Therapists' Perent
Improvement in Patients Who Are NeurOties ,. 147
55. Phipps Clinic Therapists-Means andStandard Deviallions ...... 149
56. Intercorrelation of the PTQ flersonality Variables
Nonhospital Psychiatrists and Residents 153
.1. ,'
 CONTENTS

; 57. Intercorrelation of the 'A-B Predictors, Drug-Tree

;

.
1Relationships, Nonhospital Psychiatrists And; Rekiients ., ... .. ... .., 154., .

58. Personaliiy Correlates 'of the' A-B PiledictOrs- and Clusters,

1

Drng-free'Relationships, Hospital Psychiatrists ..

ind.Resident . . ...... . ,..... . . . . . .: ... . ...... . .,.. . . ..... . ...

, . .

psychiatrists and Residents

60. PTQ Variables Psichiatrists and Residents . .,. . . ....: ....... .
a.
-.

61. PTQ Variables Selected Groups . ... . . . .'. . . . . ..1. .,. . .,. .. ... .f-...
.

62., ./. Intereorrelation of the PTQ Personality Scales, Loysville .. ..... . . .

63./ SVIB Items Correlated With Doctors' Prescribing Drugs
.
.

04, Varirnax Rotation of Principal Components , -B Clusters '

. .;
for Their Schizophrenic Patients . . : . . ;.A ... . . . . . . . 163'

' and PTQ Personatity Scalesa"sy:ChiStrists 'and Residents ... . . . 17

.

List of Figures
Sehemati Diagram Criterion vs,. Correlational Studies .., . . .
Hypothetical Correlation Between Improvement Scores for
: c. ,16
,,.
....,
),
-;1. -,

Two Diagnostic Groups Showing Possible Effect oil. : .,-, . .. .,

Eliminating Unsuccessful Helpers .. . . . . . .. . . \62 4 ---

EffectIveness Patterns for Various Disorders ..... . . .... . . . . . :I:77 .7'''
4,,,.

Phipps Residents' Correlations at Two Time Periods

.
...,5,...,
.1 ,

rl

"es
.
,

"? A.

i
 , For group A (effective therapists) takf,group/B(adfeei'effective):-.A-B)s,
attached to any variable thatrelates to effectiveness_ It oae-or,amither
of the mental or-behavioral disorders. ;

"A-B Scale" More properly an A =B Predicto , thistefects-to231tetas from the

Strong Vocational Interest Blank Whit sere found bry. yOlwahni.,0 and
Betz to relate to therapists' effectijOenieSs1WithTsChiz Irtienice;(Pee:Page
17).'Also called '"W-B 23."
A.B. Predictor AnY collection ofiteias thatiii related iiAcess,Withi.parti-
ctlar diagnostic group. A- predictor usually it con=
.

,
tains more than one'A-B cluater.," .1 _-
A-B Cluster Any group' of items within -a single Az:B predictor Which -arejap-
propriately,/intercorrelated and can., be Presumed telieiong .to a single
dimension .or personality characteristie. (in.appendix:2; A7-8- Chuiters are
called "A-B Subscales.");
ABDRUG
, . /,. -.
predictor -forAerapists' success with schizophrenics,
when drugs are prescribed; (See page:48J This predictor 'Is comparable.
.to TOTI,iS among the drug-free predictors, ,_ - ,- r.:
Alpha A-coefficient ranging fraia 0 to 1, and measuring the degree of internal
e

sistency (unidimensionality) in a group of items. Usually calleid '"ceef-:

ficltenAlpha." (See Page 109.) 4

BC-.1, BC-2, BC-3, and BC-4 See below: Quay 1, Quay 2, Quay 3, andaQuii3t 4
D SVIB items that correlate with effectiveness with depressive and only
with depressives, and are' included in -the PTQ.- This is 'a.';:dePresOve- '
specific, drug- free predictor. (See page 42.) _',,' , 1 )
DBUSNS An A-B cluster within D. These items, all related to.bariitieSe'par-
suits, are rejected by therapists effective with depressives. (See-page43.).
DCIVIC An A-B cluster within D. These items seem to repreient a broad
social concern and a rejection of rugged individualism. (See,page 433,
DN Items that correlate with effectiveness With both depOssives and
neurotics. It is a weak, drug-free predictor. (See page 42.) .

D OTHR Items that correlate with drug-free effectiveness with' depressives

but found too late in analysis to include in the PTQ. (See page 42.)
CT Electroconvulsive therapy. 7., ,,
- r, , . ,

F Scale "Authoritarianism," items that me ure ethnocentrism indirectly.

(See "Antidemocratic," page 136.) . .

Form C A second revision of the PTQ. (See. appenclix

Gamma A rank-order correlation coefficient,which can be interpreted as the
proportion of nonreve 'sags in rank. (See page 1053 ' , .

MMPI Minnesota Multiph sic Personality Inventory'. A personality teat con-

cerned primarily with psychopathology.
MP Manifest personality scale. Items in such a scale are measuring ,,what
', they appear to be measuring. By contrast, in criterion-based Scales (A-B
predictors and chisters, the F Scale, and the Independence-of Judgement
Scale) the items. may not look at all like the dimensions they measure.
.
 id3BiltylATIONS

MPRV. Criterion of theripist effectiveness. In the Phipps data, it is the per:

cent of patients improved. In the LOysville data it is an average of ratings
of improyement by judges and helpers. A letter or number is added to
PRV to indicate the diagnosis ()Ulm clientsio which it was computed.
(see pages 41 and
.

MPRV D Percent improved of depressives (Phipps Clinic,"- 191'

MPRV Percent improved of neurotics (Phipps Clinic).
MPRX, S Pereentimproved of schizophrenics (Phipps Clinic).
MPRV 1 Average rated improvement in neurotics (LoYsville).
MPRV 2 Average rated improvement in situationals (Loysville).
MPRV 3 Average,rated improvement in subculturals (Loysville).
MPRV .10 Average rated imptiayement in Manipulators (Loysville)..
MPRV 5 Average rated improvement in conformists (Loysville).
MPRV 6 Average rated improvement iii asocial (LOYsville).
N 'SVIB items that eorrelate with effectiveness with neurotics and &Ay with
neurotics, and are included in the Pt'Q.' This is a neurotic-specific, drug-
free predictor. (See pa e 421
N OTHR ms th rre ate with drug-free effectiveness with neurotics
.but'we e o tqo late to be included in the PTQ. (Seepage 43.)'
NSOLVE An A-B. cluster within N. These items relate to problem solving
' and are rejected,by therapists effective with neurotics. (See page 444.)
NQEST1 An A -B. cluster*within N. The items are heterogeneous but include
a liking for literature. (See page* 44 and 73.)
NQEST2 An A-B cluster within N. The meaning is unclear.
PTQ Personal Tendencies Questionnaire (See appendices 3, 4 and 8.)
QUA Y1 Quay-and Parsons classification for "Inadequate-Immature?' Specif-
ically, Quay 1 stands for a 15-item scale that is supposed to predict success
with this type of delinquent. wit .

QUAY 2.. Similar to QUAY t for Neurotic-Disturbed."

QUAY 3 Similar to. QUAY 1 but for "Unsocialized-Psycyoathic."
QUAY 4 Similar to QUAY 1, but for ,'Socialized-Subcultural."
r The Pearsonian product-moment correlation coefficient.
S SVIB items that correlate with effectiveness with schizophrenics, and only
with schizophrenics, and are included'in the PTQ, This is a schizophrenic-
.. specific, drng-free predictor'. (See page 42.)
SD SVIB items that correlate
. with drug-free eggctiveness with both schizo-
phrenics and depressives. (See page 413
S OTHR SVIB items that correlate with effectiveness with schizophre.nics
but were found too late to ,be included in the PTQ: (See page 423
UEST An A-43 cluster within S. These items are heterogeneous. They
may, possibly represent'extroversion. (See pages 43 and 47.)
SXPRSA An A-B cluster within S. These items have in common a liking for
'active, involved, social expressionAeee page 43.)
iSXPRSS An A-B cluster within S. Social expression, but not so active and
involved. (See page 43.)
SVIB The Strong Vocational Interest Blank. In the present studies it refers
only to the 400-itein Form M.
TAT The Thematic eppereeption Test; a projective personality
.
test..
TOTL D° All those SVItl items in the PTQ that correlate with drug-free- ef-
fectiveness with depressives.
TOTL D ..'D + SD :)- DN.'(See page 41.) , \ -

fy
 ABBREVIATIONS

TOTL N All those SVIB items in the-PTQ dad correlate with drug-free ef-
fectiveness with neurotics.
TOTL N m N + DN. (See page 41.)
TOTL S All those SVIB items in the PTQ that correlate with drug-free'effec-
tiveness with schizophrenics.
- Toni S = S + SD. (See page 41.) .
V.A. Veterans Administration, specifically a stAdy done in outpatient clinics
by McNair, et al.
W-B 23 The 23-item "A-B. Scale" of Whitehorn and Betz. Specifically, the
variable consists of the original scores-for the Phipps Clinic residents as
computed by Betz on the system indicated in table 1.
W-B V2 A score, coniputgd from the PTQ for 22 of the 23 items and highly
correlated with the W-B 23. (See pages 32 and 106.)
YDC Youth Development Center, specifically the one at Loysville, Penn-
sylvania.

13,
 Chapter I
Overview
Studies of the effectiveness of tlie psycho- committed to these studies are miniscule com-
social therapies have been plagued on the one pared to many other studies of psycho-social
'hand by "null' findings (no effects, no dif- theripies. Of course, verification of worth can
ferences, etc.) and on the other with inconsis- come only with repeated studies that. are
tent findingo(findings that don't "add up"). The ,carefully evaluated.
reason may be that the various psycho - social The studies 'to be presented also, illustrate
therapies are not themselves clearly defined comparative research strategies (comparing
and are rarely themselves 'measured" when neuroses with the schizophrenias, etc.) as. Op-
research is done on them. It is proposed that posed to Categorical studies (of a single
the therapies be defined by the personality disorder). Again it is not asserted, that com-
characteristics of therapists, and that the arative studies are' superior, 'but only that
salient dimensions of these therapies be de- this approach has probably been undervalued
. -fined by personality characteristics of effective in the overall distribution of research efforts.
therapists. Evidence of the validity of these The .comparative approach facilitates con-
premises is summarized in, chapter II. sideratipn of one issue that has wide ramifica-
This is not to suggest that studies of tions in the definition of sejvices, and their
therapists' personalities are more. useful than design and evaluation. This issue miliht be
studies of therapists' behaviors or of thera- posed: To what extent do .the various mental
peutic processes. Rather, the various 'ap- disorders require similar therapies (the
proaches.are viewed as complementary. °It is generalist approach) or .does 'each disorder re-
much cheaper to study therapists' personali- quire its own specialized therapy (the differen-
ties than: to study their myriad behaviors. tial hypothesis)? The findings of the studies
Moreover, understanding therapists' per- presented here, as well as of the other studies,
sonalities attaches meaning to their behaviors. are' very much in support of the differential
Such understanding gives us broad hints about' hypothesis. In this way, the comparative ap-
what therapists do and it suggests what to look proach serves to define the generality of find
for in iherapeutic processes. There are, unfor- ings from- thercategorial studies. The differen-
tunately, in many' clinical settings, rooms full Of tial hypothesis has widespread implications for
tapes and films that'have never been analyzed. the design of mental health programs, but
Simply defining what we should look for is a Much more research is needed before these im-
prodigious task in itself. plications can be elaborated with' confidence.
Accordingly, considerable attention is given The comparative approach also helps us to
in chapters II and 'VII to the probable cost-ef- define "regions" or -"rubrics." What are the
fectiveness of various research approaches. limits of the region called "mental disorders"?
Guidelines are set forth to help-us 'in 'exploring Findings to 1presented suggest that there is
uncharted areas. In the absence of research on considerable overlap between the region called
comparative research strategies, no conclusions "mental ',disorders" and the region called
are possible. But the thesis advanced that "delinquency.7 While this overlap has been
research purposes sometimes require simPle, recognized for many decades, there has been
naturalistic designs. Such designs may only little progresain defining how the two regions
be more cost-effective, but ibly more effec- differ, or whether,' indeed, 'there are two
tive ill absdlute terms, than e complex, cost- regions or many. The present findings suggest
ly, contrived designs. The studies reported that j venile delinquents ate even' more
here illustrate such simple designs. It is impor-. heter geneous than are the mentally
Cant to, emphasize that the resources that were disord red. There is no one correct approach to
1
 EXPLORING THE PSYSHO-SOCIAL THERAPIES

delinquents, but rather an- ap opriate ap- Y reversal of the "A-B Scale" are supported in
proach to each type of Idelinq t, some of the various waysAndeed, the findings clearly sug-
approaches being in opposition to each other. gest that the semantic reversal probably _ex-
Great progress has beiOn made in recent years plains many of the inconsistencies that have
in the di ferential treatment of juvenile delin- been observed in the various studies using the
quents. "A-B Scale," and that the functional reversal is
The e studies to be presented are concerned unlikely to be found with consistency unless \,
w'th ne-to-one therapy. The social nature of the therapists and patients are drawn from
su "psychotherapy" is emphasized and sug- certain defined.populations. (2) A new scalein-
gestions are made that certain characteristics volving the rejection of problem solving is
of one-to-one therapyiewile generalize to milieu found to characterize therapists who are effec-
therapy and to other psycho-social therapies. tive with neurotics both ,at the Phipps Clinic
The present research grows out of the work and among neurotic delinquents at tilt?
of Whitehorn and Betz with schizophrenics 'at Loysville Youth Development Center. This is
the Henry Phipps Clinic of the Johns Hopkins taken as evidence in support of Freud's con-
Hospital. The findings serve to clarify their cern that the' neurotic client be allowed to
work and to explain the anomalies in findings solve his own problems. (3) Another new scale
of others who have tried to work with their suggests that therapists effective with
"A-B Scale" predicting success with schizophrenics take an active, involved in-
schizophrenics. New light is shed on the "func- terest in the patient. 'This supports the find-
tional reversal" of this scale7 the fact that it ings of Whitehorn and Betz that schiZo-
tends to reverse as , we pass from none phrenics require active, 'participative leader-
diagnostic group to another. In addition, there ship. (This requirement may not hold when
is' new evidence presented that the scale tends drugs are prescribed.) (4) But regardless of
to reverse its meaning (the "semantic rever- whether drugs are prescribed, evidence is
sal") as we pass from hospital to honhospital presented that schizophrenics must be treated
therapists. with tolerance and understanding.
While many statistical findings are There is considerable attention to the
presented, some of them must be viewed as assumptions underlying this research and to
tentative, pending replication, because the the implications of 'the findings ;for' future
methods that are used are not established research. Of particular importance is the
ones. However, there are some findings' that careful definition of the population of- clients
are :.replicated in more than one sample. (1) and the population' of therapists that are being
Both the functional reversal and the semantic studied.

4.

r.
ChapterI1
IntroduCtiOn
These studies will be reported chronologi- (3) a definition o appropriate intervention
"cally-,r "as they happened." I am concerned not (free association) These three types of
Only ith what we learn but also with how we variables outcome, control variables, and in-
learn. A chronological presentation makes. it . terve,ntiOns are presented in detail else-
more clear how the knowledge was gained, .where (Dent, -1966).
errors were made, and how best to pro-. Progress; has been made in the definition of
ceed ith further 'studies. .
outcomes, sometimes calls t . criterion
variables, or dependentVariable0Waskow and
.

Thvi lines of inquiry what we know, and Parloff,' 1974; Berzins, Bednar, and Severy,
how est we learn are interwined. Some ef-
fort rill' be ,made to separate them, for some 1975).' With respect to the control variables,
readers will be conterned primarily with the the most important seems to be diagnosis.
clinical substance,: not the methods and While progress here is slow, new 'distinctions
strategies of research. Insofar as possible, and new criteria are continually. emerging. In-
technical matters are relegated to footnotes dividual, .diagnoses continue to be somewhat
and appendices in order to maintain they con- unreliable, but diagnoses derived from a 'con-
tinuity of clinical content. sensus of several sources have been found to
Unfortunately, there' are some issues, of be very, useful in statistical -analyses of the
substance which turn upon issues' of method. mental disorders. The importance Of diagnosis
For example, some people are convinced that can be seen by looking again at the illustration .
the psycho-social therapies are ineffective... in the last paragraph. If we subStitute "phobia"
This important issue requires some considera- for "repression," then "free association" is no
tion of methods in order to ,interpret data that longer the 'appropriate intervention. -

are presently available. But first we need to The word "evaluation" is used ambiguously -
define where we are. in clinical practice. Frequently, it means
My interest in, the effectiveness of the diagnosis rather than evaluation (the actual
psycho-social therapies arose from an interest measured outcome). Insofar as diagnosis car
in the organizational design of effective mental ries a prognosis, evaluation may be' implied.
health service programs. I came to the convic- This confusion of two concepts which may be
tion a dozen years ago that we could not design statistically related. but are conceptually dif-
or evaluate mental health programs very well ferent contributes to the pejorative, dangers of
unless we underkood the effectiveness of the diagnosis and inhibits our understanding. of ef-
elements of those programs. While there 'is fective and ineffective treatments. If a process
some evidence to support this conviction, it is schizophrenic recovers, many clinicians do not
not the kind of proposition that lends itself to look for the reasons; they simply conclude it was
ready proof. In any event, .I set out to study the not process schizophrenia. Small wonder that
effectiveness of elements of mental health ser- psychiatric diagnoses are not respected:
vice programs. I soon found myself in trouble It is with respect to the interventions (the
here, for the elements were not well defined, at therapies, the independent variables) that the
least not with the. kind 'of precision that a situation is most confused. On the one hand, we
researcher needs. now have precise definitions of certain kinds of
For almcist any kind of problem it is useful to drug interventions and these interventions are
have (1) a statement of the desired outcome specific to certain diagnoses. On the other
(reduce anxiety), (2) a definition of the problem, hand, with respect to the psychosocial
sometimes called a :diagnosis (repression), and therapies,' we use such words as "psycho--
, . 3

1t;
 EXPLORING THE PSYCHO,SOCI L THERAPIES

analytic," "nondirective," "milieu," and so corrected by the client if ijdoes enter (195I,
forth. As descriptions or whitt therapists ac-
.
P. 42).
) .

tually do, these words are not at all precise. Thus both Freud and Rogers with their very
.

One solution to this prOblem is to attempt to di ferent techniques hold a common view that
describe the personality of the therapist. This to hnique should and can prevent-the intrusion
chapter is concerned with the rationale and the
if
empirical evidence supporting such an am
proach, and with the issues tha must be dealt
of a therapist's personality into the "thera-
p utic process. It is a reasonable generaliza- .

ti n that most cognitive. therapists, behavior'

With in defining the psyCho-social therapies. th rapists, and behavior, modifiers would agree
I wi h this position. In. contrast; Carl Jung wrote :
1

in 934:
Why Study the Personality
of. the Therapist? It is in fact largely immaterial what sort of tech-
nique he uses, for the point is not the technique '
but the 'person who uses the technique ...., the
Why study the personality of the therapist?. personality and, attitude of the doctor are of
This question is particularly appropriate since supreme importance whether he appreciates
two of the giants of psychotherapy, Sigmund this fact or not .... the Freudian school asserts
Freud and Carl. Rogers, tended to view the per- that this has nothing to do" with psychoanalysis.
Psychoanalysis is evidently .a technique behind
sonality of the therapist as something to be which the human being vanishes, and\ which al-
kept out of therapy. Both of them emphasized ways remains the same no matter who practices
technique.. it. (1984, p. 159-163).
Freud wrote in 1913:
I adhere firmly to the plan of requiring the pa-
I. Therese Benedek reviewed the
tient to recline upon a sofa, while one sits behind tra itional view. Of the countertransference,
him out of his sight .... Since while I listen, I re- na ely: by emphasizing technique and sitting
sign myself to the control of my unconscious the patient, therapists can keep their
thoughts I do not wish my expressiorao give the ow personalities out of the therapy. She ques-
patient indications which he may interpret or, tioned that this was possible, and.after giving
which may influence him in his communications
(1959, vol. 2, p. 354). many examples, concluded that "a therapist's
perSonality is the most important agent in the
In 1951, Rogers wrote: therapeutic process" (Benedek, 1953, p. 208).
There can be no doubt that every therapist, As we shall see below, a formidable array of
even when he has resolved many of his own dif- clinicians and clinical researchers (Menninger, .

ficulties in a therapeutic relationship, still, has Prinzhorn,.. Riernann, Strupp, and Sullivan)
troubling conflicts, tendenciei to project, or un- agree with Jung and Benedek that the per-
realistic attitudes on certain matters. How to sonality of the therapist is a critical variable.
keep these warped attitudes from blocking But the disagreement is more apparent than
therapy or harming the client has been an im-
:portant topic in therapeutic thinking. real. 'To deal with it we must understand the
In client-centered therapy this problem has relationship among the following three sets of
been minimized considerably by the very nature variables: (1) the fornialized or idealized tech-
of the therapist's function. Warped or unrealis- piques, (2) "this actual behaviors of individual
tic attitudes are most likely to be evident therapists in therapy, and (3)' their per-
wherever evaluations are made .... any sonalities. Of particular interest is a study of
theripy in which the counselor is asking himself
"How do I see this? How do I understandthis therapeutic styles which was done by Marylou
material?" the door is wide open for the per- Lionells in 1967.
sonal needs or conflicts of the therapist to dis- Lionells built her conceptions of therapeutic
tort these evaluations. But where the coun- styles on the works of Gilbert and Levinson
selor's central question is "How does a client see (1956), Sharaf and Levinson (1957), Hollingshead
this?" and where he is continually checking his
own understanding of the client's perception by and Redlich (1958), Strauss et al. (1964), and
putting forth tentatite statements of it, dis- Sundland and Barker (1962). She was par-?7- -...-
tortion Wised upon tlig counselor's conflicts is titularly influenced by the -work' of Hans-
much less apt to enter, and muchmore apt to be Strupp (1955 and 1960). Strupp had foUnd that '
 INTRODUCTION
,
he was unable
. to. discriminate among pectation that we will.SOmeday find `that cer-
therapists' behaviors on the basis of their pros tain kinds of psychotherapy are less akin to
fession or their theoretical .school. He conclud- other kinds of psychotherapy than they are to
ed that the chief differencea among therapists certain khids of milieu t herapi.4 someday;
_ with respedt, to their techniques are probably when we understand hoW various kinds of rela-
determined by their personalities (Strupp, tionships affect outcomes. Thus, although the
1960, p.. 265, 288,.'307). Lionells therefore set research to be presented all in the one-to-one
outto answer the question: What determines a context, it is' hoped that the relationships CA-
therapist'i style? Is it training (psychiatrist, plored can ultimately be differentiated within
psychoanalyst, psychologist, social worker), and generalized acros§ our' current psycho-
thls.oretical oriehtation (Freudian, Sullivanian, social therapeutic modalities. .

Rogers n, and so forth), or personality (as Understanding the appropriate relation-

) measuied by the TAT and a' self-rating ifiven- shiPs may facilitate generalization,- i.e., types
: tory)? of relationships may generalize more easily
What is. therapeutic style? This question re- than specific behaviors. The specific behaNiiors
quires a brief discussion. There is a great deal appropriate to a particular relationship may be
of concern with what acutally goes on in different in different modalities.
therapy. What does the therapist do? Such an Consider the' word "nurturantW This word
. emphasis serves the current-F-sa-cntffi-c----c an-be-conaehzed_to describe a iherapigt but it
Zeitgeist of behaviorism, and it also serves cer is also descriptive of one-half of a. social rela-
tain 'professiohal, needs that will be discussed. tionship. If the social relationship is reasonably
further below. From the standpoint of research stable, then the other half can be presumed to
Strategy the emphasis on behavior has been a be reciprocal: accepting nurturance. Consider
mixed blessing. While behaviorism has made now the myriad behaviori' that might be
.

our thinking and our research methods more observed which might be indicative of a, nur-
strecise, there have been two major failings. turant relationship. 'Consider also that for
First, there is a tendency to get lost in a mass Many of these behaviors, we would need to
,of complex detail that doesn't seem to add up knOW the age and sex combinations involved in
to anythink,,and second, there is a tendency tp order to be certain that they indeed indicated
sight a the fact that therapy is a relatioii- nurturance. It might be nice to have all this
ship: `It is described not simply by the behavior detail. On the Other hand, if nurturance is of no
Of thetheiapist and of the patient but also by significlince in the therapeutic xelatiOn-
the relationship betwein them. In fact, Sloane ship there is no present knowledge of
let al. (1975) found that' the relationship thia-- theU there is a question that we ought to
measure's were particularly potent predictors devote 'Our, energies to the detailed study of
of. Outcome.
t nurturant behaviors. If we can characterize
The cumbersome phrase "psycho-social. relationships' at a more global level, as a
therapies" is Used in order to keep firmly prieihninary test, we can determine Which
before tis the fact that there are many of relationships should be studied more
psychotherapies and they are defined not sim- i n4e nsiv
ply by what thelndividuals do,' but also hy.the The iniportant thing here is that therapeutic
reciprocal meanings of their acts, i.e., the rela- Style is' not simply a set of behaviors. It also.
tionships. between the individuals. All of the defines characteristic relationships ketwOn
psychotherapies, are not only psychological, therapists and clients.
they
the are also sOeial.- In defining therapeutic styles, Lionells
The cumbersome phrase has another pur- derived her.data from lengthy interviews with
pose, namely, to -avoid artificial boundaries 177 psychoanalysts, psychiatrists, psychbl-
that tend to be set between psychotherapy and ogists, and social workers. Factor analysis of
behavior therapy, between psychotherapy and specific therapeutic behaviors yielded five
milieu therapy, between psychotherapy and styles which Lionells labeled "egalitarian,',''
. various group therapies, and so forth. ,These "dogmatic "Itormalizing,': "pragmatic," and
boundaries tend to limit our vision. It is my ex- "authoritarian." Exaniples of the behavior
 6 EXPLORING THE PSYCHO-SOCIAL THERAPIES

items are "the therapist generally Mies free prescribed technique, we must reluctantly ac-
association" and "the therapist will sometimes cept the fact that we don't really know very
use confrontation techniques." Through factor much about what was evaluated, what thera-
analyses, these behaviors were clustered, a peutic styles were used in the therapy.
particular behavior being associated with one The optiMtim research strategy considers
or more therapeutic styles, or alternately, a not only what it is that we seek to understand,
style being defined by i claster of behaviors. b also hOw much understanding we currently
Lionell's analysis is interesting but lengthy har . We have little systematic knowledge of
and comPlicated. Sumniarizing, she writes: the ycho-social therapies. Under these cir-
"Given a certain persOnality configuration, the cumstances, we should explore: measure large
therapist will tend to adopt one therapeutic numbers Of variables as inexpensively as possi-
style over another, to a large extent, ble with a view to defining those areas where
regardleigiorthe theoretical school into which' further expenditure of resources will pay off.
he hits Veen indoctrinated and the profession The personality Of the therapist is more easily
- he hiss :chosen" (Lionel's, 1967; p. -247); While measured than behavior; personalit3i will tell
profensional .'training and theoretical position us a good bit about what the therapist dOes;
de influence style, they are relatively minor and understanding the personality of the
compared to the,influence'of personality. Thus, therapist will give us understanding of the
.

,if one wants to kn.aw more quickly and easily meaning of his or her behaviors.
/What a therapist does in 'therapy, one does not ' In 1932, Harry Stack Sullivan wrote:
.- inquire "about :training or therapeutic iden-:
4. ification. One inquires into his or her per- The self mediates in most attempts at communi-
' sonalitk cating information. Two people talking together
Perhaps we should not be §urprised that the say verbal combinations more or less entirely
therapist's style is related more to personality self-consciously. The two personalities inte-
than to training.' All psychotherapiSts are grated into the total situation within which this
self-to-self conversation is occurring communt:-..
,.' trained to be aware of themselves, to try tobe cate more or less, as it were, under cover of the
true to themselves, to be genuine. In a sense verbal interchange. A penumbra of personality-
then, we train therapists to expreiss them- meaning is attached to the culturally, standard-
selves. More importantly, since there is very ized words. It is.conveyed from one to another in
little sure knowledge of how they should relate themeasure that there is empathic linkage from
similarity of personality. (Sullivan, 1972, p. 830,
to .different kindS of patients, it is only natural . emphasis in the original)
that they should relate to them naturally.
Perhaps, at some future time when we have The logical consequence of Sullivan's posi-
sounder bases for telling them how to relate to tion is that we must approach the psycho-social
various kinds - of patients, we ill find that theraiiies through the personalities of the
is more important then rsonality. therapist and the patient, not simply because it
There is yet another way to vi w these find- is efficient to do this, but because it is the only
ings. When we speak of technique, we are talk- valid approach. The outcome of therapy is.
ing about prescribed technique, e.g. psycho- determined, not by what is said and done, but
analysis; What actually goes on in -psycho- by what is meant and .understood. Sullivan's
analysis" is quite, variable, Technique refers position is that meaning and understanding are
therefore to the prescribed behavior for a par- a result of a complex subliminal interaction of
ticular type of therapy; it does not describe the personalities.
behavior of therapists. For that, : we
actual, No attempt will be made here to appraise
muss conduct detailed observations,' or, we the positions that have been taken by various
may approximate that behavior by .attempting expert observers of the field. While the find=
to measure the therapist's personality. ;:' ings to be presented below shed light on some
Finally, and most troublesofne, if research- of the controversies (see chapter VI), we will
ers have evaluated psychotherapy, and they need 'a great many studies of therapists' per-
can tell us only (hot it was "psychoanalytic" or sonalities and of therapeutic process before
"Supportive"' or some such theoretically these issues can be resolved.
 Tai

INTRODUCTION 7

some Current writings about "the - therapist

The Good Therapiflt vs. the variable" which seem to assume a unitary trait.
Appropriate Therapist As he sees it, all good therapists have certain
multiple traits in common.
How does the therapeutic style, as indicated In contrast with Prinzhorn's general
by the personality of the therapist, affect the assumptions about the good therapist, Carl
outcome of therapy? Is it that certain kinds of Jung emphasized in 1934 how therapy must be
therapists are "good" in the sense that they differentiated for different kinds of patients:
are effective with most disorders under most
conditions? Or do different disorders require The psychbtherapist should no longer labour
different skills and different kinds of under the delusion that treatment of neurosis
demands nothing more than the knowledge of a
therapists? The first of these views has been technique, he should be absolutely clear in his
called the "generalist" approach; certain ." own mind that psychological . treatment of the.
therapies or therapists are seen as-. effective . / sick is a relationship in which the doctor is in-
across the board. The second .approachi, volved' quite as much a the patient . ... All the
assumes' that different disorders require dill. more significance, th efore, falls to the
general attitude of the doc , who must know,
7.ferent treatments; it will be called the "flit: himself well enough not. to deitroy the peculiar
ferential approach." (See, for examgle; .1:
values of the patient entrusted to his care, what-
Kiesler's grid model, 1971). . ever these may be. If Alfred Adler were to re-
These two views have persisted in then quest analytical treatment of his old' teacher
literature for more than 40 years, each endfirs- Freud, Freud would have to adjust himself to see-.
ing Adler's- peculiar psychology, even . to the,
ed by various clinicians and researchers, who, point of admitting its general right to exist; for -
generally speaking, appear not to know C41the there are innumerable people whose. psychology
views of others. Until recent years;/ the is that of the son in need of prestige. If, on the ,
literature' ha4 been quite discOnnected..Iti ;con- other hand, I were to analyse Freud, I would be,
sidering these two views we are no longOr ask - doing him a great and irreparable .wrong if I
failed to take elaborate account of the very
ing whether personality is impOrtajit; 4
but ,real historical significance of the nursery, the
rather how is it important. importance of the entanglements of the family
.Let us look first at the "generalist" ap-. romance, the bitternesi and gravity of early-
proach. In 1929, Hans Prinzhorn *rote as acquired resentments, .and their compensatory
follows: acconipaniment by wish-fantasies which un-
.happily cannot be fulfilled .. ,. (1964, p. 164,
The principal problem piVc etdpy is the emphasis in the original)
therapist. Anyone who, speaking an expert,
maintains that personal 'gifts and ap itude play Thus Jung agrees with Prinzhorn that the
no part here, speaks against bettyr nowledge, therapist's personality is critical but he em-
darkens counsel for the uninitiaeed, and. en-
courages the irresponsible experimenting phasizes the differential' aspects of the thera-
those who; unsure of themselves, love nothing tic relationship. Note that when Jung
bette#othan, with, their analysing, to play the write bout Freud's personality, he frequent-
vamMte to other minds. On the contrary, there ly means "extrovert" and when he writes
is:in fact a specific psychothOd peutic gift which about Adler, he frequently' means "introvert."
is exceedingly rare. It compiises three things:
1. Wide and sure knowledge;of human beings, no Thus, he is writing about therapists' adapting
matter in what degree of. consciousness 2. to types of patients and not just to individuals.
Edsy self-objectivizingHelimination of the 'The remainder of this section is concerned
pdvate-ego)-- 3. :Innate capacity' fo-r leader- with the generalist frame. Wilhelm Reich (1949)
ship (instinctive vital certainty of aim). Jo addi- recognized that the therapist's individuality
tiOn, there are certain desirable qualities of
character and intelligence, among them this, would affect the choice of patients, but he
rare one: freedom from immature traits, from fioped that the training analySis would
.

what is -neurotic and infantile. (1932, p. 330-331, establish .,"the necessary plasticity, of
emphasis in the original) character," (1949, p. 149). .Hence, the good
For Prinzhorn, there is one psychotherapy, not therapist is flexible and can deal with a yariety
several, nor many. Yet he insists that several of patient types: This flexibility can be' achiev-
traits are invohred. fie would not go along with ed through a technique.
.11

r-,
)
 EXPLORING THE BUCHO-SOCIAL THERAPIES

_ Also in the generalist frame, Karl Menn- 'Sol Garfield, and Allen Bergin studied the ef-
inger had the following to say: fectiveness of therapists-in-training who'were
The psychiatrist as a person is more important
working with a wide variety of patients (no
than the psychiatrist' as a technician or scientist. specific diagnostic group). These studies con-
What he is has more effect upon his patients stitute a kind of test of the generalist
tfian anything he does. Because of the intimate hypothesis. Hardly any of their measures cor-
relationship between patient and psychiatrist, related with outcome (1971b). However, those
the value system, standards, nterest and ideals therapists with elevated scores. on certain
of the doctor' become very i portant ... their
effectiveness as therapists d pends in large part MMPI stales had less success °than the
upon the stature and breadth of their own per- "healthier" ones (1971a).' These findings have
- sonalities. (1952, p. 156. empha is in the original) not been replicated; but they are certainly sug-
Holt and Luborsky (1958), usin supervisors' - gestive of Prinzhorn's concern that the
ratings as a criterion, obtained a de cription of therapist not be neurotic and immature.
:the "good therapist.': A siinilar design was Lambert, -Bergin, and Collins (in press) use
used-6y Knupfer, Jackson, and Krieger (1959). the phrase "psychenoxious therapist" as a.
When' researchers of the -Psycho-social general characteristic _.of therapists. Bergin
,therapies discover a characteristic of and Siinn feel that the "differential effec-
therapists or of therapeutic relationships tiveness of techniques is not well established"
which is related to effectiveness, it is not un- (1975, p. 525). A's we shall soon see, there are
common for them to assume that this is a others who have concentrated their efforts in a
general finding, not limited to the type of pa- differential frame. As usual, the resolution is
tients they studied. -There is a tendency to not all-or-none. After a consideration of thedif-
'view specific findings in.. a generalist ferential hypothesis, an attempt will be made
framework. to synthesize it with the generalist approach.
Carl Rogers and his associates (Rogers et a1.
1967) were among the first to show* systemati-
cally that a particular kind of relationship
might be helpful while its opposite might be -The Differential Hypothesis
damaging ,to patients. They worked with
schizophrenics, but the variables of "wariiith," The earliest systethatic use of the differen-.
"unconditional regard," and "accurate em- tial hypothesis' is in the work of Whitehorn and
pathy" had already been conceived in a larger Betz. The studies to be'repotted proceed from
context, and were viewed as being their work. They began their work in the early
characteristic of a "good" therapeutic relation- -1940's, their publications ranging through the
ship (Rogers, 1954). While the generality of late 1940's, 1950's and early 1960's. They set
these findings continues to be debated (Bergin out to understand what, kind of psychotherapy
and Suinn, 1975; .Lambert. and DeJulio, 1976), .is effective, with schizophrenics. They' early
the powerful constructive influence of this recognized that those therapists who were ef-
research makes it a landmark along with that fective with schizophrenics were not par- .

of Whitehorn and Betz (1954), to be. discussed titularly effective with depressives. And from
. further below. the beginning tbey were concerned with the
Bertram Karon and his associates, working personality of the therapist and his or her rela-
also with, schizophrenics, have developed a tionship withithe patient. The therapist pro-
concept and a TAT measure called "path°- vides "an opportunity for the patient to have
. genesis." They have found that this variable is the experience of being unconditionally under-
negatively associated with therapist effec- stodd by another human being': (Betz, 1946, p.
tiVeness (Vandenbog and Karon, 1971). Karon 252). "... the therapist's effectiveness comes to..
(persona communication)' feels that this 'is a lie in the relationship the' pktient forms with
general characteristic of ineffective therapists him as a person .. .." (Betz, 1947, p. 272, ,em-
although he is quick to add that it has not been phasis in the.original).
tested for therapy with other: than schizo- Frieda Fromm-Reichmann (1950) had much
phrenics. to say about the personality of the therapist,
 INTRODUCTION

and she specifically endorsed the differential If 216 assun4-thalthere are a number of dif-
hypothesis: the "therapist should not. expect ferW dimensions involved in the `therapist
to be capable of treating persons suffering variable," some of which are characteristic of
from any type of personality disorder ... he all therapists who are effective, while others
should learn what type of patients respond are specific to those therapists Who are effec-
best to his personality" (1950. p. 40). tive with specific disorders, the issue of
The Whitehorn and Betz research stimu- general vs. differential becomes a kind of
lated hundreds of empirical studies of the dif-. rough proportion to be determined over a fair-
ferential approach to the personality of he ly large number of empirical studies. To what
therapist; these will be treated further below. extent are there general abilities, traits, or
There' were, however, some. developments Skills which are needed by all therapists deal-
which were quite 4ndependent of this stream. ing with all kinds of human disorders or.cott-
In 1960 Fairweather et al., published their tiersely.to what,extent:are the skills that,are
pioneering, systematic study of four thera- needed specific to-SpO'cific problems?
peutic programs with three, broad patient The Indiana Matehing Project (Berzins, 1974)
groups. This is one of the first 'studies to' 'does not answer this question across the board,
demonstrate how specific are the effects of but X, is an example of an important approach
specific herapies. to the question. Ten therapists in a university
In 19 4 Riemann questioned agaik the tradi- Counseling service treated 751: patients whose
tional sychoanalytic emphasis on technique characteristics were measured on diniensions
(Chen ,1966). Riemann's owns typol- such as "avoidance of others," "turning against'
ogy schizoid, depressive, hysteric,,,and com- the self," and so forth. For the therapists, a
pulsive-- is extended to the analyst. In addi- number of, personality dimensions were
tion to describing each type of analyst's measured. If a particular dimensiokshows as a
behavior, he attempts to match patients and "mairleffect," i.e., significant fOr all patients, it
analysts (Riemanu, 1968). Generally speaking, can be considered general. If, however,, it
he favors like kairs, except for the schizoid shows an interaction with patient characteris-
therapist whom he regards as problematic, tics, i.e., effective with certain patients but not
possibly best with hysterics (which leaves the others, it is differential. In this particular.
schizoid patient With no analyst at all). study there are,. roughly twice as many dif-
Beutler (1976) has reviewed a number of ferential factors as general ones, thus pro-
evalnatiOn;stuies, classifying them as to tyke viding more .aupkort for the differential hy-
of therapy and type of patients. For example, pothesis than for the generalist.
hesUggests that behavioral treatmentamay be Fairweather et al. (1960) found many, many
more effective in dealing with habit patterns treatment-by-diagnosis interactions and only a
t, while psychotherapy proceduree may be more .few simple treatment effects. Their study blf in-
'effective with adjustment problems. Differen- patients in three broad diagnostic groups
tial processes are considered alpo by Goldstein strongly supports the differential approach.
and Stein (1976). .
Whether or not general' factors are found
In recent yearithere has been a dramatic in- will probably depend on a variety of considera-
crease in the number of systematic empirical tions. It is possible that inpatients are more dif-
studies which assume the differential approach ferentiated than outpatients in' their pathol-
and the consequent desirability of matching ogy, thus perhaps enhancing the importanceof
clients and therapists. A variety of theoretical differential factors.
frameWorks have been used; These studies are How experienced are the therapists? A
in two. distinct areas: juvenile delinquency and group of therapists-in-training undoubtedly in-
mental health/eounseling. -The former will be cludes some who will not remain in clinical
discussed in chapter V. The 1 r have been work. 'In such a group, among the general
carefully reviewed by Berzin press). For dimensions may be those that in effect are dif-
our purposes it will be useful lo at one of ferentiating therapists froin -nontherapists,
these studies for the light it heds on the dimensions that would not shoW in a study 'of
generalist vs. the differential approaches. experienced therapists (see page62).
 EXPLORING THE PSYCHO-SOCIAL THERAPIES
/.
In the Indiana study, the pitients were out- must speculate about why certain researches
patients, and the, therapists, were quite ex- are' unacceptable. Such speculation may
ciperienced: It might well be a typical middle-
stimulate studies and - may ,increase our
range study. If so, we might anticipate that dif- understanding, of our blind spots.
ferential factors are going to be far, more fre- Freud felt that these matters were so cora-
quent than general ones. plex they defied codification (1958, p. 123).
There is another approach which is much Many clinician's feel that each therapeutic en-
eglimpler (though not so informative). This is counter is distinctive; there can be no general-
simply to ask whether therapists who are ef- izations: Justhy such a' negative view should
fective with one type of patient are likewise ef- hold for personVity, but hot for technique,, is
fective with other types. In the studies to be not made 'deaf!'
reported we will ask both kinds of questions. Sometimes,2When these issues are con-
Again,We shall find that the weight of evidence, sidered, there is an underlying pseudoscien-
is for differential factors, not general' ones. tisna: studies of behavior are considered to be
The clinical literature is otherwise, that is, more scientific than studi personality..Em-
the weight of printers' ink is in general ap- phasis on technique is les shamanistic than is
proaches. Most of the literature presents emphasis on the thera Off
prescriptions for therapeutic behavior without Still another view t emphasis ,on tech.-
specification of goals or situations. It either nique absolves:-the clinician of responsibility.
assumes or extols the virtues of a particular So long as we are doing what "should" be done,
therapy; contraindications are presented more we Cannot be blamed for a bad outcome. It is
as exceptions than astthe rule. Till's is not just.),difficult for persons not engaged in this enter-.
the case,literatlAte.Mizenithe research report;; prise . to understand the feelings of responr.
contain much discusSion, bf the good therapisa;,, sibility immense, lonely, responsibility that
in the abstract.' . clinicians feel. " -

'We accumulate a mass.of informationper-

'. :Sonal, confidential, private, much of it .un
Why So Few Studits? verifiable whichi, even under the best of cir-
Cumstances is difficult to share very Widely.
Actually, there have been hundreds of And wp must make ,judgements that we, know
studies of the personality of the therapist. But will be received very carefully,- sometimes
these studies have usually been.: small ones, even with homage. ,In this lonely responsibili-
more often than not "labors of lOve" of a single ty, therapists have one support : tpchnique.
investigator or small group of investigators This lonely responsibility is most awesome
with paltry financial resources. The big studies when we consider the possibility that we might-,
have been studies of "process," of technique. In :damage our patients, a possibility so threaten-
view of the opinions quoted above of such com ing it is usually denied. The most carefully
netent clinicians as Jungi.Menninger, Fromm- documented instance of physiCian's denial that
Reichmann, and Sullivarrckne cannot help but they themselves were at fault is in Colby's
wonder: Why so little attention to the per- (1960) account of S_ emmelweiss. He writes it so
sonality of the therapist? beautifully, one feels that one cannot sum-
Not long before he died, John C. Whitehorn, marize, but should reproduce the entire seven
himself a former president of the American pages (p. '44-50). But summarize we -must:
Psychiatric Association and a pioneer of Semmelweiss had great difficulty convincing
studies of therapist personality, told me that his associates that they were responsible for
one of the leaders of American psychiatry had the deaths of the new mothers; that it was they
complained to him about the Whitehorn-Betz who were carrying the germs from the cadaver
research. This was Whitehorn's way of telling room to the delivery room. If such an outcome
me that the area is controversial. Unfortunate- rests on each of us as individuals, it is very
ly, we have no systematic studies of research tening indeed:-
strategies. We seem to prefer expert opinion, . But if the outcome 4oes rest on each of us as
In the absence of research on research, we individuals, we can 'never be rid of the threat

, 4 r
 INTRODUCTION
' If '.-
Similarly it is asserted here tha human
..;

until we understand h0* we affect the

7.
out-
come. )
behavior is ":'usually phenotyPieTa the follow-
Does this mean that .we are responsible for ing senses: : , q . .,

all _of our failures? NO.' Most patients are Ira-- 1. Any single behavior is a resiiltant of more
bedded in a web of,interpersonal relationships than one genotypic force.
Of which the therapistis only a part, and usual - g. A particular kind of behavior. In 'different
,

ly not the'most imiportiiiit part. What happens individuals may result .-frolii different..
to the patient is a function of all the forces .genotypic forces. .
. Q .,

within and around him. 3. The same genotypic:, force- may. result. ,in
'By the same token we cannot claim credit for different behaviors 'in different indiacl-
all our successes. What we can claim credit fox: uals.' . .

is a high batting average. There is little doubt ' .

that we can improve our batting average by While we are concernedjwith explaining
.

choosing patients who fit ua:Personally. Just b avior; we can do-so only if We understand,
how restrictive we sliouldi.be is a question th genotypic forces which shape, it. So loniva:a*
which is quite a few ; years: ,beyond present we look only at behavior we will never iinde*"
knowledge. For relearehers'have not yet iden- stand it any more than Aristotle could. tinder-'
tified the relevant variables. When that' has stand a waterfall or a fire. ' ., ' .

been done we : must discover the extent to .

At present, there is little understanding of
which therapists can adapt to, the, varying pa- the psycho-social Therapies. What therapists 0,,,°
tient requirements. There is some eviden6e learn in training is primarily...a function of .

that some. therapist variables are quite stable where they. got their . training, different
over long periods of time (see appendix 4). On centers being conmitted to one.,Or another ap-
the other hand, some observers are confident proach, there being fesiv'law's that ielate typeS "

that therapists can adapt if they know what to* .

Of therapy to, goals, problems';: situations, 'etc --4-,-s_.
do (e.g., Ricks, 1974, p. 294). What therapists tend to do in this "social" "
These personality issues will be treated Itherapeutic) situation is influenced somewhat
-more _,systematically in the 'next section, but by their, training but is primarily an exp ession
readers' not interested in psychological theory of their personalities.
may skip iti if tney:Wish. Personality is defined as an, organize set of
predispositions to behave in certain waSrs., It is
an energized structure that (derives from
Theoretical Summary ...various origins.' "Teniperament" frequently
refers to physiological predispositions: that
The following is a 'statement of the derive from the genes and_from environmental
theoretical position on which this research-. insultsdiseases, injuries, and so forth. But
rests. While I believe the postion is consistent m h of personality consists of memory traces
with available empirical findings, the available from st learning which 'result, in certain
findings are clearly not adequate to affirm the characte istic reactions in certain situations.
theory. (The var origins of personality are not in-
Human behavior is viewed here as being dependent; for example, past learning can pro-
"usually phenotypic," not usually genotypic. duce a psychosomatic reaction which changes
The distinction between phenotype and , temperament.) Behavior is viewed as a product
genotype, as used here, is not limited to of the personality and the-situation. Newcomb
genetics, but is a broad scientific distinction (1950) emphasizeThat personality is oriented
which can be illustrated by Aristotle's both inward and outward.
classification of "earthly" and "heavenly,"
bodies (Lewin-, 1931). Rain is earthly while the 'These assertions are "statistical.7 There Amy well be ex-
rising fire is heavenly. This is a phenotypii ceptions. For practical purposes, the behavior of a rat in a
description, it being quite beyond Aristotle to maze may be assumed to represenix genotypic force, learn-
understand the genotype, ' avity," as the ing, just as a steel ball on an inclined plane can represent
primary force explaining b phenomena. gravity.
 12 -EXPLORING THE PSYCHO. - SOCIAL. THERAPIES

Can individuals adapt to variouthittiations? The position here is that in a "bootstraps"

If we knew what to tell therapists to dO,'ciOUld science we make use of what we have, hoping
they follow our instructions or are they 146,Und always to learn and to improve. Certainly,
by their own predispositions? Probablyi, the there have been, imajor improVements in
therapist can follow instructions (technique) to kagnosis in recent years. For example, the
a degree., but,within lin-ririnarobably, Certain unipolar-bipolar classification'of depressions',
kinds of reacTions are very deep-and automatic appears to be more precise, reliable, and pro- '
b
as, for example,- when 11) therapist of a. par ductive than were some earlier classifications.
ticular -origin reacts angrily to a slur ,:on his or In clinical studies We frequently have dif-
her origins.. Other behaviors are, probably
. ficulties achieving -adequate samples. Thereis
Under the'. -controi of 'the iheraPiSt- if only it a tendency to include patients that might bet-
were known what behaviorware*Appropriate. ter have been eiclUded. Failure of diagnosti-
This issue is critical to 'the "generalist" vs. cians to agree may arise if the patient is suffer--
the `differential hypothesis.", If -Reich' is 'cor- rtg from more than one, disorder, or from a
-. rect in: his hope that training anaiyiis,v
establishes the necessary: ``plasticity of
character," then thelfne generalist is differen.-
order WichI
t
we have not yet adequately
ined. d' described; in neither case does the
patient long in a study.
tial andjhe issue disappers. One "taped ofidiagnbsis° is receiving inCreas-
There little doubt . that therapeutic train- irig., attention:, how, fine or Coarse. are the
drig, -` and particulaily analysis, requires a criteria? In , discussing ,1**40ime dis-
r
therapist to examine his -own reactions and to orders, glerraiin-41973YarirdeS-Itir' efiltraliatie
atteinpt to control them. On the other hand; rather than a unitary approach, i.e., that there
. .
theiapptie:training 'ttitztalso tend to 0110 cer- are a, variety of affective disorders with dif-
tain restrictions on plasticity of character. ferential responses to various treatmenti.
'What little evidence we have about therapists=" But this issue can bCvie'wed quite different-
suggests that their, behavior is still largely ly: For; soiree purposes, it may be useful to, treat
determined 'by theirpersonalities: the depressive disorders as unitary, for corn-
Understanding personalities 'facilitates the Parison with the schizophrenias, for example.
, interpretation of behaviors. Personality vari- :Indeed, we will consider the possibility that
ables can shed light on misunderstandings that certain therapist behavior is required to deal
can occur between fherapists and clients. This with an even broader class: the functional
is not to say that personality studies should psychoses. Diagnosis is here viewed as hierar-
take precedence over studies -of therapeutic chical: certain broad classes divided into
process. Rather, thetwo should go hand in hand. subclasses, analogous to the families-genus-
Since therapy_ is a social situation, it is ex- sPecies classification in botany. Some
pected that, 1.the social dimensions of per- therapeutic requirements may distinguish the
sonality authoritarianism, nurturance, defer- broad 'asses, while other requirements, the
ence, etc.are particularly fruitful areas for subclaises. The issue then becomes the levtl of
understanding the effects of therapy. inquiry, and, this may be determined by the
availability of patients, therapists, and so
forth.
Diagnosis - At one extreme, the broadest of all classes
all mental and emotional disturbancesis the"
In contrast to the generalist approach, which generalist apprdach described above. At the
assumes that therapies are good or bad, the dif- other extreme are those who assert that there
ferential approach assumes that an interven- are no classes;Deach, patient is unique; eaat
tion is good or bad depending upon whether it therapeutic, relationship is unique; indeed, each
is appropriate to the problem. It therefore encounter is unique. Between these two ex-
places considerakle weight .on being able to tremes are various levels of generality; at each
define problems.) Diagnosis of the mental level there are undoubtedly regularities across
.disorders is so difficult that some have types 'of patients and therapists. A knowledge
despaired and thrown out the idea completely. of these regularities should help us to improve

9 .

t.J
 .,., .

,....-,INTRODUCTION

our averages, and avoid, in Prinzhorn's terms, r'llthaviora.1," "Itfigerian;" etc.2 We have seen
"playing the vampire to other minds:" from Lionells' work that these descriptions,
really tell tri vdy little about what` went on in
therapy, Por whatwent on was more a fwletion
Of the personality of the therapist t65-n i of his
training or his 'school identificatiom,What was
Is Psychotherapy.Effecilve?, fnztnipulated was v,ifrimprecise;
_
and quite I

variable (Wexler, 1975). Under?citcumstances

In a differential frame, this qu tion makes where _the, independent variable contains so
no sense, as. Sundberg and Tyler e aginized iri m "error," we are .unlikely, to find any
1962: "... probably unanswerable: But Many tisticialy's,ignifieant_qucome: $oreover, we
inovP1-:"4 be:-..Tacke,tOwa4,d_butaainillg:!saufikiii- are"unlikely to find Significant- differences be-
dc;iiendabie. information ihout wlearlitind tween, types of therapy fiat are so, peorly, con-
therapeutic techniques,, What kind ra- trolled: Lubors15y, Singer, and Luborsky 975)
pists, and what kinds of treatment conditions in their detailed evaluation and summary o
hive the most beneficial effects on specific fisting studies come to the toncluaion that there-
kinds of `patients" (1962, P. 435). are few significant differences. Of course, null
. "If we were to ask someone, even the man in findings, can result from, any kind of error, in
the street, whether drugs were effective, he selection, Meashrement, and hztndling of data.
would probably respond with a question: But slice we genexally know a great deal more
"What's' your problem; buddy? Headache, about the measurement of outcome variablti
cancer, or just bored?" Long ago, medicine and control variables than we know about th
shed the unrealistic assumption: the good drug. experimental variables the psycho-socia
Today,, many social-psychological alchemists therapies-- it is here that we must concentrate
,continue to present their summa bona, each of future efforts. We must discover, 4define, and
troubles un-
which will; cure our troubles - troubles measure the salient dimensions of the psycho-
defined. social therapies.
In fact, so much, of the literature of the Let us turn now to the :second and most
psycho-social therapies is written in the troublesome problem with the question: Is
generalist, frame, it is. hardly surprising that psychotherapy effective? Inherent in this ques-
some, should attempt to answer the' geheral . Lion is an absolute judgment; it implies a com-
question: Is psychotherapy effective? One of parison between psychotherapy and no
the first was Eysenck (1952). Very recently,
Donald T. Campbell (1976) proposed that we 'The exceptions 'are of interest and two will be considered
"test the effectiveness of our therapeutic alter-, here. In 1952 (more than two decades ago) DaVid Grossman,
natives, finding out which ones work better." acting as a lone investigator-clinician, attempted to vary his
Nowhere <does he suggest the possibility of dif- dwn style (interpretation vs. reflection) across experiinen-
ferential effects. How ubiquitous is the tal groups. Moreover, he proyed that he had done it. It is un-
fortunate that his findings were confounded with otherdif-
generalist type of thinking! ferences in his experimental groups, but it Is' to his credit .

Note that _Campbell asks "which ones work that he had included measures that 'permitted the discovery
betterr' Even in the generalist frame it is dif- of this confoUnding. His design is certainly replicable. (It is
;ow-. ficult to answer the absolute question: Is of interest that :his 'article appears In the Journal Of Con-
psychotherapy effective? There are two suiting Psychology immediately following Eysenck's null
findings. Juit why Eysenck's findings should be sa well,
reasons. First, in most of the presently known while Grossman's are .not is an interesting question'!
available studies, psychotherapy is .ill defined. for the sociology.of knowledge.) The important thing about
Second, being an absolute qatstion, it implies Grossman's study is that when a therapist attempts to vary'
what is probably impossible: a comparison wit his style, although the personality of the therapist is not
a sample of patients who are receiving no eliminated, it is at least controlled. Similarly, in the
tliVrapy at all (the so-called "control group")., Fairweather et al. (1960) study mentioned above, two
investigator-therapists served for all therapy groups, thus
With respect to the first problem, in most ex- controlling for the personality of the therapist.' Would that,
isting studies, the therapy (the independent we also had measures of .the personalities of Grossman, and
variable) is meagerly described as 'analytic," of Fairweather et al.

2G.
14 EXPLORING THE PSYCHO-SOCIAL THERAPIES

therapy pt all. It is possible to conceive of a While we cannot be .certain of the=meaning,

sample of patients who have received, say, no it is useful to speculate. Bergin and his
drugs; but one cannot Conceive of a sample of associates feel that the increase .in variance
patients to whOm nothing has been done results from the deterioration induced by the
psycho-socially. As soon as a problem or psychonoxious therapist. This is in the.,

disorder is suggested, something is done, par' generalist frame. It is equally plausible to

ticularly in these days of "cassettes," "rap ses- assume that the increase in variance is due to
sions," "bibliotherapy," "writing therapy," and the mismatching of therapies and patients. In
so forth. if a priest visits a person regularly, is this speculation we simply assume (ivith
that _pers an adequate "control"? We will Lionelli) that whatever went on in the therapy,
.never be able o give absolute answers for the it was a lot of different things, some of which
psycho-social therapies, nor should we. We can were appropriate and some were not so ap-
only comparuyarious kinds of interventions. propriate to the paiticular patients involved.
(see pages 8748). These speculations lay the groundwork for
Is there nothing then that can(be said? Many the distinction between an efficacious therapy
of the studies of the psycho-social therapies and an effective one. Although the words "ef-
have shown a large increase in variance, i.e., ficaciousr and "effective" are frequently taken
some of the patients show dramatic improve- as synonyms, it is possible to distiquish be-
ment while others .show no improvement or tween them. The former is a powerful therapy
even a worsening. This finding is not so fre- that may or may not be correctly applied, e.g.,
.." "good" therapy but pot for some patients. Ah'
quent .for studies of drug therapies.'
effective therapy is a powerful therapy cor-
rectly applied. Probably the best interpreta-
'The increase in variance in the psycho-social therapies is tion of - currently available evidence is that,
disputed. Lambert, Bergin, and Collins (in press), review the tpsychotherapy and other psycho-social thera-
Issue with considerable rigor. They correctly conclude that
the increase in variance does not constitute proof of pies are indeed very efficacious but they are
anything but that it is a phenomenon that must be dealt currently only modestly effective. We should
with. May (1971) doubts the evidence. He is'greatly corkcern- hope for more, if only we' knew .when to, use
ed with- an increase in the variance relative to.a control which therapies.
group. If he cannot find a control group, or if the control We turn now to the pioneering/ efforts of
group fails to meet his criteria, he throws the study out.
Looking at May's own, study, we find a large increase in Whitehorn and. Betz to discover what kind of
variance for the milieu-therapy group (control group?), and therapy is effective for schizoPhrenics.
for the psychotherapy-alohe group (which appears really to
beg psychotherapy-plus-milieu-therapy group). No such
lagge increases are found for the drug therapy' groups. Whitehorn and Betz
(Conipare May and Tuma. 1964, p. 364, with May, 1968, p.
182). Some explanation is needed for the large increases in
variance for the psycho-social therapies and the large dif- In 1941, . John C. Whitehorn b ame
ferences in variance between the psycho-social therapies Psychiatrist in Chief and Director of the H ry
and the drug therapies. Phipps Clinic of The Johns Hopkins Univers'
An increase in variance can be found if, and only if: (1) the Medical School. This was the 'chair of Adolf
outcome measure in an interval or ratio scale, (2) applied
"before" and "after," and (3) the averages do not crowd the Meyer, the most 'prominent, psychiatrist in
extremes of the scale. Unfortunately, the studies to be America. Meyer's -)psychobiology" explained
reported below do not meet Lhese conditions and therefore the mental disordeFs (including schizophrenia)
do not help us with this question of increase in variance. as "bad habits." He thus provided what ts
In addition to the disagreement about the increase in essentially a psyChogenic explanation, and ac
variance, there is also a lack of consensus about the extent
of deterioration among patients in psychotherapy. Compare cordingly he emphasized psycho-social in-
Lambert, Bergin, and Collins (in press) with Gomes and terventions.. It was not that Meyer did not
ArmstrOng (1976). But the question of actual deterioration know his biology, or his residents, for-that mat-
is lid really at stake here. If the increase in variance is due ter. It was simply, as one of the residents put
simply' to the fact that some patients failed to improve, the it, "difficult to see where the biology came hi."
issue remains: Could it be that some patients are gqtting
constflerably less effective therapy than they should be Personally responsible for many of Meyer's pa-
getting? tients, was Barbara J. :Betz, a research
 INTRODliCTION 15

oriented psychiatric resident who had just training of young scientists today (Stansfield,
complbted a study of the somatology- of 1975). He noted that when certain doctors went
schizophrenia (Betz, 1942). Thus,- although on vacation there were dramatic changes in
there was interest In somatology, therapy was their patients. There was' one doctor in par-
-essentially long term and psychasocial at the ticular; when he entered the room, schizo-
time Whitehorn became ditector. phrenic patients "froze." In fact, Whitehorn
Whitehorn was known then as a biochemist. came to use the presence, of this doctor to
He had been brought up in a sod home in diagnose schizophrenia.
Nebraska, worked his way through school, and, Not only did he observe, he counted. He
for 17 years at 'McLean and Massachusetts -counted even his own interactions with others.
General Hospitals dkveloped methods of At the age 'of 12' he had been shot in the eye
chemical assay which were the predecesiors of with an air rifle, and one of his eyes changed
those in -use today. , color while the other did not. Sometimes pa-
In those days, when a bioch,emist needed tients would remark on this. He soon
blood, he drew it himself. Repeatedly, when he -discovered that it was schizophrenic patients
sat down to morning coffee, the psychiatrists who would so remark. In fact, the count got to
would grumble about particular patients.-From 25 schizophrenics in a row, unbroken. (The next
some of these patients he-had 'drawn blood hat son to comment was'the director of a great
Very morning, and they had been completely ,medical center. So if yoic have eyes of different
cooperative with him. It 'puzzled him that he colors, you will still have to be careful in using
didn't . have trouble with these difficult pa- this, diagnostic testi)
, tients. It was when some of his "controls" "im-
He searciliti for 'hyperglycemia in excited proved", that he began to appreciate the power
patients but was unable to find it even when of the interview: He devoted a vats. deal of at-
the patient was chasing, him with a chair: His tention tO interviewing skills (WgEehorn, 1944,
kiend, Walter B. Cannon, found this hard to 1947). He was concerned with the role of Com-
believe and came to see for himself. Together, munality in mental health "(Whitehorn, 1954).
they watched a woman dancing on 'her bed, He sought, to establish a healthy social corn-.
muttering about the "wisdom of the body," in- munity on the wards.
dicating she knew precisely who her important There came a patient who was mute. When
visitor was (though she was not a biologist). he was given sodium amytal "to - get some
'But she refused to acknowledge his presence. material" the whole, ward went silent.
When Whitehorn inquired why this patient WhitOhorn recognized thi4 sign of negative'
had improyed ,so little after 8 months in the group feelings. All of these things indicate his.
hospital, her psychiatrist responded, "Why continuing concern for accurate diagnosis' and
don't-you take the case?" He did; she was well in effective ,treatment, and for the relation be-
a week, and she stayed that way. "I had simply tween diagnosis and treatment.
offered the patient the opportunity to establish In 1938, Whitehorn becaine professor Of
contact?! Thus was born his interest in the per- psychiatry at Washington University in St.
sonality of the therapist. Louis.-:Tkere he got acquainted with Carlyle
Soon he found the ,younger staff members Jacobsen who was doing studies with the
were sneaking out to the chemistry lab to talk Strong Vocational Interest Blank ASVIB).
over their patients witirhim. Patients, too, in- Whitehorn asked that it be mailed to a number
dicated their appreciation. (Until then, he had of his friends.
resisted the title "psychotherapist. ". Now he Not long aftee.he took ov-etthe Phipps Clinic
found psychiatrists waiting on him to find out at Hopkins, he asked Barbara Betz to study in-
his secret. And so he- left the laboratory to tensively ~a subgroup of the schizo-
work in the clinic, where he was soon much in phrenias the obsessive-compulsive ones
demand as a doctor's doctor. and their reactions to treatment. This clinical
Whitehorn was always observing, metic- pilot study resulted in her 1946 and 1947
ulously. This kind of exploratory observation Rapers' cited above, which contain many of the
is illustrated here because it is so lacking in the hypotheses to be tested in their later work.
16 EXPLORING THE PSYCHO-SOCIAL THERAPIES

They ha4,already begun to administer, routine- doctors who had a high rate of improvement
15', the Strong Vocational Interest Blank iSVIB) among their schizophrenic patients and "B"
to all incoming residents. octors with a. low rate. Using this criterion, A.
In many'respects the Phipps Clinic was an , vs. B, the Whitehorn-Betz studies proceeded in
ideal place to conduct research into the effec- two parallel but essentially, unrelated
tiveness of the psycho-social therapies: -pathsone involving the reconds the therapist
a. A small clinic, its social climate was rela- produced, and the other, responses to the
tively homogeneous. items in the Strong Vocational Interest Blank
b. Clinics at Hopkins tend to be relatively (SVIB).
autonomous. Whitehorn could and did protect Considering first the analysis of the doctor's
both researchers and clinicians from intrusions records, there were two of particular impor-
from outside the Clinic. tance; the Personal Diagnostic Formulation
p. As an inpatient setting,. it provides a and the /Therapy and Progress in Personal Ad-
greater' opportunity to Control the forces im- justment. On the basis of these recordi
pinging upon patients. AnOther way of saying Whitehorn slut A113et.. found ' that the A
this is that therapy is the major source of in- -therapist: ,
fluence on each patient, not the forces in the (1) ... indicates in his personal diagnostic
larger social environment. Forces which are formulation same grasp of the personal meaning
not part of the therapeutic system are "noise." and motivation of the patient's behavior, going
They make it more diffiCult to ascertain the ef- . beyond mere clinical .description and narrative
fects of therapy., biography; ,'
d. The high staff-to-patient ratio means that (2) ... in his formulation of strategic goals in
the treatment of a particular patient, selects
the assigned doctor is not a person remote personality oriented mill, i.e., aims at assisting
from the patient jiut rather one who is actively the patient in definite modifications of personal
working with the patient. adjustment patterns rather than the mere
e. .A high staff-to-patient ratio is essential to decrease of symptoms ...;
the maintenance of good records. (3) . :. in 'his darto-day tactics makes use of ,
"
.

"active perional participation," rather than the

The' records of the Phipps Clinic included patterns "passive permaissive," "interpretation
several systematic mesures (Whitehorn and and instruction," or `practical care." ,

Betz, 1954). (1) Since 1914, each daythe nursing There is a similarly high association between
staff has marked a behavior chart indicating improved condition at the time of .a patient's dis-
each patient's behaviors that day. Also record- charge and the development by the patient,
while in treatment, of a trusting, confidential
id each day are the pa t's (2) social behaviok relationship to the therapist. (Whitehorn and
with other patients, an 3) participation in Betz, 1954, p. 331).
clinic activities: (4) .At discharge, the disposi-
tion Qf the patient is noted whether to the This analysis is really more..informative and
commimity or to another hospital. These useful than the A-B Scale derived from SNUB
records served as -background material for a analysis to be reported below. It is notable that
judgment of "improved" or "not improved" no one has attempted ,to replicate this analysis
made by the clinic director, the chief resident, of records. Iit-contrast within a year after the
and the therapist. 'All these activities were A-B Scale was published in 1960 an attempt
routine, not part of a 'research protocol. No ex- was underway to replicate it and there have
perimental manipulations were involved. No been hundreds of funther efforts. .
judgments of improvement were made while Whitehorn (1972) felt that the A-B Scale had.
'therapy was in progress. Only after discharge, "distracted attention from the primary issue"
was this issue. considered. of the origil A.B. research: "the difference in
Whitehorn and Betz computed for each modes of dealing with schizophrenic patients."
therapist the percent of schizophrenic patients Betz (1972) sees clearly why the scale has
improved. (They also computed the percent of eclipped the clinical da'ta; it is "the easiest data
depresSive patients and of neurotic., patients to. use." The records data are not easily
who improved. This will be explored in chapter replicable. Indeed the necessary documenta-
IV.) The therapists were then divided into "A"'I' tion is 'not _ayailable inmost -clinical settings.'
 INTRODUCTION 17

Perhaps if we can understand the scale, it Roughly half of them seem to have a common
will .help us with the priniary issue: the differ- meaning: the "A" therapists tend to reject
ing Moderof treating patients. manual and mechanical occupations and ac-
F pf: o*C: tivities. This half of the items .dominates the
scale. If there ,is a common. metalling to the
The."4-B Seale" scale, it concerns the meaning of these items:
Actually, there area number of "A-B Scales"
Whitehorn and Betz conducted a number of (Kemp and Stephens, 1971). Some investigators
analyses of the strong Vocational Interest have sought to lengthen the 23-item scale,
Blank. Of these, the one which has stiniulated others to shorten it, others to make it more
the most research is their item analysis which homdgeneo-us, and still others to revise it, op-
yielded 23 items that differentiate the A and. B timize it, and adapt it to revisions of the SVIB.
therapists (Whitehorn and Betz, 1960; Betz, Most frequently used, though, is the 23-item
1967). These items are presented in table 1. scale or scales derived from,it:

Table 1.-- Strong Vocational l st Blankt itemi Which Differentiate

A' and B Therapists
Item No. Item Response
*1
. 17 Building Contractor *D
19 Carpenter L *1 *D
59 Marine engineer
60 Mechanical engineer L *D
68 Photoengraver **D.
87 Ship officer
90 Specialty salesman L *D
. 94 Toolmaker L I
*1
121 Manual training *D
122 Mechanical drawing L *1

Drilling in a [military] comp,any *L De

151
185 Making a radio-set L .1 'D*D .

Adjusting a carburetor L *I
187
189 Cabinet making L *1 'D
216 Entertaining others L *I *D
218 Looking at shop windows *I
290 Interest public in a new machine through public
addresses (rather than develop, design, etc. the
new machine) *L *

311 President of a society or club (rather than

secretary, member,, or committee chairman) *L D
356 Many women friends (rather than few women friends) *L D
367r Accept just Criticism without. getting sore *Yes 9 No
368 Have mechanical ingenuity Yes 9 *No' -

375 Can correct others without giving offense *Yes No

381 Follow up subordinates effectively *Yes No

1Indifferent,
*Signifies the characteristic response of therapists of whose patients a high proportion improved. The response
withotit an asterisk is characteristic of therapists who had .a low proportion improved.
tForrn M (400 items).
 L.!

18 EXPLORING THE PSYCHO-SOCIAL THERAPIES

I am not greatly concerned here with these to be predictive for blacks but not-for whites
various psychometric efforts. My concern is to even though they were originally validated on
throw light on the meaning.of "A-B" by study- both groups. The authors offer a number of
ing its correlates. However, one question is too possible explanations but are not themselves
important to be left to the psychometricians: very impressed with any of them. Here again
should there be internal consistency, homo- the test literature illustrates, the difficulty 'of
geneity, or common meaning among these generalization and the need for repeated
items (Razin, 1971)? A small digression into the validation on various populations.
area 'of selection tests should shed some light Because the determinants of therapeutic
on this qUestion. success are probably multidimensional, "A-B",
In developing their scale, Whitehorn and should never have been called a "scale." The
Betz were not interested in selecting word conjures up psychometric qualities which
therapists, but rather in characterizing them. these items do not have, and should not have. '
However, the "A-B Scale" could be used to By the nature of the task they perform they z

select therapists. Moreover, its construction is should not be homogeneous or internally con-
similar to that which is used in developing a sistent. Rather various items should represent
test battery for selecting new employees! For various traits significant in the treatment of.
these reasons, it is useful to consider selection sChizophrenia.
tests and what the outcome of such test con- Hereafter,, the phrase "A-B predictor" will
structions can be. stand for any collection of items correlated
Almost any illustration will do, but a recent with therapist success but which collectively
one is particularly interesting. Fox and do not have scalar properties. Most of the col-
Lefkowitz (1974) developed a test battery for lections of items presented are A-B predictors,
entry-level employees in an electronics not A-B scales.
manufacturing firm. The details need not be If, within a given A-B 'predictor, certain
reported here, but after correlating many tests items can be shown to be appropriately cor-
with performance measures, they came up related with each 'other, this subset of items
with three that were. predictive: small-parts, will be' called an A-B chister. As compared with
dexterity, tracing, and plotting. Note that even A-B predictors, A-B clusters are useful becauSe
those tasks that are relatively simple are likely there is greater likelihood of discovering the
to require several abilities. Why then should deeper personality significance of la cluster
we assume that the treatment of a schizo- than a multidimensional predictor. On the
phrenic, probably a very complex phenomenon, other hand, predictors are useful because they
should be represented by-a single homogene- indicate combinations of traits that may be re-
ous trait? Many researchers have been com- - quired for a certain type of patient. An A-B.,
mitted to making the "A-B Scale" unidimen- cluster will be called a scale if we can asiigS a
sional (homogeneous items). It is not uncom- meaning ("construct") to it.
mon .to read about the "therapist variable" as There will be one exception to this ter-
though it were a singloebnitary phenomenon. minology. "A -B. Scale" (in quotation marks) will
However, Seidman et al. (1974) and .Dublin et 'refer to the existing literature of the 23-item
al. (1969) have recently used multidimensional predictor or variants of this predictor which
-.analyses with some success.. have been used in other studies (table 1).
In addition'to multidimensionality, there is a Accordingly, the original 23-item "A-B
second' set of findings for the electronics Scale" is really a 23-item A-B predictor. Within
employees. The tests so developed turned out this predictor there is one cluster consisting of
items relating to manual/mechanical occupa-
tions or activities. The A's reject these items.
`There is an important difference in the use of personali- To know that a therapist .who is effective with
ty tests in research and in selection. When used in selection, schizophrenics is one who rejects manual ac-
and the subject knows this, there 'A a premium on knowing
the "right.answers" and in distorting one's answer toward tivities does not give us very much understand-
the right answer. When used in research, this problem is , ing of psychotherapy. When one thinks about
not so serious. See pages 108-109 and 151. being a carpenter, one might, think about: the
 INTRODUCTION 19'

pay, or the supervision, or the intrinsic perionality disorders and no relationship at all
satisfactions, or the weather, or any number of fot the neurotics.
Otherihings. In fact, different substrates of'a. '.A\ If the A's are more effective with schizo-_
population- will think of different things: Thus Orenics while the B's are more effective with
this central cluster of the original A-B predictor'`` outpatients, then -the "A-B Scale" is the
probably has a variety of meanings for the ultimate in differential treatment. It goes
deeper personality structure of different beyond the notion that the treatment must fit
groups of therapists.. I will refer to a cluster the disorder. It says that -certain treatment
consisting of such items as multisemantic. which is beneficial in the case of one disorder is
In contrast, a manifest scale is one where decidedly not beneficial for another disorder.
most of the items seem to have a common Small wonder that Carson, (1967) called it a
meaning for personality. Such scales are prob- "critical variable."
ably more constant in their meaning In any event, this intriguing touchstone of
groups of therapists,' and they are mope ikter- diffeiential treatment soon attracted a number
pretable in a therapeutic setting. For example, of younger: workers who, quite suprisingly as
we will find a second cluster of items that seem Betz (1972) notes, made it work in pseudo-
to have in common certain aspects of social ex- therapy -- laboratory analogues =--: quite dif-
pression., ferent from the Phipps lOng-term psycho-
In summary, then, it was assumed in the pre- therapy. with real inpatient schizophrenicsg.kt
sent studies that the "A-B Scale" is not a scale.. the' same time, several investigators (May,
It is a Multidimensional, multisemantic predic -' 1968; Bowden et al.:1972) tried the "scale" with
tor.- Uniaveling its multiple meanings is an in- athizophrenic inpatients only to come up with
teresting scientific detective story. null findings.
Continuing theanalOgy, "the plot thickens." And.so the plot thickens still 'further:
Almost immediately after the publication of 1. Being empirically based, the ."scale" has
the "A-B Scale," McNair, Callahan, and Lorr no apparent meaning for psychotherapy.
(1962) attempted to replicate the Whitehorn-. The meanings attached to it are almost as
Betz ffindings. Tbey found indeed that the- numerous as the investigators who have
"scale" worked, but in the reverse direction. used it.
The 40 therapists and 40 patients were in out- 2. The "sCale" fails to "work" in some clini-
patient clinics of .the Veterans Administra- cal settings where it would be expected
tion.t to work.
Since many V.A.. outpatients are neurotic, it 3. The "scale" "Works" in settings where it
was irnmediately. inferred that the, A's were might well be presumed inappropriate,
more. effective with -schizophrenics, while the e.g., therapy analogues using college stu-
B's were more effective with neurotics. In fact, dents as pseudotherapists :reacting to
as is .the case in too many clinics.; studies, the taped material.
diagnosis' of the patients is unknown. McNair
reports that 82 percent of the patients in V.A. It would be a mistake here to try to review
clinics atlhat time were neurotics; the remain- the several hundred ,studies 'that have been
ing'18 percent were personality disorders (per- done with this "scale." Several reviews are
sonal communication). It is possible that the available (Razin; 1971; Ratin, in press; Char-
personality disorders are responsible for the tier, 1971).
negative relationship observed between the As an example 'of the problems, and .the
"A-B 'Scale" and therapist's, success. This frustrations, we, will consider one study.
would be the case if there were .a strong Draper : (1967) attempted to replicate the
negative relationship among, say, six or eight Whitehorn-Betz research.. Instead of trying to
Work with records, he had these young, rotat-
'If this were a real detective story. I would allow the reader ing, medical interns rated by their psychiatric
to form his hunch that this reversal was due to the scale's supervisors on the dimensions that Whitehorn
multiple meanings. While some of the reversals to be
discussed appear to be due to its multiple meanings, this and Betz had extracted from they (page
one.appears not to be. 16 above). His findings tended to, confirm
 EXPLORING,THE PSYCHO-SOCIAL THERAPIES

the Whitehorit-Betz findings.. However, his A great advance was made by Whitehorn
data from the Strong Vocational Interest and Betz when they sought Systematic obser-
Blank showed relationships reversed from vations. Many thoughtfUl clinicians had con-
those of. Whitehorn and Betz. Paradoxical as cerned themselves with the personality of the,
these 'findings are, they are exactly what therapist as a. critical variable, but it was
Would be expected from the findings to be Whiteborn and Betz who collected systematic
presented below. observations on more than a thousand patients
It has been suggested tha the "A-B Scale" is and several score of therapists.
a waste of time since it was developed chiefly The state of psychotherapeutics in, the early
in a period before the introduction of the 1940's was not unlike the state of aeronautics
neuroleptic drugs. The impact of drug treat- in the early 1900's. Wilbur and Orville Wright
ment upon psychotherapy is an important did not have any college education, but they
issue in this research. However, it is probably did pursue some systematic observations.
Unreasonable to assume that all schizophrenics They "set up a wind. tunnel Where they could
are going to be drugged heavily and regularly, observethe performance of airfoils. What they
now and forever. We know that therapists dif- learned was quite contrary to what was then
fer in their use of drugs. Some prefer to make "known," and it was disquieting to the
minimal use while others feel that dosages academic experts of their day. 'Moreover, their
should be "adequate." way of learning, systematic empirical'observa-
If it were true that all schizophrenics are to tion, remained for decades the only, satisfac-
be drugged, it would serve to magnify the im- tory approach to the problem. It would be
portance of these data. Researchers every- many decades before accurate theoretical
where agree that the psychoactive drugs mask predictions could be made about airfoils.
the 'disorders, aild make it difficult to know While the task of the Wright brothers was
what is disorder and what is drug effect. Many not an easy one, still it was possible for them to
researchers go to great pains to collect data prove dramatically and conclusively the
from patients who are atleast temporarily free superiority of their knowledge over the
of drugs.. Viewed'in this light, the Phipps Clinic prevalent knowledge they flew an airplane:
data betome a vast resource that would be Would that it were as easy to demonstrate the
most difficult to reproduce. worth of the findings of Whitehorn and Betz!
Chapter III
The First St udy---A Pilot Investigatioq
My first efforts to find the correlates of the 4. That "young male adults" might have
Aq3 predictor were as by-products of another personalities different ,f from those of psy-
investigation. This other study does. not con- chiatric residents at the Phipps, Clinic. .
cern us except to say that it was a study of pic-
tures drawn by mental patients. I became. in This first Personal Tendencies Question-
terested in the way, "raters" reacted to these naire (PTQ) contained in addition to the 18 A-B
pictures (Dent and. Kwiatkowska, 1970). To items, 'items from 12 personality scales. Ex-
measure the raters' personalities I put amples of these items can be seen in tables 3, 4,
together a questionnaire derived largely from ands. 2
Irvin L. Child's works in aestketics. The The respondents were (1) 26 male normal
resulting 12 personality dimensiorls were net volunteers living 'in the Clinical Center at the
only useful in that they were related to National Institutes of Health and (2) 34 males
"aesthetic judgment," many of them could Eilio in class in psychology at a community college.
be expected to define the way different people I wondered how these respondents might com-
react to different kinds of psychopatfiology. pare with the psychiatric residents at the
It would be of interest to understand how A , Phipps Clinic. One would expect mental health'
and B therapists react to various kinds of professionals to have an interest in how other
psychopathblogy. Accordingly, the question- people feel. Certainly, those who are not 'in-.
naire included 18 of the 23 A-$ items. Of the 13 terested in the .feelings of others are not likely
items in the main cluster, rejection of manual to become psychotherapists. There were eight
and mechanical activities, I included 8, and ex- items in the questionnaire which might in-
cluded the remaining 5 as redundant (table 1, dicate such . an interest, and these, eight items
page 17, Nos. 59, 68, 87, 121, and 189). were used to construct a scale called "Em-
The purposes of including the A-B items in pathic Interest." This scale became the key.to
this questionnaire were frankly exploratory. the analysis.
Quite independently of reactions 'to pictures, I If we look at the 60 respondents as a group,
sought the correlates of the main cluster, and I there are no significant correlates of the A-B
hoped to-tease out some other clusters from predictor. However, when we looleteparately
among the remaining 10 items. The assump- at those with high empathic interest and those
tions were: with low empathic interest, some interesting
_

1. That the A-B predictor is multidimen- possibilities emerge. These results are
sional; i.e., it contains more than one ern-
presented in tables 2 and and. can be sum-
marized as follows:
,.pirical cluster.
2. That the dimensions might well represent
various aspects of dealing with schizo- 1. In table 3, there are hardly any correlates
phrenics*. of the A-B Predictor among those with
3. That the main dimension, rejection of loW empathic interest, but somewhat
manual and mechanical interests, could more among those with high empathic
not possibly mean the same thing to interest. This is what we would expect if
women as it 'means to men. Since the the subjects with high empathic interest
Phipps residents were predominantly were more like the' original validation
male, only 'males were to be used in group (Phipps residents) than those with
analysis. low empathic interest.
21
 22 EXPLORING THE PSYCHO-SOCIAL THERAPIES
9

Table 2.Intercoirelation of Selected Scales

(Pearsonlan es)

31 College Men 29 College Men

With High With Low
Emphatic interest Emphatic interest
8 A-B items 8 A-13 items
representing 9-item representipg 9-item
rejection of . F Scale rejection of F Scale
mechanical mechanical
interests interests

A-B predictor (18 items) .90 .39 .88 c .05

8 A-B Items representing rejection
of Mechanical interests .47 .as
Significant at .05 level (two:tail).
= Significant at .01 level (two-tail).

2. 'In table 2, for those with high empathic social affairs. They were scored as a scale for
interest, one of the 12 personality scales item analysis.
is correlated with the A-B predictor. Tables 4 and 5. present. item analyses' for
There is a negative correlation with the F eight items in the main scale, rejection of
Scale, Authoritarianism, which is an indi- manual and mechanical activities, and the six
rect measure of ethnocentrisin (Adorn° et items in the active social- area. The itemAor-
al., 1950). related with the "reject manual" scale suggest
3. The items representing rejection of nonauthoritarianism, tolerance for complexity,
manual activities dominate the predictor kegression, and so forth. The items correlated
in both halves of the. sample:- In the high- with the "active social" scale also include one
empathic-interest group, the rejection of from tolerance for complexity, but the dign of
manual activities explains all of the cor- the 'correlation is reversed. In fact the items in
relation of the predictor with the F Scale table 5 reveal an active structuring, though of
(table 2). course not a complete reversal Of the tolerance
4. In table 3, there is a tendency for the item in table 4. Thus it appears that the A=B predic-
correlations to reverse as we pass froth tor contains at least two scales that are very
high-empathic-interest subjects to low. (In different from each other.
theext-chapter, this tendency of the "Ai The findings serve as background for the
B Seale" to reverse its meaning will be study to be reported in the next chapter. The,
called "the semantic reversal:') findings thus becomassumptionswhich could
Considerable attention was given to the 10 be elaborated in clinical terms as follows:,
A-B items not part of the Main cluster. Six of 1. The A-B predictor is limited in applies:
the 10 items showed:a modest amount of inter- tion to settings which are primarily
nal consistency. The average intercorrela,tion psycho7social in their treatment orienta-
was .22. These items include (1) liking-4o be tion. It will not "work" in somatically
president, (2) drilling in a [military] company, oriented. settings, and indeed might even
and (3) interesting the public through public reverse its meaning' in such settings.
dresses. The A's also feel that they can (4) ac- 2. Authoritarianism stands for -"CustodiS1-
cept just criticism without getting sore, (5) cor- ism" (Gilbert and Levinson, 1956) and
rect others without giying offense, and (6) negative attitudes generally toward
followup subordinates effectively. These items serious mental disorder. This might well
seem to reflect a certain active initiative in be associated with a particular there-
 A PILOT4NVESTIGATION 23

Table 3. Correlation of the AB Predictor OS items) With Items in Other

Personality Scales

314.1en With High 29 Men With Low

-Empathic in113rest Empathic Interest
Scals(s) and Item Scale
Answer "A"
Ansi. .Gamma Ansr. Gamma.

F Scale and the Tolerance for Complexity Scale:

People fall-very' naturally into distinct classes,
such as the strong and the weak.. F .65 .02

Unquestioning loyalty is the first requirement of

good citizenship. F F .60 .01

. F Scale and the independence of. Judgement Scale:

What youth needs most is strict
rugged determination, and the will to work and
fight for family and country. F .48 .03

The Tolerance of Complexity'Scale

The man who truly loVes a woman must regard her as
F 48 .02 .34 .11
the best in the world in every important respect.
At the end of a textbook chapter, a good :Summary is
more valuable than 'a set of thought-provoking'
.32 .10 1 .14
questions.
The lndependenceof Judgement .Scale:
It is easy for me to take orders and do what I am
.48 .01
told.
The happy person tends always to be poised, courte
ous, outgoing, and emotionally controlled. F .43 .04

The best theory is the one that has the best

.62 .01
practical applications.
Some of my friends think that my ideas are
impractical if not a bit wild.
The Regression in the Service of the Ego Scale:
I can detect in myself no strong antisocial
impulses of the sort which, under certain circum-
stances, might lead to crime. .64 001

in_listening to .a lecture, I often am amused by

-thoughts of double meanings or_ possible puns which
the lecturer probably doesn't intend. 4- .69 .04

The Tolerance of,Ambivalence Scale:

s A beautiful sunset would be still more beautiful if
it lasted longer, and were not a tragic reminder of
. how transitory everything good is. F .64 .02

Gamma is not recorded unless p value is. .15 or less.

/7
24 EXPLORING THE PSYCHO-SOCIAL THERAPIES

Table 4.Correlation of the Score for Eight A-B items Which imply Rejection
of Mechanical interests With items in Other Personality Scales Among
31 College Men With High Empathic interest
Scale "A"
Scale and item With Which the "Rejection" Score Was Correlated An. An. Gem-
mel* swer ma p

FScale and the. Tolerance for Complexity Scale:

Unquestioning loyalty is the first requirement of good citizenship. F .79 .0006
People fall very naturally Into distinct classes, such as the
strong and the weak.

F Scale and the Independence of Judgement Scale:

What youth needs most is strict disciplin , rugged determination,
and the will to work and fight for family a d country. F .62 .003

The Independence of Judgement Scale;

The happy person tends always to be poised, courteous, outgoing
and emotionally controlled. F F .51 .02
It is easy for me to take orders and do what I am told. F F .45 .02
Science should have as mu"ch to say about moral values as religion
does. .42 .04

The Regression in the Service of the Ego Scale:

I can detect in myself no strong antisocial impulses of the sort
which, under certain circumstanoes, might lead to crime. F
I enjoy letting my thoughts wander 'aimlessly, and find myself think-
ing about all sorts of unusual and unrelated things. T .03

The Tolerance of the Unrealistic Experience Scale:'

Optical Illusions and other experiences that put you in conflict about
?hat is real and what isn't are on the whole quite enjoyable. `.45 .05

The Tolerance of Ambivalence Scale:

A beautiful sunset would be still more beautiful if it lasted longer,
and were not a tragic reminder of hov transitory everything good is. F F .56 .05

Need for Order:

I prefer that my hours of eating and sleeping be regular, not
changing from dai to -day. lincluded for comparison with table 5). F .40 .18

peutic style, say, for example, efforts to specifics will not be confirmed. We shall find
decrease symptoms as opposed to trying that empathic interest does. not necessarily
to understand dynamics. See item (2) on distinguish the psycho-social from the somatic
page 16. orientation, and that authoritarianism is dif--
3.- Some sort of active social initiative may i ficult to measure among mental health profes-
be involveii, perhaps reflecting the "parti- sionals: Few mental health professionals
cipation" discussed. kr Whitehorn and dorse any of the items in the F Scale. But men-
Betz. See item (3) on page 16. tal health professionals do vary on dimensions
which are usually related o-the F Scale, dimen-
While the general outlines, of these assump- sions which piobably-are correlated with their
tions will be confirmed below, some of the reactions to types of pathology.
 25
A PILOT INVESTIGATION .

Table 5.--torrelation of the Score for Six "Other" A-B items With Items In.
Other Personality Scales Among 31 College Men With High Empathic interest
ASAP!
Scali and itein With Which the "Other Score Scale , .
.

Was Correlated, An- An- Gam-

swer swer ma

The Tolerance. for Complexity Scale:

No one can be sure of conquering his diffi-
culties; willpower is not enough. .48 .02

The Regression in the Service of the Ego Scale:

When Finn alone orperfOrming a task'requiring'.
no mental work, enjoy doing things like
rhyming wards, discovering puns, and makirig up
.53 :01
tittle songs.
Need for Order:
I prefer that my hours of eating and sleeping
be regular, not changing from .day to day..
.66 .02
(Cf. table 4)
Preference for Decision Making:
I prefer a job that requires making difficult .
deCisions rather than f011owing instructions
T T .44 .04
carefully.
Preference for Work:
I;Prefer to climb a flight of stairs rather
than taking an elevator up. .44
4-
 Chapter W

The Second .Study--Hospitalized

Schizophrenics, DepregAives, and Neurotics,
The pilot investigation was sufficiently in-. timal but are nevertheless useful for our pur-
tripling that I decided to, design a larger-cor- poses here.
relational study of mental health professionals This revised PTQ, including a larger number,
and other relevant groups. Actually, much of of personality scales and. revised A-B scales,
this second study was not "by design." Rather, was designed, for administration to a variety of
as is the case so often in exploratory studies, groups representing various degrees and kinds .
the research opportunities unfolded as the of mental health specialization. I, hoped that
study proCeeded. TO understand the limita- the psychiatrists who participated in Ahe
tions of this study, it is necessary to deScribe original. Whitehormlletz studies would be
its original design. among these groups.

Study Design Design of the Personal Tendencies

Questionnaire (PTQ)
The study was oriented to the schizo In order to reach relevant groups th' ques-
phrenias. (See appendix 2' which 'was written tionnaire had to: be short, short enoug
about the time that this study was designed.) the average person could answer it in ha
The fact that the "A-B Scale" might have im- hour. This was achieved, but only by the most
plications for the treatment of the neuroses careful selection of content. Not only were per-
was given some consideration, but, the per- sonality scales selected with care, each item
. sonality scales-included were chosen primarily was required to justify itself. For example, if
because...of their: possible relevance for treating an item were such that most psychotherapists
schizophrenici. answer it in the same way, it was excluded,
I hoped to have the opportunity to .reanalyze ' since it contributes little to measurement.
the original Phipps Clink data, using more (This controversial. procedure is, discussed 'at
powerful and more accurate statistical tech. length sin appendix 1.) The PTQ S presented in
niques which might yield a largerand better set appendix 2:
of A-B items than the original 23. This The personality scales included are
reanalysis was Made possible by Dr..Joseph H. presented in table 6 and the items are
Stephens who has Preserved the data; checked .presented in appendix 4. As presented,. they I.
and auguniented it, adding more patients and reflect cortiderable analysis of internal con-
more doctors, and assembling much more' infor-, sistency that,was conducted after the data had
mation about the patients and the doctors (e.g., been collected. The appendix tables indicate
drug therapy, process/ponprocesi, etc.). The how each item was selected,-and if not used as
reanalyses presented below continued after it was originally selected, why it was not so
the revised Personal Tendencies Question- used
naire (PTQ) was put into the field, because There are'
re doubtless many, other facets of the
More and more Phipps data became available, treatmentiof mental disorders that could be in-
permitting more accurate analyses of the 5orporated into such a questionnaire. ClearlY,
Strong Vocational Interest Blank (SVIB). Thus we can get from a questionnaire only what has
the A-B items included in the PTQ are not op- been built intb it
26

39
 Table 6,Scales Included In the. Revised. Personal Tendencies Questionnaire

No. Coel.

fabl /of )Mphe

Source
No. )terns

31 1.5 .49 Child "Tolerance for Complexity"

Need'for Closure

Regression in the $erVIce of the Ego 32 / 22 ,?7 David Singer

33 7 .24 Various sources

Empathic Interest
34 5 A6 iviireis and Garrett."ProteStant Ethic"
Work Ethic ,

Need for Order, 35 5 .39 Various ce°

Tolerance of the Unrealistic' Experience 36 7 .56 Child "Tolerance otthe 'Lino:Ostia Experience"'

37 7 .25 Child "Tolerance of Ambivalence"

Tolerance of Ambivalence
38 4 .32 Child.."Preference for DecisisMaking"
Preference for Decision Making

High Personal 'Standards 39 Written for the PTO' IF

40 7 .39 Child, "Super*" z
Remorse

41 16 11 Adorn,. "F Scale" plUithres others

Antidemocratic
42 .51 Barron "IndOpendence of Judgement"
individualism

Deference Anxiety 43 7 .53 Child "Deferenceilnxlety"

'Child "Nt,irturanoe Anxiety" tii

ikturturance Anxiety 44 3 .38
F
Science and Moral Values 45 1 Barron "Independence of Judgement"

46 3 .50 Child "Preference for Intellectual Challenge"

Preferehce for intellectual Challenge
,
Extroversion 47 7 .74 Child "SoCiabllity"

`48 4 .55 Rokeach "Dogmatism"

Fervor

Life Satisfaction 49 5 56 Written 'for the \PTO. z

50 3 34 Child, "Preferen6 for Woe
Work Activity
1013,17, Stronglocatiortal Iniereit Blank (SVIB)
Various 1kB Predictors gild SCales

training,
*Coefficient Alpha Is computed for 133 psychiatrists, psychiatric residents, arid clinical psychologists in
V

.10
 EXPLORING THE PSYCHO-SOCIAL THERAPIES

The Respondents The groups of respondents are shown in

table 7. It was assumed that the first group,
Since it is assumed that the A-B predictor the original Phipps Clinic psychiatrists, would
has different meanings for different groups, it have a psycho-social orientation since that was
is important t6 be able to describe the groups. the .orientation of the Clinic when they were
A great deal of atikeetion was given to response there. It turns`out that they now vary in their
rates and to increasing the response rates in orientation.
order to be able to define as precisely as possi- In order to get a sample of somatically
ble the nature of the groups that were studied. oriented psychiatrists, I turned to another
It was assumed that, somehow, the ,psycho- survey being conducted, at the time: The
social vs. the somatic orientations are involved Followup Study of Former NIMH Trainees and
in understanding the A-B predictok. But it was Fellows. In that questionnaire there were some
not understood how difficult a task it is to gtt questions about types of therapy utilized. A
an empirical handle on these orientations. number of. male psychiatrists were selected (in

Table 7.Male Respondents to. the. Revised Personal

Tendencies Questionnaire
Group Response _Description
Psychiatrists 26 45% The orginal Phipps Clinic
psychiatric residents, exclud-
ing those not now in clinical
work. Collected by mail.
Psychiatrists 27 56% Matched to above on age, but
. selected to be more somatically
oriented. Collected by mail.
Psychiatric residents 36 ? Collected by training directors.
Subtotal 89

Clinical psychologists in training 22 Collected by training directors.

Counselors, house parents 51 910/0 Loysville (Pa.) Youth .Develop-
ment Center. Collected by.
George A.'Furse wl Me living
at the Center.
Normal volunteers 22 76% College student s living at the
NIH Clinical Center.

In addition., there were some female respondents: 5 psychiatrists, 17 clinical psychologists in training, and 21 normal
volunteers.
in the text, "133 mental health profesSionals" includes the following groups: 89 male psychiatrists or residents, 5
female psychiatrists, 22 male and 17 female psychologists in training.
Questionnaires were collected during the perjod June-December 1973 except for the counselors' and house parents'
which were collected in July 1974.
to,
 29
HOSPITALIZED SCHIZOPHRENICS,' DEPRESSIVES,-AND NEUROTICS

the 'same age range as the former _Phipps not at all comparable to those presented in the
residents) who indicated that they' used last chapter.
chemotherapy at feast as much as any other
kind of therapy. This indicator also turns-but to
be inadequate to the purpose. Female Therapists
The third and fourth groups are mental
health trainees. SinCe A713 studies have been While most studies of the psycho-social
done on psychiatrk residents and clinical therapies have been conducted oh male
psychologists in training, it seemed desirable therapists,' there are probably more females
to collect some questionnaires from these than males engaged in the psycho-social
groups. For this purpose, some clinical training theia,pies.
directorsdconsented to give the questionnaire The A-B predictor comes from a version of
to their trainees. The actual response rates for the Strong. Vocationsil Interest Blank (SVIB)
these groupS is probably quite variable, very which was designed fbr males, .Moreover, the
high for some groups and low for others. content of the predictor-is male oriented. Even
The nexegroup in table 7 are the counselors in these liberated days, it is unlikely that "not
and house parents at the Loysville (Pa.) Youth wanting to be a machinist" means the, same
Development Center. Most of the eligible staff thing to males and females. Moreover, many of
Were persuaded to participate. For this group, the petsonality scales in' the PTQ were
we have not only the PTQ, but a variety of validated on male popnlitions and this shows
other- data as including their success in the wording Of certain questions,
rates. Together these data constitute a pilot It was my hope that if we understood the
study developing methods for applying "A-B Scale" we could generalize it to, females.
therapist personality research to a milieu set- Some females 'were included in the study, and
ting. This self-contained pilot study of some efforts to specify the limits of generaliza-
therapeutic influence is reported separately in tion are presented in appendix 7.
chapter V. But I cannot feel sanguine about these ef-
The last grbup is one of "normal volunteers" forts. It is possible that the sex of the therapist
at the N.I.H. Clinical Center, college students itself changes the therapeutic relationship.
who serve as live-in "controls" for medical ex- -What is needed is a complete study comparing
periments. They were cooperative and consien- women who are variously successful with dif-
tious participants in this research and were in- ferent types of patients. In the meantime, my
.cluded because college students have been apologies for concentration on the males in this
Fused as 1.ipseudotherapists" in a large number study. Until that larger study is done, we can
of,analogue studies, some of which have tended hardly feel confident about the meaning of the,
to confirm the original Whitehorn-Betz find- various A-B predictors. and clusters for' female
jogs. therapists.
The characteristics of the normal volunteers
changed dramatically between the time that
the pilot investigation was done (chapter III
above) and the time this study was undertaken. Psycho-Social vs. Somatic Orientations
In the earlier period, they were typically
freshman and sophomores who had not yet As was indicated in chapter III, the pilot in-
chosen a field of specialization. Their inten- vestigation suggested the possibility that the
tions were about as varied as for any cross sec- A -B.' predictor has different meanings for dif-
tion of college students. By the time the second ferent groups, and that the 'meaning might well
study was begun, normal volunteers had be different for therapists with a somatic
become typically juniors and seniors, many of - orientation from those with a Psycho-social
Whom were oriented toward the helping pro- orientation. This might, for example, explain
fessions. In fact, nine of thege volunteers were the null findings in certain hospitals. Accord-
psychology majors, and nine were preinedical ingly, the orientation of the therapist was to be
students. Thus, these normal volunteers are , a major control in the larger study.

4n
30 EXPLORING THE PSYCHO-SOCIAL THERAPIES.

Actually, there are three dimensions: the If, however, the psyc hiatrists (the first two
psychological, the social and the somatic groups in table 7 combined) are sorted into two
(Strauss et al., 1964). Among psychiatrists and . groups, those whose primary affilitation was
psychologists, the psychological and the social with a, medical school or hospital, and those
tend to be negatively correlated with somatic, priVe practices, clinics, or CMHC's there was
and there is therefore some justification for a-lot-quite-significant difference in response
combining the three into a single dimension raris: 45 percent to 58 percent. This difference
from psycho-social to somatic., For discussion was puzzling; most of the first group are in
purposes, the dimension is treated as a medical schools, while most of the latter are in
dichotomy, but we must keep 'in mind that any private .practice. A psychiatrist in a medical
particular group can fall anywhere on the school probably feels a greater need to respond
dimension. to an NIMH survey than does a psychiatrist in
Truly somatically oriented psychiatrists are private practice. Thus, the "real" difference
a small minority, but they carry large patient might be larger than that observed.
loads, and are therefore an important minority. Finally, the Follow-up Study does include a
They are not particularly interested, in inter- qnestion about interest in psYchopharma-
view studies and questionnaires.:Henry, Sims, cology. The responses to this question also did
and. Spray (1971) abandoned their objective of not correlate with reported use of chemo-
studying such doctors because only 29 percent therapy. In fact, among the four variables
of them consented to be interviewed (personal hospital/nonhospital, empathic interest,
communication). chemotherapy, and interest in psycho-
I recognized the problems of trying to deal pharmacology -- there are six possible correla-
with these varying orientations. For several tions? but only one of them is significant.
reasons, I chose not to attempt to measure Hospital/nophospital is correlated with in-
orientations directly by including appropriate terest in psychopharmacology, .32 (gainma
questions in the PTQ. (1) The PTQ was already significant at .05, two-tail).
too long. (2) It is a general questionnaire free of It was probably naive to think that empathic
specific controversial clinical issues. (3) In- interest would correlate with orientation. But
troducingsuch issues might color responses to why is chemotherapy so little correlated with
general questions, and it might damage others? It is possible that chemotherapy is
response rates. more a function of institutional requirements
As an indirect measure, "empathic interest" than of the doctor's preferences. We shall find
was expected to be higher among the Phipps other support for such an interpretation.
Clinic graduates (assumed to be psycho - social) The reader will forgive: me for jumping
than for the second group who had indicated in ahead in time to a portion of the reanalysis of
the Follow-up Study that they used chemo- the Phipps data. The material logically belongs
therapy at least as much as any other type of here, although it did not become available until
therapy. It was disappointment and consterna- long after the PTQ's had been collected.
tion when I discovered that this "somatic" In January 1975 Joseph H.'Stephens made
group; reported empathic interest as high as available some data for the former Phipps
that reported by the Phipps graduates. residents showing the proportion of schizo-
At the same time, I was trying to compare phrenic patients for whom, drugs were pre-
respondents and nonrespondents. There were scribed, as well as other characteristics of pa-
no differences between respondents and tients (ECT, sex, and so forth). These data
nonrespondents in background characteristics, made a number of new analyses possible. For
or even, for the Phipps group,. in A-B scores present purposes we can ask the question;
from the SVIB's they had taken years befort. Does the doctor's prescribing drugs relate to
Specifically. there was no difference between his later returning thelnQ?
respondents and nonrespondents in reporting . There are some qualifications to the answer.
the use of chemotherapy in the Follow-up The Phipps Clinic residents were a very
Study. This seemed odd in view of experiences unusual group. For example, nearly all of the
other researchers had had. ' later residents are now associated with
 31
HOSPITALIZED SCHIZOPHRENICS, DEPRESSIVES, AND NEUROTICS

'Table & PTO Response Rates, for Former Phipps Clinic. Residents
°Ai, Returning
Whether Resident Prescribed. the PTO
Drugs "for Schizophrenics

.60%.
Prescribed no drugs-

Prescribed drugs, but for less than half of his patients

Prescribeddrugs for half or more of his patients 20%

medical schools. This makes it useless to try to in psychopharmacology), hospital/nonhospital

check their earlier drug-prescribing with their is the most attractive, (1) It does show some of
present hospital/nonhospital status as we the expected difference in response rate. (2) It
would like to do. Since it 'is one of the few cap- is available for all psychiatrists and psychiatric
tive samples of psychiatrists of varying .per- residents. (3) It is easily replicable. (4). And it
suasions, we must learn what we Can from it. can be used to classify studies. For example,
Table 8 shows -startling differences in the Draper study discussed on page 19' ih clear-
response rates. Of those who prescribed no ly. a : hospital study.- Moreover, since the
drugs, .60 percent returned the PTQ. Of those therapists were medical interns, they Were un-
Who prescribed drugs for less than half of their doubtedly omatically-oriented as a group. We
schizophrenic patients, 50 percent responded. will therefore 'use the hospitanonhospital
And of those who prescribed 'drugs for half or groups as reflecting the somatic vs. the psycho-
more of their schizophrenic patients, only 20 social orientation. However, we should note
percent returned- the questionnaire.' Thus we that somatically oriented psychiatrists,
get the expected relation between chemo, because of their lower response rates, are prob-
therapy and response rate. There is possibly ably underrepresented in the hoipital group.
an age effect here. The no-drug group are Another way of saying it is that the hospital
older, having done their residencies earlier.. group is probably nearer to the middle' of the
Other surveys have shown that older people dimension, psycho - social to'. somatic. Thus a
are more likely to respond, but the differences partial reversal in 'correlations between the
are never of this orderl, hospital and nonhospital groups, may be sug-
As noted, above, the current report of chemo- gestive of a complete reversal if we could get'
therapy use from the Follow-up Study does not data from those with a fully somatic orienta-
correlate with the Phipps residents' current tion. '

behavior in responding to the PTQ. But their For the clinical psychologisti in training we
_earlier (by about 15 years) drug-prescribing as do not have data on hospital affiliation. We will
found in the Phipps records does predict. therefore concentrate our attention 'on the 89
.strongly their current responding. Needless to psychiatrists and psychiatric residents, 46 of
say, there is no correlation between their cur- whom are in nonhospital settings-and 4$ are in.
rent report of chemotherapy use and their hospital settings.
earlier drug-prescribing. Nor is there any cor- Further discussion of the somatic vs. psycho-
relation if we consider only those who -Com- social orientation is presented in appendix 8.
pleted their residencies in 1959 or later. All of
this casts some doubt on whether their current
report of chemotherapy use reflects their own', What Is the Meaning of the "A-B Scale"?
professional preferences. Table 9 presents the ,PTQ correlates of the
Of the four variables (hospital/nonhospital,
empathic interest, chetnotherapy, and interest "&P Scale" for various subgroups in the pres-
111.

4 1".0
 82 EXPLORING THE PSYCHO-SOCIAL ,THERAPIES

ent data.' We Will. be interested in the pat- ble that other personality measures not includ-
terns presented in this table, not the specifics. ed in the PT_ Q might correlate as well or better
The patterns presented there have been found than these measures do, What we can say is
to be stable over a variety of analysis formats. that the scale has different meanings for dif-
If we look first in table 9 for the underlined ferent groups. (It is different again for the
coefficients, those significantly different from counselors and house parents.) It cannot be
zero, we are drawn to column 3, the nonhos- generalized, but there is a greater likelihood
pital psychiatrists. Almost a third of this group that undergraduate psychology majors will
are from the original Phipps residents. Almost reproduce its original meaning than will
a fourth of column 4 are also former Phipps psychiatrists in hospitals.
residents. In all of this complexity we need to keep
The nonhosPital psychiatrists and psychiat- separate two aspects of the reversal of the A-B
ric residents in column 3 are presumed to be predictor. The functional reversal refers to the
nearest to the psycho-social orientation of the differential hypothesis that the A's are more
Meyer-Whitehorn Phipps Clinic.If we were to effective with schizophrenics, while the B's are
break column 3 into three groups former more effective with some other diagnostic
Phipps residents, other nonhospital psychia- group, .notably, V.A. outpatients. The semantic
trists, and nonhospital psychiatric residents reversal refers to the fact that the correlates
there are no significant differences in means, of the A-B predictor are reversed in sign in dif-
or in covariance Matrices among the three ferent groups of therapists.
groups. As between the groups in. columns 3 We have no,,'evidence that the functional
and 4, there are only a few differences in reversal is simply a semantic reversal: It is
means, blit there are sUbstantial, consistent reasonable to assume that the Phipps Clinic
differences in covariance matrices. and the V.A. outpatient clinics were similar in
There is a tendency for the signs of the cor- having a psycho-social orientation. We will
relations to be reversed between eolumns. 3 need to look at this question in more detail
and 4. Thus when theSe two groups are combin- when we have all the findings before us.
ed in column 1,. we get correlations which are We will turn away now from the original
smaller Ind less representative of either 23-item A-B predictor. It will appear in some
group. We conclude that for psychiatrists as a tables for information purposes. But our atten-.
whole the "A-B Scale" has very little meaning. tion will: be toward new predictors and scaleS
Whatever meaning it' has for nonhospital which will tell . us more about the differing'
psychiatrist- the meaning ..for hospital modes of treating patients. Aetually, we seek
psychiatrist tends to be just the reverse... to dispose of all: the multisemantic SVIB
If we look now at the nine psychology majors Clusters and replace thein with personality
and the nine premedical 'students we see a `,dimensions which have mo re meaning for
Similar tendency to reversal. Moreover, the therapy. tut we must not didcard the. SVIB's
nine psychology majors are' More like the until we have learned all we can from theni
nonhospital psychiatristS than either of the ,For that we will study. the .Phipps datit'in
other two groups. greater. detail than has previously been done.
We cannot say'from these data what precise- But one caveat before we leave the original
ly the meaning of thel"A-B Scale" is; It is pOssi- "A43 Seale." The fact that it reverses its mean-
ri
ing does not imply that the treatment of schizo-
, !The "A-B Scale" is the WB-22 Predictor, 22 of the 23 phrenics should be reversed. This question
items in the original scale. One item was excluded from the gets-more attention in the last two sections of
PTQ because It had not performed well in a number of this chapter where it will be shown that the
analyses. Also, as explained in detail in appendix 1, 'the W-B 23 fails to predict success with schizo
WB-22 Scoring is some hat different from that for the phrenics when drugs are prescribed, but where
- 2!-item 540 re,ented in table 1. In spite of these dif- it will also be shown that the correlates of the
fereoepS As *22 i correlated .964 with the scores on
uted by Barbara Betz for.the original .,W-B 23 are similar to the correlates of an ap-
0,iiippisrrisiiii!;n. It is therefore a reasonable substitute for propriate predictor Cr success with schizo-
lhotter phrenics when drugs are prescribed..

4,t) .
 PTO Personality Scales
Table 9.-Correlation Between the WB.22 Predictor and.

(Parasols') r's)

(8)
(2) (3) .
'.(a) ' A517.' OF ,

p for Volunteers pfor 'z ',,

: ,' Psychlat.rists.
..

(3), PaychOli .(6) ' (5))

Normal None

minus ogy Pre' minne.' i

Hospl Hospi.
.

Psychia. Volun.
1:.1:'1''! :::,

tiers MI MI (4) Majors. met U.) .. r

trials
,

PTO 'PersdnalRy Scale

g
ttl

18 46 '43 9 9 o
Numberof Respondents . 89 .

** .06 -
* z
-.43 -.71'
Need for Closure .
.43 . -.15 / ;04
tqw
*it* 0.
'Ti -.16 1T., -.24 .-17: .38.
4J.
Regression in the Service of theEgO
*a*. ..
-19
.33 ,.,f27. .08
Empathic Interest .-,..04 ..09... .
o
.
- ** 19 rii
/ ..06
Work Ethic ,. '. -12 .34 -.44 ,.07
z
-.13 -737 -.01 * -.50 72 ,..,,

n
..,-1-8
.

Need for:ONO ... '.

.
** .14
.22' -.31 ..21 -.02
ToleranCe of the Unrealistic Experience .!-.06 ' .M7
v
.13 -3.. 38 .7.03 ' tii
, ,M4 .' .15
Tolerance Ambivalence. ro

-.24 .14 .03 , .24 -.40 z

Preference for Decision Making .09 m
.18' .22 .37
/2 .30 ..28
High Personal Standards *ft
-.29 ..09 -.59 ,79
Remorse -.11. .25

7-.01 -.03 ..,.18 23 .03 -13 P

'Antidemocratic ,

Individualism M1 -.13 ; .24 -It ** -.03 .20 >

z
'.02 .17 .04 o
.16 .08 .28
Deference Anxiety **
-,93 z
.26. -AO .36 . Ai .33
Nurturance Anxiety

Science and Moral Values II " -.32 -2 .05 . .09 . -16 v

z
-,29 ,50
-.17 -A4 -.16 -.18
Pre,ference for Intellectual Challenge
''' .48 .3
17 .40 : ,;1 .34
Extroversion

different from zero at' the .05 level,

Underlining indicates that a correlation is significantly

columns being compared differ from each other at the ,10 level,,
* The two correlation coefficients In the

They differ at the 45 level,

` They differ at the ,01 level,

coefflclentsi.ere rounded to two digits,
Significance levels were computed, before the

48
 EXPLORING THE PSYCHO - SOCIAL 'THERAPIES

Further Analyses of the Phipps reduced and the number of items increased on-
Clinic Data ly a little as a result of the better techniques:
At the time, I was having long talks with Dr.
.When I asked Dr.. Joseph H. Stephens about Whitehorn and with Di: Betz. It became evi-
the possibility of doing further analyses of, the dent that they had done considerable work on
Whitehorn-Betz data, I had two things in mind: the therapists' success rates with depressives
First, :I Was curious Oda how 'well the A-B and with neurotks: They were aware and had
predictcor worked for the female therapists. noted in some of their writings that doctors
Second, I hoped to increase the number of were not uniformly successful with the three
items the A-B predictor and thus facilitate types Of patients. They felt that depressive and
the isolation of more scales within the predic- neurotic improvement rates tended to be much
tor. As. I. proceeded still other important pur- higher than schizophrenic 'rates, and that the
poses presented themselves: developing variation in- the depresiive and neurotic rates
similar scales for depressives and neurotics, was of sufficient to warrant analyses. It is my
and controlling for drug therapy effects. belie h that we ought to let the data ,koit, us
I was delighted to discover that Dr. whether the variation is significant; thaeis, in a
Stephens had acted not simply as a 'caretaker bootstraps science we simply ask whether the
of the Whitehorn-Betz data. By systeinatically results of analysis are, meaningful and sug-
Searching the files, he had been able to add gestive of further hypotheses.
more patients and more therapists. And he had Accordingly, I asked Dr. Stephens whether
vastly increased the number of other measures there were in the files any rhea on the doctors'
available for each therapist success -rate -S with depresaives and neurotics.
As for the female therapists that had been Indeed he found such data for nearly all the
inclAded in the original analysis, I had con- doctors. (These "percent-imprOved" rates 'are
siderable doubts that a predictor, the central the original work of Whitehorn and Betz. They
dimension of which was the rejection of manual have not been cheaked by Stephens.) For most
tasks, could mean the same thing for them as it of the doctors, then, we. haVe 'three "percent;
meant for males. Perhaps it might in the improved" rates:. for schizophrenics, for
liberated 1970's, but hardly in the 1940's and depressives, and for neurotics. Each of these
1950's when "Rosie, the Riveter" of World War generates a set predictor items from the :
II fame was still a new phenomenon. Indeed it SVIB.
turned out that the A43 items that are predic- The depressives- include both psychotic and/a?
tive for men are not predictive for women neurotic depressiVes.Aibout equally divided:
(Stephens et al. 1975). Actually, the women as k The data currently available do-.114t permit us
group were slightly more effective with schizo- to compute separate indices for the two types
phrenics than were the men. Moreover, as a 'of depressives. This' is unfortunate, but still it
group they tended to reject the manual occupa- will be of use. compare the treatment of the
tions. As a grouip, then, they tended to ar- schizophenias4ith that of the depressions, as
tab:daily:reinforce the central dimension in the discuised On page 12 above%
original predictor:. HoWeer, Within the group The neurotics were' brought, to the phippa
.
of females, this dimension. is not predictive. Clinic from far and wide. They were very severely
As for the second objective, increasing the disturbed, probably, quite different from the
.. number of:iterris in the A-B predictor, I hoped outpatients of NcNair, Callahan, and Lorr,
that by using more of the data and more power- which are presumed to be Chiefly neurotic but
ful and accurate statistical techniques, I could included also some personality disorders (page
generate a larger number of items that relate 19). I did not expect, therefore, that these data
to success with schizophrenics. Actually, the on neurotics would necessarily constitute a
number of items generated is primarily a func- test of their finding,
.tion of the number of doctors. By eliminating At the time theleclatalar depressives and
female therapists and controlling for drug
. neurotics became .1:0e.ilple, the Personal
therapy (below), the number of doctOrs.
was Tendencies Questionnair was already into its

4t7
 35
HOSPITALIZED SCHIZOPHRENICS, DEPRESSIVES,-AND NEUROTICS

final formal review for the . U.S. Office of frOin tables 10-13. Taking out 12 doctors who
Management and the Budget. Some analyses of prescribed drugs makes a difference. This and
the new data for depressives and neurotics other findings to be reported confirm that the
were done quickly and an extra page of SVIB difference is important and should be explored
is was added to the PTQ. as fully as possible even though the number of
2 Meanwhile, I had been doing considerable doctors prescribing drugs might be con-
work with respect to the last of the four objec- siderably less that 35. First, though, we will
,
tives:.the introduction of the neuroleptics. Dr. explore the "drug-free therapeutic relation-
Betz felt, and Dr. Whitehorn.tended to agree, ships."
that _these drugs change the doctor-patient.
relationship in schizophrenia. The successful
..doctor's personality is one half of a relation- The Phipps DataDrug-Free
ship, the other half presumed to be a somewhat Therapeutic Relationships .r1
homogeneous rubricthe schizophrenias. If
the drugs change the latter, they change the The best way to understand the contents of
relationship, and Indeed also the former.. tables 10-13 is to consider the summary
By using the 47 doctors who completed their variables presented in table 14. Many of these'
residencies before 1959, I hoped to approxi -, variables are not independent of each other;
mate this drug-free relationship. The item e.g., the, same items -may be used in more thah
analyses. for -these 47 doctors are presented, in one predictor. It is not suprising, then, that
appendix 5. They are of. interest partly because there are -so many correlations which are'
they explain which SVIB items were included significant at the .01, level.. In studying this
in the PTQ. They are also of interest for com- table,. we will be concerned not so much 'with
parison with later analyses. significance levels as with whethera particular
In January 1975 Dr. Stephens furnished still coefficient' is 'high enough or low enough for
mere data: the percent -of each doctor's those purposes for which we, mean to' use it.
schizophrenic patients whq (1 were women, {2) The variables in table 14 are of several.
were procesd vs. nonprocess, (3) got ECT, (4) 'types. W- 23 and 22 are not part of the
got insulin shock, (5) got drugs. These data re- reanalysis. }Lather, they are closest to the "A-B
vealed that of the '47 doctors, there AV% 12 Scale" that has been used in other studies.
Who had prescribed drugs. Thus there were in They are included to show their relation to the
fact only 35 doctors .who had prescribed no new predictors and scales. The three MPRV
drugs for their schizophrenic patients. variables are the criteria of successpercent
The SVIB's were analyzed again, using the of patients improved for the three diagnostic
35 doctors only.. For, although the PTQ could groups. Reis from these three criteria that all
not be changed at this point, the scoring of the of the remaining predictors and chisters are
PTQ could. be changed.-The item analyses of derived. The three TOIL: predictors are all
the 35 doctors are showfn in tables 10-13. those items. in the PTQ which predict a
The datkregarding prescribed drugs are for- criterion: Since some items predict more than
schizophrenic patients only. We do not have one criterion, the TOTL, predictors are not
simitar data for the depressive and neurotic pa- statistically independent IA each other. On -the
tients. The antidepressant drugs were not' in- other hand, S, D, and N include only the items
troduced until .seN4ral .years after the distinctive to a single criterion. These predic-
neuroleptics. On the other hand, the neurolep- tors are specific to the specific diagnoses. The.
tics are also known as antiPsychotics. They'. OTHR variables include the items that were
might well have been prescribed for psychotic-i not included in the PTQ but 'should have been,
depressions. Therefore, the analyses of the 35 had these analyses of the Phipps data been
doctors is probabjy the best approximation to c.,.? finished before the PTQ's were .collected. The
lower part of the table concerned with
the "drug-free relationships" for all diagnostic
groups. "clusters," i.e., groiips of items from within the
various Prsdictors which are inteirrelated
.

Casual inspection shows that the tables for

47 doctors' 'in appendix 5 are quite. different with each other.

.1.
5 0.
 EXPLORING THE PSYCHO- SOCIAL THERAPIES.

Table 10.-SVIB'Items Correlated. With Therapists' Percent Improvement .

Patients Who Are Sahizophrenic,. Depressive, or Neurotic

(Doctors who did not prescribe drUgi for schizOphrenics)
36 N 34** N = 32**
Item Number Schizophrenic Depresdive.. Neurotic
SVIB PTO Item Gamma p Gamma p . GaMma
SD items: (6),
17 3.31 Building
. Contractor* --.45 .02 -.39. .04
19 3-32 Carpenter* -.51 '.01g... -.53 .01
56- 3-41 Machinist -.35 .05 . -.36 .05
60 343 Mechanical.
Engineer* (-.26 .15) -.45 .01
94 3-52 Toolmaker* -.45 .001 -.41 .03
188. 3 -71. Repairing Elec.
wiring .37.: . .05 .57 :004
155 Excursions .41 .06) .37 .09).
162 Animal zoos .62 .01 .37. .10)
209 , Adjusting 1
difficulties
of others .50 .10) ( -. 2 .09)
'DN items (4)
8 3-29 Auctioneer-. DBUSNS -.36 .09). -.44 .05
.103. Arithmetic . (-.38 ,.10) (-.58 .06)
105. 3-56 Bookkeeping 013USNS -.51 .01. -.52 .02
4-4 Foreigners NQUEST2 (-.37 .09) -.68 :004
280.' 4-7 Athletic men NQUEST2 (-.41 .06). ..04
287 Sell the
..machine -.50 .04 .09)
301 Luther Burbank
"plant wizard" 33 .08) ..37 .09)

imparenthesesare relations petween the .05 and the ..10 level-of significance: 7) . .

* One of the dottors did not have enough depresdive patients to yield a.rellable improvement rate; three. of t
doctors did not have enough neurotic patients. ..
.
.
Original 23 items.. .
 HOSPITALIZED SCHIZOPHRENICS, DEPRESSIVES,-AND NEUROTICS

Table 11. SVIB Items Porrelited With Therapidts' Percent Improvement

In Patients Who Are Schizophrenic
, (Doctors who did not prestribe drugs for schizophrenics)
N 35

Item Number
S1/15:Fra Item

13 Auto nepairman ,04

59 3-42 Marine Engineer* .
SQUST 4MQ1,7

150 Collecting postage stamps . 48

158.3 -64. ConVentions SQUEST .43 -102
'170 Snakes :56 t 004

184 ''3-67 *Social problem movies SXPRSS . 47

-189 3-72 Cabinetmaking* SQUEST -.37' 06
Continually .changing activities .38
225
283 Discover an improvement in th design of the machine .0
-.46
iF
.01
.02
289 Tech others s_the use of the achine

Chautfeur vs. Chef (B) .44 .01,--

323
341 3 -19Work involving few details vs. many details (A) -.47 .008
SXPRSS .01
352 3-28 Nights spent at home vs. away (A) ( .03
Accept just criticism without getting sore* SXPRSS .4e
367 1-2
SXPRSA .45
376 1-62 Able to meet emergencies quickly and effectively
SXPRSA AO '.03
383 1-68 Stimulate the ambition of my associates
Between the :0$ and the 10 level:
.53 .06
9 3-30- Author.of a novel
Lawyer, Criminal .32 .07 .,
51.
Locomotive Engineer .09
55
.32 .08
71 3-46 Poet
132 Shop Work -.30 '.10

Tennis .54 . 06
140
145 Poker -.31 . 09/

195 Arguments 7-.34 .07

,221 3-75 Expressing judgments publicly regardless of
criticism .
SXPRSA . .31 .08
independents in. politics .39 .08
276
291 Salary received for work -.31 . 09

Opportunity to understand lust how one's superior . .37 :10

297.
President of a society or club* SXPRSS .43 .08
311 1-18
'333 Tangible returns vs: activity for its own sake (A) -.32 .07

. .
(A) The 's prefei the second choice. (B) The B's prefer the second choice.
I Origins 23 A-8 Items: .
 EXPLORING THE PSYCHO-SOCIAL THERAPIES

Table 2. SVIB Item Correlated With Therapists' Percent Improvement

in Who Are Depressives
(Doctors who did not Prescribe drugs for schizophrenics)
N = 34

Item Number
SVIB PTO -.Item Gamma

29 Dentist -.42 .04

p8 3-37 Floorwalker DBUSNS -.81 .01
42 Hotel Keeper or Manager -.46 .02
48 3-40 Labor Arbitrator .42 .02
64 .3-44 Office Clerk DBUSNS. -.67 .02

"* 68.3 -' Photoengraver* DBUSNS -.43 .03

80 3-48 Retailer . DBUSNS -.70 :001
90 3-50 Specialty Salesman* DBUSNS -.62 .006
99 3-53 Whalesaler DBUSNS' -.65 .001
109 Civics ' DCIVIC .41 .05
. .
_

1.
166 3-65 Musical Comedy . DBUSNS --.71 .03 ''''
180 "Popular Mechanics"
. --.43 .05 - sr.
190 Operating macCiinery 7-.50 .01
197 Interviewing prospects in selling -.40 .04
245 4-3 People who have made fortunes in business DBUSNS -.57 .004
, ,-,_}.
260 Side -show freaks .55 ..01.
306 J. P. Morgan, financier .. DCIVIC- -.37 : .04
338 Work in a large corporation with little change of 4
--._

becorhing president vs. work for self (B) DCIVIC .: .37 .03
373 2-3 Am always on time with.my'work -2.46' '.02..

Between the .05 and .10 level:

25 Civil Service. Employee -.47 .06
54 Life Insurance Salesman -.61 .06
73 Printer i -.35 .09
74 3-47 Private Secrefary
. DBUSNS -.35 .08'
95., Traveling Salesman `-.40 .10

1100. Worker in YMCA, KOFC, etc'. -.38 ...07

1.16 Histoty .38 .09
135. Typewriting -.36 .07
156 - Smokers .33 .09
257 Nervous people -.31 .10

279 4-6 People wbo -chew' gum \ :42 .1Q

379 Have good judgment in appraising values, .41 s .08
397 Tell jokes well vs. never tell jokes (A) -.36 .08
Original 23 A-B items.
(A) The A' prefer the second choice.
, (B) The B's prer the second choice.
 'HOSPITALIZE)) SCHIZOPHRENICS, DEPRESSIVES,-AND NEUROTICS -89

Table 13.-SVIB Item; .Correlated With Therapists'Percent Improvement

in Patients Who Are Neurotics
(Doctors who did not proscribe drugs for sehlzophronics)
N so 32

item Plumber. p
SVIB PTO- Item Gamma

18 Bookkeeper -.54 .05

101 3-55 Algebra NSOLVE .01
113 Geography '.54 .04
115.3 -57- Geometry NSOLVE -.66 .02
119 '3.58 Literature NQEST1 .49 .05

148 3-61 Solving mechanical puzzles NSOLVE. -.47 .03 .

149 3-62 Performing sleight-of-hand tricks NSOLVE -.61 .01

171 Sporting pages .02
173 3-66 Detective stories NSOLVE -.51 .03"
;185 3-68 Making a radio set* NSOLVE -.44 .04

186 3-69 Repairing a clock NSOLVE -.46 .03

240 Optimists -.49 .03
4-5 People who talk very'slowly NQUEST2 .52 .04
269
303 Thomas A. Edison, inventor .68 .01
310 John Wanamaker, merchant .
-.50. .03

Chairman, Educational Committee .47 .04

316
350 Playing baseball vs. watching baseball (A) -.51 .03
NQEST1 .03
359 3 -23 Jealous vs. conceited people (A)
. 390 Usually ignore other's feelings vs. consider (A) -.52 .02

Between the .05-and .10 level:

7 Athletic Director -.36 .09
11 Aufo Salesman -.38 .10
15 Bank Teller -.52 -.07
21 .3 -33 Cashier in a bank , NQEST1 -.46 .07
70 Playground Director -.30 .10,

Psychology .48 _ .08

129
130 Physiology .53 .06
142 Taking long walks .43 .06
Vaudeville .43 .06
165
198 Interviewing clients -.46 .10

217 Bargaining ("swapping") -.36 .09

236 Energetic people -.72 .09-
Create a new artistic effect, i.e:, improve beauty .36 .10
286
275 Bolshevists .42 .07
318 Chairman, Membership Committee -.42 .07

353 2-35 Reading a book vs. going .to movies (B)

358 Tall men vs. short men (A) -.52 .07
386 3:24 .Smooth out tangles and disagreements
between people -.62 .06
.399 Frequently make wagers vs. never (A) -.57 .08

Original 23 A-B Items.

(A) The A's prefer the .second choice.
(B) The B's prefer the second choice:
 Table 14, -The Criterion Variables-Pertent improved of Schizophrenic, of
Depressive, and
of Neurotic Patients. Intercorrelated.Viith the A'B Predictors and the A.B
Clusters
(Phipps CH* Thomplits Vlho Did Not hots Drugs for Any of tisk Sthirohronk Pstlents)1

PRIM Irrowykl
Al Pirko
Vorlii I
Table *obit WI 23 VII 22 .MPRV I MPRY 13' MIRY N TOTL I TOR D TOTL .10 ON $ D. N I OTHR 0 OTHR N OTHR

IN.13 22' 1 .634

MPRV S .458/

NAV D .482

MPRV N Zvi .267

TOIL S .803 .623 .705 .337 117

-r
TOTL .782 ri .326 '154 .335 .611

TOTL N 108 446 .374 .730 213 a

SD 10. /14 . 544 197 .822 le .440
41

DN 10 470 479 .161. .443 .548 ,354. , 070 172 .491

0.
$ 11

=
.546 513 .731 160

= 166
031 219 OM 487 .159

0 12 515 '555 18 171

mwm=
ne 136 .338 035 450 .08

N 13 178
.335 -,051 7302 120 .115 416 .968 .364 .587 .018 .252

S OTHR A33 432 .711 .234 -.04 069 .214 -.082 .419 011 .763 .161 -.109

OTHR 128 ,057 .554 .285 .129 ..323 348 793 .128 ,cti .315 .392 .040'

N OTHR .010 -.81 .1713 142 128 -044 .167 .532 .009 .422 -.120 065 L 052 193

*I 23 W.I 22 MPRV S , MPRV 0 MPRV N TOIL $ TOTL D TOIL N SD ON I 0 N S OTHR 0 OTHR N OTHR

SXPRSA 1 t' .28 -.028

.211 A76 -.012 .392 .077 -.017 041 .021 .565 .08 -.030 .506 -.132 -.082 '

S XPRSS

SOUEST
11

i1
, -
A01
=
.424

,
492 .185 .122 141

8
.254 .223

-08
-
.415; .182 .815 .079 112. -583 285 -.066

=/ .08
,534 003 .8 .245 .58 520 .033 r .078 85 -.062 -.016

DBUSNS i2 .831 Y 130 150 .870 240 AN .922 .442 103 136 .130 .903 .311 148 .171 .145

DCIVIC 12 .124 20 ,OY -606 229 256 .18 217 -,032 .184 i. .341 ,134'
,..192 172 -093
zr,
NSOLVE 13 257 A08 -.022 .28 .605 218 112
A84 A32 051 .001 .259 130 7119 AN .
=
,
.380

NOEST1 13 .247 214 .130 217 .485 .258 2.8 142 28 ii .185 .275 SU .181 .100 250

NOEST2 13 272 282 .020 .239 .530 '.235 ' 200 .551 114 .840 ' .18 .131 7449 .090 .097 .254 r: ri

5J The number of cases Is 35 except for MPRV 0 whin W 34 and MPRV N which is 32. The COMI1110111 111 Penonien f Single underlining Indictee that r is significant it .15; double
underlining, it 01. Many of the variables are not independent of othdr vviables.1,e., some Items of measurement are common tomore thin one variable. Among them debendendes are the
,'
following: TOILS w, SO + S. TOIL D w SD 4 DN + D. TOIL N DN + N. Variables contained In the triangle marked 3 are independent In the serge that Merles no ovorlippIng items,

'
 HOSPITALIZED SCHIZOPHRENICS, DEPRESSIVES: AND NEIIROTIeS 41

Clusters hive the potential of being scales, percent receiving insulin schock, or the per-
i.e., having a common meaning and represent- cent male.
ing a specific therapist characteristic. Predic- MPRV D. For each docto,, the percent of his
tors are probably multidimensional, containing depressive patients who improved; dePiessives
several clusters, and indicate what combina- Include both psychotiCt neurotic depres-
tions of therapist factors. are needed to deal sives in about equal ,preportions. This
with a particular diagnostic group.. criterion, not previouiti used f4 item analysis,
. Of particular importance in table 14 is is available for 34 of the 35 Odors.
triangle 1, showing the intercorrelations MPRV N. Similarly for each:of 32 doctors,
among the three improvement criteria. The the percent of his neurotic patients who im-
correlations are relatively low, only one or proved:
them approaching the .05 level of significance, TOTL S. (SD + S) All those items in table's
MPRV S and MPRV D at .834 (.336 needed at 10 and 11 which. are predictive of success with
.05). The relative independence of the three schizophrenics and which are included in, the
Criteria permits us to hypothesize that items PTQ.
related to each of them may stand for differen- TOTL D. (SD + DN + D). All those items in
tials in treatment required for the three tables 10 and 12 which are predictive of sue-
disorders. The absence of a negative correla-' cess with depressives and which are included
-tion between MPRV Sand MPRV N suggests in the PTQ.
that the functional reversal of the "A-B Scale" TOTL N. (DN + N). All those items in tables
will not be found in these data. If the functional 10 and 13 which are predictive of success with
reversal is valid, Awl the inpatient neurotics neurotics'and which are included in the PTQ.
at the Phipps Clinic must be different from the Triingle 2 shows the intercorrelations
outpatients at the V.A. among these "total- predictors " as they are
Considering now each variable in table 14, presented in, the PTQ.. As compared with
they can be defined as follows: Triangle 1, some independence has been lost.
23 and W-B 22. W-B 23 is the original Certain items significant for 35 doctors had*
23-item scale. In fact, this variable consists of been significant for the'47 and were not includ-
the original scores for the 35 doctors as com- ed in the PTQ. Scores from these items are
puted by Dr. Betz. In the W-B 22 one item (216, presented at the extreme right of table 14
"Entertaining Others") is omitted. The other under the headings: S OTHR, D OTHR, and N
22 items are included in the PTQ. The correla- OTHR. Note that the correlations' among these
tion of .964 between the two scales indicates three are very low. (Lower right-hand corner of
that they measure vractically the same thing. Triangle 3).
MPRV S. For each doctor, the percent of his Some of the items had been deliberately ex-
schizophrenic patients who improved. This cluded. Very few young people today have ever
criterion variable correlates .458 with W-I' 23, seen "Side-show freaks," and such an item is
' a respectable correlation but not high enough not now meaningful. These various items that
to assume that the W-B 23 is satisfactorY ,are not--included are further discussed below"
predictor for the 35 doctors. The W-B 23 items under S OTHR, D OTHR, and N OTHR.,
were derived- fro_ m a group of doctors that in- Triangle 3 presents the intercorrelations
cluded many of the 35 doctors analyzed here among eight predictors which are: (1') ex-
whb prescribed no drugs, but it also included haustive in the sense that all predictive items
some female doctors and some who prescribed Are in-dne or. another predictor, and (2) indepen-
drgs. dent in measurement in that no item appears in
Por the 35 doctors, the doctors' "percent of more than one predictor.
schizophrenic patients improved" is not SD. This consists of the first six items in
significantly correlated even at the :10 level, table 10. It is an important predictor for the
with any of the following variables: the percent following reasons: (1) It is the only one of the
of schizophrenic patients who are process predictors that is also a cluster; i.e., the itemi
schiZophrenics, the percent receiving ECT, the intercorrelate and can be "labeled." The
42 EXPLORING THE PSYCHO-SOCIAL THERAPIES

average intercorrelation (gamma) for 'these predictor and is the best single predictor for
items id .63. The "label" is the "rejection by. the success with depressives. We shall see that it
A's of manual and mechanical' pursuits." (2) is dominated by a single cluster.
This cluster correlates more highly with the N. All those items in table 13 which are in-
"Original W-B 23 than any other: .884. In fact, it cluded in the PTQ. This is the neurotic-specific
represents the main cluster from the 'original' predictor. It contains at least three clusters.
23-item scale. (3) While the original 23-item All of the above predictors, from TOTL-S to
predictor was developed for schizophrenics, N, are built from items included in the PTQand
this cluster is also predictive of success with are designed for the analysis. It is important to
depressives, in fact, the correlation with look also at those items not included in the
MPRV' D (.544) is slightly better 'than for PTQ to see what might be missing. These
MPRV S (.443). As such it might represent a items omitted. are contained in S OTHR, Et'
trait which is essential to the treatment of OTHR, and N OTHR.
psychoties. .
S 0 R. All those items in table 10 and 11,,
The limitations of SD are two: (1) There are signific t at .05 or better with the "percent of
other predictors and scales which are better schizo renics improved," and not included in
predictors of success with schizophrenics and the P Q. This predictor correlates .42 with SD
with depressions thin this one., (2) It is and . with S. Most of its items belong in one
multisemantic, i.e., the label -we have assigned place or the other. Specifically, by their cor-
to itrejecting manual pursuits 'has dif- relations, items 13 and 162 belong in SD. Items
ferent meanings for different population 150, 170, and 225 belong in S. This is not to siy
groups. that nothing was lost by omitting them; their
It will be noted that there is no SN predictor. inclusion would probably increase the reliabili-
There are no items which are predictive of suc- ty of the other scales substantially. It is to say
cess with both schizophrenics and neurotics. that there appears to be little new content in 'S
One item, "Adjusting difficUlties of others" ap- OTHR.
proacheS significance (.10 for MPRV S and .09 D OTHR. By contrast, D OTHR does seem to
for MPRV N) but the sign reverses. "Geog- contain new material. Of the items in tables 10
raphy" in talge 33 might also have been includ- and 12, significant 'at .05 'or better with the
ed here. It correlates: positively with MPRV N "percent of depressives improved," and omit-
(.54 at .04) but negatively with MPRV S ( .51 ted from the PTQ, there are only three that
at .09). Neither of these two items was included clearly belong in D. These items are 29, 42; and
in the PTQ. 197. It can be seen in table 14 that D OTHR cor-
DN. Four itemg near the bottom of table 10 relates only .395 with D. There clearly is
have some predtive power for MPRV D and something in D OTHR not covered by other.
MPRV N. Actually, three of them are not very predictors.
strongly related to MPRV D and the resulting Item 260, Side-show freaks, is of interest
Predictor is stronger for neurotics (.548) than because of a remark that Dr. Whitehorn made
for deptessives (.443). When we try to build on one occasion. When I asked him for his
clusters from the predictors, two Of the items clinical impressions he stated that a doctor
will go into a cluster for depressives, and two who was successful with depressives was rare-
into a cluster= for neurotics. This DN is not a ly on "odd ball." Rather he was conservative
.strong predictor; its.- .items might well have person. There was no, suggestion that he was
been distributed in D and N below. conservative politically or, socially, only in his
S. All the items in table 11 which are includ- person. "Side-show freaks" is, of course a badly
ed in the PTQ. These items migl_it be regarded dated kern. Perhaps in some future study of
as schizophrenicspecific. This is the strongest' success with depressives, someone will con -'
single predictor, of success with schizophrenics. struct a `new scale of personal conservatism.
It contains at least three clusters discussed Three other items are of interest. These doc-
beloW: tors who are successful with depressives say
D. All those items in table 12 that are includ- they like Civics (#109), dislike J. P. Morgan,
ed in the PTQ. This is the depressive-specific financier (#306), and they would prefer to iork
 43
HOSPITALIZED SCHIZOPHRENICS, DEPRESSIVES, AND NEUROTICS
r.
in a large 'corporition rather than work for oriented doctors shown in table 63 which is
themselves' (#338): "These items are discussed more casual.'
under DCIVIC below. ..SXPRSS. "Scale predicting success. with
N OTHR. Most of the items correlate with Schizophrenics the items' of which _appear to
other predictors or clusters. Specifically, items have something to do with social expression,
303. and ,310 correlate with NQEST1 below, not necessarily active." The fiftr items shown
item 240 with DN above, 171, 350, and 390 with in table 11 have an intercorrelation of .50. Item
NSOLVE' below. 367, accept just criticism without getting sore,
A-B Predictors and A-B Scales. All of the actually correlates is well with SXPRSA as it
above, with one exception, are A-B predic-. does with SXPRSS and could havle been' in-
tors, collections of items that are correlated cluded in either scale.
with a common criterion, but not necessarily SQUEST. "Cluster predicting success with
correlated with each other as one might expect schizopiirenics the meaning of which is ques-
if they constituted genuine scales. The one ex- tionable." The three items shown in, table 11
ception is SD. Here the average intercorrela- appear to have no common meaning although
tion of items is .63; moreover, these items haVe the average intercorrelation is .74. We look to
a common frame ofreference, the rejection of the PTQ to help with this scale. SQUEST does
'manual and .mechanical pursuits, as in the not correlate at all with SXPRSA and only
original W-B 23. We have seen, though, that moderately with SXPRSS. However, it can be
this rejection does not have a common meaning. seen in table 14 that it does correlate .504 with
across different groups. When SD is substi- SD.
Juted for WB-22 in table 9, the table is changed We turn now to the scales for depressives,
very little. where SD is also relevant.
It is our task now to try to extract from the DB.USNS. "Cluster predicting success with
predictors what we can in the way of clutters. depreisives the items Of which appear to relate
For present purposes we will accept a collec-' to disliking business pursuits." .The 11 items
tion of items as a cluster if the items are ap- are shown in tables 10 and 12. Their average in-
propriately intercorrelated even though we tercorrelation is .51. This cluster dominateS all
are unable to say at this time what the common the D predictors: - Although it correlates with
meaning of the items might be. It remains for -SD .703, it is not significantly related to sue-
the analysis of the PTQ to suggest possible cess with schizophrenics. This cluster is
meanings for some of these clusters. In addi-. multisemantic, like SD. We can look to the PTQ
tion to SD, there are eight clusters in the lower for ideas about its meaning. .

part. of table 14. Most of them are either DCIVIC. "Scale predicting success with
multisemantic or their meaning is entirely am- depressives the items of which may relate to
biguous. For only 'a few can we guess at the social concern." The doctors who are successful
'meaning. If we can be reasonably sure of the with depressives say they like Civics (#109),
meaning we will call it a scale. 1
dislike J. P. Morgan, financier (#306), and they
The intercorrelations of the eight clusters would prefer to work in a large corporation
are presented in table 15. Considering that the rather than work for themselves (#338). While
items ire derived from three criteria, thee the meaning Of such 'a scale is problematic, it
clusters are suprisingly independent of eaeh seems likely that psychiatric resident's, vin-
other. Tfiis is consistent with a hypothesis that tage 1945-1958, learned in civics that J. P.
a number of different traits are to be found Morgan was one of a group of monopolistic
among successful therapists. financiers Whose business philosophy was ex-
SXPRSA. "Scale predicting success with - presSed by Vanderbilt: The P9hlic be damned.
schizophrenics the items of which: appear to These three items could possibly be held
have something to do with active social 'expres- together by a philosophy .that "no man is an
Sion." The three items can be seen in table 11. island," a rejection of "rugged individualism."
The intercorrelation among these items in .51. Note that the doctors' rejection of P.,.Morgan
Social expression here is both active and in.- is not correlated with DBUSNS. This is an en-
volved in contrast with that of somatically tirely different dimension.
 44 EXPLORING 'hIE PSYCHO-SOCIAL THERAPIES

Table 15.7Intercorrelation of the A-B Clusters.

(35 Doctors Who Did Not Prescribe Drugs
for Any of Their Schizophrenic Patients)
(Pearsonlan r's)

VARIABLE SXPRSA SXPRSS. SQUEST DBUSNS ()CIVIC NSOLVE NQUEST1

SXPRSS .384

.'SQUEST . .050 .341

DBUSNS .149 .074 .062

DCIVIC .082 .021 .109

NSOLVE .120 .120. :036 .403 .093

NQUEST1 .177 .269 L,.129 .261 :161 .293

NQUEST2 .010 .397 .146 .115 .031 , .308 .246

For correlations with SD, see bottom part of table 14.

None of these DCIVIC items are included in NQEST1. The three items are shown in table
the PTQ. Nor does the PTQ contain a scale for 13. The average intercorrelation is only .44.
social concern. One item contained. in the PTQ, The best intercorrelation 'is between liking
"Labor Arbitrator" in table 12, is clearly a literature and not wanting to be a cashier (.75).
business-related occupation: However, far The meaning of this cluster,will be considered
from rejecting this item as they "do other in chapter V'(table 45):
business pursuits, doctors who are successful 1VQEST2; The average intercorrelation
with depressives show a pieference .for it. The among these three items is .69. The doctors
period 1945-1958 was a period of considerable successful with neurotics are -"indifferent"
labor-management. strife. It is therefore rather than "liking" foreigners and athletic
reasonable that this item should correlate with men. They like people who talk .slowly (see
DCIVIC, if indeed DCIVIC represents a broad tables'10 and 13). The possible meaning of such
social concern. In fact, "Labor Arbitrator" cor- a collection of items is certainly problematic.
relates as well with the three items in DCIVIC Accordingly, for further analyses in the
as-they do with each other. Unfortunately,.the PTQ, there are six predictors: TOTL S, TOTL
intercorrelation is not high, .40: D, TOTL N, for the total drug-free relation-
We turn now to the three clusters predictive ships with schizophrenic, depressive and
of success with inpatient neurotics: neurotic patients, and S, D, and N, represent-
NSOLVE. "Scale predictive of success with ing that portion of each relationship which is
inpatient neurotics and which appears to be specific to that diagnosis. SD, a predictor for
related to rejection of problem solving." The both .schizophrenic and depressive patients
seven 'items are 'presented in table 13.' Their happens also to have cluster properties. DN is
average intercorrelation is 46. 'If "sleight-of- a very weak predictor for both depressive and
hand tricks" were .left out, the average inter- neurotic patients. No items were found, that
correlation would be .50. predict success with both schizophrenic and
L,)
 HOSPITALIZED SCHIZOPHRENICS, DEPRESSIVES, AND NEUROTICS 45
,
neurotic, patients, tiuggestiiig a: considerable say about r3? Suppose r, were .60 and r, were
differentiation between these diagnoses. AO, what might be the value of r,?. Statistically,
Within the predictors, there are clusters, the and theoretically, it could be anything from
items of which are intercorrelated. Until we -1.00 through- zero to + 1.00. Some relevant
can assign a meaning to a cluster, it is not prop- data are presented in appendix 5.
erly a scale. Of 'the clusters, there are three The 'point is that the "correlates of. cor-
which have a probable meaning for personality relates" presented' here are hunches, useful
and might be called personality scales: SX- hunches, but of,lesser predictive validity than
PRSA, SXPRSS, AND NSOLVE. There. are the correlates of the,original criteria. We shall
' two-that can be labeled, but their personality see that certain obierved patterns. of correla-
implications are unclean_ SD and DItUSNS. tions are very difficult to explain on the basis
Finally, there are three that are complete of chance, quite aside from statistical tests of
enigmas: SQUEST, NQEST1, and. NQEST2. significance. But our intere'st is' in patterns
The 'PTQ correlates will provide some hints rather than in specific correlations.
about some df them. The PTQ correlates are shown in table 16.
Next we will consider these PTQ correitktes Two t f the PTQ scales are not shown' sines
of the drug-free relationships. After that we they have no significant corrglates with any of
will return to further analyses of the Phipps the AB predictors and scales (Fervor ai Life
data for' those doctors who 'prescribed drugs Satisfaction). It we count all the possible inter
for their schizophrenic patients. -.- correlations, excluding, WB 22, we find that
almost a quarter of them are significant at the
.05 level.
This table is computed-from data for the 46
nonhospital psychiatrists and residents.
PTQ Correlates for the Drug-Free Similar data for 43 hospital psychiatrists and
Relationships residents is presented in table 58, :appendix 7.
In that table less than a tenth of the correla--
:tions are significant. In the two tables, if we
How do these A-B predictors and ciusterg consider only tlie' five predictors that are
correlate with the more manifest personality measured independently, the percentages sig-'
scales that are included in the PTQ? The nificant are 27 percent and 8 percent respec-
research processes are presented schematical- tively. This is consistent with our assumption
ly in figure 1. The syinbols and arrows on the that' the drug-free predictors are more ap-
left side represent the analyses of the Phipps propriate for nonhospital therapists than for
, data where A-B predictors, clusters, and a few hospital therapists.
personality scales are derived directly from In table 16, the patterns suggest that some
the criterion. of success. Unfortunately, most of therapist traits are more general than others.
the predictors and clusters so .derived are not One or two seem to be correlated with all three
meaningful. We now seek their meaning by in- diagnostic areas, others only for the schizo-
, eluding them in another instrument which also iihrenias and depressions. Still others seem to
includes a number of manifest personality (MP) apply only to a single diagnostio group, and in
scales, and determining the" correlations the case of the schizophrenias only to certain
.represented by hoping. thus to be able to predictors for that diagnostic group. The cor-
make meaningful clinical interpretations.. We relates of S are very different from those of.
will call this "chaining correlations," or "cor- SD. Further interpretation of this table is con-
relates of correlates." sidered in Chapter VI, Clinical Interpretation.
It is obvious from the figure that the correla- Let us return now to the Phipps Clinic data
tions we' really seek' are those represented by for those .doctors who prescribed drugs for
r, on the diagonal. Given r, and r2, what can we their schizophrenic patients.

vat
.,,46 EXPLORING THE PSYCHO-SOCIAL THERAPIES

-Figure 1
SCHEMATIC DIAGRAM CRITERION VS. CORRELATIONAL STUDIES
Criterion
Subjects
at Time Period 1

C (Percent of Patients Improved)

Other. Subjects
at Time Period 2

Clinical Interpretation -

The Phipps DataPsychotherapy Plus treatment. To this issue we have alreadygiven

Drugs for Schizophrenics some attention. Appendix 8 provides support
for the studies of Hollingshead and Redlich and
We noted above that, if drugs change- the others, that, indeed, the two types of doctors
behavior of schizophrenics, then hey change are different.
the _therapeutic relationship, and resumably Of the 17 therapists at Phipps who finished
thereby the -therapist traits essential to suc- their rescdenciis in 1959 or later, there are 12
cess, Beyond the change in the' patient, drugs who prescribed drugs for half or more of their
may' imply a change in the theiapist, namely schizophrenic patients. We will assume that
that therapists who prefer. to use drugs may be these 12 are somatically oriented.-We now ask
differ6tit from those who !infer psycho-social the question: what SVIB items are correlated
.

4
6 4,r)
 w.

Table 16, Personality Correlates of the Al Predictors and Clusters, Drugfree Relationehilis

Pearsonian r'e N 46

Jot Populate

for Schinsphrenice

Personality Umiak

App.4 Tabli Number

W .1 22 TOTL S S SXPRSA PPM SOUEST SO TOTL 0 DIUSNS ON TOTL N N NSOLVE NOEST1 NOEIT2
Fa

-
-,427 -.401
.I
31 Need for ClOsure , -.266 -A16 -.267 .120 -.438 -.604 -.470 -.583 -.488 -.612 -.568
= :-.460
= -168 -139
.
,
MWMN
6.MI .111 ' OM.
.1.I NNW 0.1

-=
32 Regression In the Ly
Serrici.otthe Ego .410 :413 263 131 .147 117 .451 .482 335 .359 .166 .164,
. . ,192 , .041 268 7.043 .

0
33 EmPalhIc Interest 4 243 .3d
334 .332 .214 290 .046 .333 A34 .382 .227 .221 . 238 .193. 140 195 8
i
'34 VIA Ethlt , -,440 - 404
,

-152 -116
z
7,219 -.249 -.454
......
-154 r.-.089
......
-192 i -.183 -311: -.291 ). -10 -156 126
.

35 Need for Order -368 '7,406 -.338 -129 -164 -,206 a -167; 7.234 T.019 'i 144 -124 -117 -121 -144 -211 1,
= MO
, ,

A
36 Tolerance of the
Weal Experience .220
_
115 it9 0 MI5

'
.312 .041 .230 .330 308 , :373 '160 128 .104 ,141 -.061 153 . tfi
....
.

37 Tolerance of Z
Ambivalence '. .131 .291 ,302
ef
.479 .199 -.077 .206 .112 7.013 .027 -.021 , 187 202 194 .012 .015 ' "6
0

-
38 Preference lor
Decislon,Making .142 .295

1=
.386 .485 .445

., -
- 194 131 .285 14 .274' .112 .214 299 .185 .219 i .124 0
39 High Personal Ll
Stands* .285 .323 1.413 .148 387 201, .153, 079 -040 .029 .042
1
-!
.126 .128 . .05( .229 .017
U',I

Dl,,
,40 Ramors'e -.293 -.271 -.336 -.150 -.306 -.221 -.137 -.126 7.072 -185 -.036 -.184 -.195 -158 -.110 .473
1.
i

41 Antidemocratic , -185 -.146 -127 183 -.182 ,066 -.222 -.373 ,44,345 -13 -.295 -.152 ,:.122 ' -.125 128 -.059

42 Indlviduiliim , 42 .126 -.040 -.018, .4-175 138 154

==-
186 .391 312 .153 110 121 -001. 11 -,052;

43 Deference Anxiety .276 )71 110 190 -,067 064 243 A82 516 '.128 224 177 .171 019 ,236 -033 0
x
44 Nurlurince Anxiety .355 .216 100. .051 044 al .368 .341 ,188 ,. .242 ' 120 .401 .440 426 .080 -.014 til

Science end
.... =.. =
Moral Values -228 -016 141 -024 203
.4.
- 073
r
-169 -.120 '171 -.117 -266
_
-125 -281 -.382 -065 -100
48 Preference for Intel.

0
rectual Challenge
, ,
, -.151 7.030 .649 .190 -.117 7164 -.098 -.051 050' , -.074 -.086 -.229 -.251
.-
-194 .235 -200
CO

41 Extroversion .209 .233 317 .123 , 100 .329. 189 .214 .238, -.028 -.102 136 193 -145 p? -.015

50 work Activity - - 279 -184 7031 -.016 -,304 - 300 -.163 064 ' 110 -.085 -,163 - 158 -063 -.251 -116

Single underlining,theans Ognilicantly dillerenI Iron zero at .05, hvo.tail, Double underlining, 01

Variables In the heading are described on pages 35 to 44.

 EXPLORING THE PSYCHO-SOCIAL THERAPIES
Table 17.---SVIB Items Correlated With Doctor's Perceint Improvement In
Schizophrenic PatientS for Whom They PresetWed Drugs'
(12 mate doctors who ended their residencies in 1959 or later and who
prescribed drugs for half or more of their schizophrenic patient's)
Item humber,
SVIB PTO Item Gamma P
12 Auto' racer .68 .03
24 Civil engineer -.74_ .03
44 3-38 Interpreter ABDRUG .78 .03
48 3-40 Labot Arbitrator ABDRUG .77 .04
104 Art -.72 -.05

144 Chess .69 .04

195 Arguments .02
278 Men who use perfume .87: .02
306 J..P. Morgan, financier .76* .02
327. Repair auto vs. drive auto (A) -.92 .04

328. Develop plans vs. execute. plans (A) -.94 .01

342 Outside, work vs. inside work (A) -.77 .01
352 3-28 Nights at -home vs. nights away (B) ABDRUG .65' .05
353 2-35 Reading a book vs. movies (B) ABDRUG .77 .02

Between the .05 and the .10 level:

4 Army officer .65 .08
20 Cartoonist -.78 .07
125. Nature study G." --.70 .10
148 3-61 Solving mechanical puzzles -.54 .10
180 "Popular Mechanics- -.55 .09

Operating machinery --.57 .09

20 Teaching children .68 .09
206 Meetingand directing peciple -.71 .08
2.16 Entertaining others** -.70 .10
261 People with gold teeth .64 .06

302 -.. Enrico. Caruso, singer t. -7.56 .09

334 3-26 Taking a chance vs...playing safe (B) ABDRUG .51, .06
340 3-10 !Small pay,` large opportunities vs..
good pay, little oppKtunity (B) ABDRUG .96 .07
357 Fat men vs: thin men (B) .71_ .06
358 Tall. men.vs. short in en (A) -.66. .08
368 1-22 Have mechanical ingenuity (inventiveness .55 ..09
384 Show firmness without being easy .65 .09
393 Usually make excuses vs. never (A) -.96 .07

The sign of this item is reversed from that found in previous analyses.
Original 23.items.
(A) The A's prefer the second choice_
. (B) -The Et's prefer the second choice.
 HOSPITALIZED KIIZOPHRENICS, DEPRESSIVES, AND NEUROTICS :

with their success with those pstien ts for cern of Whitehorn and Betz that nearly all such
whom drugs'u,lere prescribed? The percent or patients improve' is valid. Almost ail doctors
patients improved varies from 67 percent, to .4 have a 100 percent improvement rate with
100 percent, .with all .average of ,80. percent. their neurotic and depressive patients. In
Table 17 presents the items that are order to study their 'success, we need another
discriminated by this criterion. eriterionbof success. (Such a criterion is un-
There are at leasVour empirical clusters in doubtedly available in.)the Phipps Clinic data.
table 17, some of which do not seem to make, See` page 16.)
Much sense. The empirical 'clusters are:.(1) 44, Considering only the schizoplirenicatients,
340, 353; (2) 12, 104, 278, 328; (3) 24,144, 261; (4) what: are the interrelationships among the
48, 195, 342. criteria and the predictors in the drug-free and
Jilf Rierdan has studied these clusters. She the drug-related therapeutic relationships?
notes that the first 'cluster interpreter, op- Table 18 attempts to relate the various criteria
portunity for advancement, and reading vs. and predittors. A new variable is introduced:
movies all' involve activity. She also observes - among the somatic therapists,, the perent im-
that the items in. the fourth cluster labor ar- proved q schizophrenic patients for Morn no
bitrator, arguMents, and inside,work involve drugs were preicribed. This is the therapists'
reasoning. She suggests ihat the. A. therapist is success in the drug-free relationship to be com-
'Much smore concerned with reasoning than pared with their success :in the drug-associated
With solving (personal communication). The in- relationship. In order to compute this figure; it
sight would be consistent with the inference was necessary to relax 'one of Dr. Stephens'
from table 16 that the A therapist has a low standards, for "percent improved." For all
need for closure. other percents, there must have been at least
Only six of the items in table 17 significant at four patients; this one is computed if there are
.07 or less were included in the 13,TQ. However, at least two patients. Two of the doctors do not
if we compute a score for these six items and have even two drug-free patients 'so we have
correlate it with a- score computed from all this sseasure for only. 10 of the 12 doctors.
items significant at .05,' theresulting r is .93. In Note that these 10 doctors' sucdess
other words', the /six items in the PTQ very their no-drug .patients does not correlate 1.1
closely reproduce the total score, and we can with their success with their drug patients.
expect that the PTQ will give us some idea Note that ABDRUG does reasonably repro-
what table 17. represents. These six items are duce the percent-improved criterion from
labeled ABDRUG and this predictor for the which it was drawn (.84) but does not correlate
drug-associated relationship is roughly Com- strongly with the ne0drug percent improved
parable to TOTL S in the drug-free relation- (.39). The drug predictor does not work for
their no-drug patients. Do the no-drug predic- ',.
It should be noted that while one of the six tors -work for these patients?
ABDRUG items also predicted success with If we compute fore these drug- prescri
schizophrenics in the drug-free relationship, it therapists their scores on the A-B predictors
is now reversed in sign. SucceSsful therapists derived fro the analyses of the 35 doctors in
in the drug-free relationship preferred "nights the drug free: relationship, we Can test' how
away!' to "nights. at. home." Successful those predictors work for both .the drug-
therapists in the drug-associated relationship' associated relationship and the' drukofree rela-
prefer . "nights at honie." We cannot help tionship among drug-prescribing therapists. In
Wondering whether "nights at home" means the first column of table 18 it can be seen that
the>same thing to the two groups. Could one, of some of the no-drug predictors also predict suc-
these groups look to nights at home as nights cess for the no-drug patients of these- drug-
alone, while the, other is thinking of entertain- prescribing therapists. This replication is of
ing people at home? some interest. Note that SXPRSA is a
We have no datajor the therapists' use of anifest scale; its meaning is reasonably
druga with their clepreSsiy'e and their neurotic evident.
patients. Moreover, during this period, the con- . .. ..
On the other hand, WB 23 and SD are
_ ..
 7.

EXPL51ZINO THE PSYCHO-SOCIAL

. y.

Table 18.--interrelaition of Criteria and 'A-B Predictors

for Therapists Who Prescribed Drugs*, ,../
(Peirsonlan es)

% Of Schizophrenic
Patients improved
No Drugs Drugs

% of schizophrenics Improved:
Patients-with no drugs 1.00
- Patients on drugs . 13 1.00
ABDRUG (Drug-Associated
)
Predictor)
-
. 39

The Drug:Free Predicto(s:.

TOTL S .54 .42
os. SD .13 .31
S .76 .39
SXPRSA .62 .30
SXPRSS .23 .03
SQUEST .47. .19
S OTHR .35 -.32
6
1/11:B 23 .44 ,24
*12 male therapists who ended their residencies in 1959 or later and who prescribed drugs for half or more of their
schizophrenic patients, except that.2 praicribed drugs for all or nearly all cif 'their schizophrenic
,
patients, leaving only .1
10 therapists for the no-drug criterion.:,
Single underlining indicates r is significant beyond the .05, double underlining beyond the .01 level (twolail test.}
. .. , .

ti

rnultisemantid; they change their meanings.as prescribed, even though the therapists. are the
we pass from hospital to nonhospital psychia- same (table 18). Till's the data suggegts that
trists. In table r4 we found that, while WB-23 is some therapeutic requirements, of schizo-
dominiated by a subscale,. SD, neither of these phrenics may remain constant when drug's are
Correlate well 'with SXPRSA. The failure of preScribed while otheri do not.. This' is dis--
WB-23 and SD in table48 could be due to their cuseed,further on pages 51 and 161.
multiSemantic nature, i.e., the therapists For now it is sufficient to conclude tflat (a)
prescribing drugs are different: the three-item predictor, SXPRSA, is repli-
In the next section we shall see that some cated in the no-drug `§ituatkn, and (b) WB-23
correlates of WB 23 hnd o SD are similar to and SD fail to predict in the'dAtg situation. The
some correlates of the dlug-associated predic- latter two appear not to be relevant for
tor, ABDRUG, thus giving somi support to the somatically oriented therapists, or in settings
idea that it is the multisemantN nature of. the :- where, rugs are a major aspect of therapy.
"A-B Scale" lOicti destroys its predictive These findings explain some anomalies in
power when dr ags are prescribed. But some of the literature. Stephens' and Astrup (1965) us-
the correlates are different. Moreover, ing Phipps Clinic data, found no relation be
SXPRSA, a manifest.scale, falls short also of a tweentherapists"A-B Score" and patient Out-
significant relationship when drugs are come. Most of the patients. in\ their..study
 HOSPITALIZED SCHIZOPHRENICS, DEPRESSIVES, AND NEUROTICS 51

received. drugs. Similarly, studies in settings We have already noted that there is, no
where drugs are used extensively have tended assurance that "chaining" correlations will,
to show null findings er reversals (e.g., Draper, give dependable conclusions. Such findings are
1967; May, 1968; Bowden et al., 1972). much in need of replicatiog. But the similar
coefficients in table 19 constitute a replication
of themselves.
PTQ Correlates When Drugs Are One might ask whether the similarity of col-
Prescribed for Schizophrenics umns 1 and 4 simply refleasthe fact that some
of Ord original Phipps residents from whom the
The predictor, ABDRUG, consists of the six predictors are derived are included in the pres-
items included in the PTQ which are correlated ent data: It is important to understand that
with the therapists' Success with schizo- Phipps graduates included in table 19 are
phrenics when drugs are prescribed. Since classified according to their present settings,
these six items together correlate .93 with the not according to their original drug-prescrib-
score of all such predictive items, it may be ing behaVior. After all, most of the 35 doctors
useful to study the PTQ correlates of who did not prescribe drugs at Phipps did not
ABDRUG. These correlates are shown in the have the option; there, were no drugs. Further,
first two columns of table 19. The first column it will be recalled that among the Phipps doc-
is for hospital psychiatrists and residents, tors who prescribed drugs- for half or more of
presumably those who would be prescribing their schizophrenic patients, only a few re-
drugs. The second column is for the non- turned the PTQ. The distribution of PTQ re-
hospital psychiatrists and residents. As be- spondents shown in table 20 reveals, that the
tween these two columns there are more Phipps therapists are' not distributed in table
significant correlations for the presumably ap- 19 according to their 4 classification in the -
propriate group than for the presumably inap- original analyses, and therefore cannot account
propriate group. for the similarity of columns 1 and '11 in table 19.
The last two columns show the correlates of Assuming that the similarity of columns 1'1'.
the drug-free predictor, TOTL S. again for and 4 is not an artifact, sand I can find no
hospital and nonhospita.I psychiatrists and evidence that it is, the findings suggest that;
residents. And again the number of significant regardless of whether drugs are prescribed,
correlations tends to confirm the appro- therapists who ate effective With schizo-
priateness of the group for the TOTL S predic- phrenics have a low need for closure (tolerance
tor. (Column 3 is from table 58 in appendix -7. of complexity), are tolerant of ambivalence,
Column 4 is from table 16 above.) aria-of regression. All Of this suggests that they
. Table 19 was sufficiently stunning that I aretolerant of schizophrenic behavior. In addi-
found it diffiCuit to belieire. I have therefore tion they are interested in other people's feel-
checked it over a number of times. Here we ings and they Prefer to make their own deci-
have two multisemantic predictors, with only sions. These. findings are consistent ,with the %.

One SVIB item in common, and that item *Hier findings of Whitehorn and Betz that the
reversed in sign. When these predictors are ap- A therapist actively participates with and tries
plied to presumably appropriate, but different to understand his patient rather than concen -.
groups of therapists, they generate some com- trating on the reduction of symptoms (see
mon correlates (column 1 compared with
.
page 16 above):
column 4).

ti
 EXPLORING THE PSYCHO-SOCIAL THERAPIES

Table 18.;-- Personality Correlates. of Two A-B Preditors for Schizophrenics-

:Among Nonhospital and Hospital Psychiatrists
(Peiirsonlan r's)

ABDRUG TOTL
(Drug-associated) (Drug-tree)

PTO Personality Scale and Non- Non.

Appendix 4 Tabte Number Hospital hosibtal Hospital hospital

31 Need fort Closure -.49 -.18 -.07 -.41

-_-,...

'32 Regression in the Serviceof the Ego .33 .17 . -.10 .41
...-_,_.-..

33 Empathic Interest .43 .19 -.18 .33

34 Work Ethic q
.18 -.03 -.05 -.41
_-=
35 Need for Order .11 -.19 .02 . =.41

36 TOlerance of the Unrealistic Experience .14 0 -.24 .32

37 Tolerance of Ambivalence .32 -.10 -.08 .29

38 Preference for Decision Making .46 .26 .17 .30

- 3g-ligh Standards -.06 .07 .20 .32

40 Remorse -.05 -.11 .02 -.27

41 Antidemocratic -.06 -.17 .11 -.15
42 individualisin .29 '.31 -.14 .13

43 Deference Anxiety . s .24 .14 .09 .17.

44 Nurturance Al-ixiety .05 -.02 .04 .22

45 Science .04 .18 .20 -.08

46 Preference for inteictual Challenge .33 .30 0 -.03
47 Extroversion .22 .28 .35 .23

Number of cases 106".°

43 '.,-, . 46 43 46

Single underlining indicates that the correlation is significantly different from zero at the .05 level, two-tail; double
underlining, .01 level.
 HOSPITALIZED SCHIZOPHRENICS, DEPRESSIVES, AND NEUROTICS

Table 20.--PTO Respondents, Hospital and Nonhospital

43 .446
Group Hospital Nonhospltal

Former Phipps Clinic Residents:

Of the 35 who prescribed no drugs for
any of their schizophrenic patients
`(criterion group for TOTL S) 10 13

Of the 12 who completed their residencies

in 1959 or later and who prescribed
for half or more of their schizophrenic
patients (criterion group for ABDRUG) 2 1

Other psychiatrists and psychiatric residents 31 32.

Chapter V

The Third StudyJuvenile Delinquents

James K. Dent and George A. Furse

The opportunity to do research in the the children of Civil War casualties. In 1964 the
Loysville (Pa.) Youth Development Center property was acquired. by the Commonwealth
opened new vistas, including the possibility of of Pennsylvania and the first youths were ad-
a self-contained study yielding more direct in- mitted in that year. There are six cottages, the
terpretation than is possible for "correlates-of- youths being assigned by age, except that in
correlates" analyses. Although this 'study in- recent years one .of the cottages has been a
volved processing a large amount of data, it diagnostic center. Here youths of all ages may
should be regarded as a pilot study because of Stay up to 60 days, most of them being return-
the experimental nature of the methods ed to, the courts for further diaposition.
employed. Each cottage has a supervisor and six staff
The study had the following general pur- members, of which three or are counselors
poses: and the iemainder, house parents. The "organ-
izational level" of the 51 "helpers" in this study
(1) Developing methods of measuring thera- ranges as follows:
peutic influence in a milieu setting.
(2) Exploring characteristics of therapists House Parent 1 (2)
who are effective with the personality House Parent 2 (3)
disorders, thus extending the range of House Parent 3 (S)
Counselor Trainee (2)
disorders investigated. Counselor I (22)
(3) Replicating findings for institutionalized Counselor II (10)
neurotics. Director (4)
Among juvenile delinquents there is usually a Only those *directors who have significant con-
sizeable minority labeled neurotic. While it tact with youths are included. Most of the
was recognized that these neurotics may well helpers had no college education. However,
be different from the Phipps Clinic neurotics, most of them had had years of experience with
still the possiblity of replication was carefully delinquent boys.
considered and planned. At the same time it Youths, ages 12 to 17 inclusive, are assigned
was recognized that methods would have to be to the Center at the discretion of the county
very different in a setting like Loysville. courts, and the various judges do not, use
Therefore, the primary objective was develop- uniform criteria for this purpose. There has
ing methods for a milieu setting. Secondarily, always been a long waiting list. The
we hoped to extend this research area to the characteristics of the Center' determine in
personality disorders. Finally, replication of part, its clientele. There is no fence. ,The
earlier findings for neurotics was hoped for, Center is located in a beautiful countryside, as
but hardly expected. beautiful as can be found anywhere. It has a
modern swimming pool.. And the staff work
The Loysville Youth very hard to make the program attractive as
Development Center well as rehabilitative.
The Loysville Y.D.C. is located in what had Director John R. Williams chuckles when he
originally lieen established as an orphanage for says: "I ask them, if they are planning to leave,
54
 JUVENILE DELINQUENTS
55

that they please stop by my Me ov the way It has been generally recognized for decades
out and let me knowthey ar that "juvenile delinquency" is a rubric or
The clientele, therefore, include a disproipor- chapter heading .rather than a classification;
tionately greater number of mild cases than Ind that within the rubric are a wide variety of
would be found in some institutions for delin- /problems', maladaptations; and/or disorders.
quents. Still, the possibility of .being sent to There is now a large amount. of excellent em-
another institution, not so attractive, must pirical work attempting to, subdivide "delin-
serve to restrain,-since there are a number of quency." Most of it involves the empirical
severely 'disturbed and disordered youths. The clustering of specific problem behaviors. Some
following is from a report of a survey of county studies, also include in the clusters certain fac-
Probation officers conducted by Loysville: tors that are considered to have significance as
causes of delinquency (e.g., loss .of a parent).
In general, the boy committed to Loysville is Some of the work also rests on theoretical con-
better labeled dependent-neglected rather than ,sideration, for example, the "maturityleVels"
delinquent. Overall, the students' problems are .

centered in the home situation which accom- Of Marguerite Warren (1966). Other significant
panyifig school problems and the boy is seen by analyses are those of Jenkins and Hewitt
probation officers as having the potential to be (1944), Jenkins (1964), and Quay and, Parsons
helped. Instead of severe delinquency ther6 are (1971).
emotional problems. For the most part, the These various studies do not yield identical
youngster tends to be younger than those
placed in other institutions and, many times, is diagnostic groups. The 'number of groups
in need of long-term service. When asked for the ranges from three to nine. A study 'by,
general characteristics of those boys whose Kobay'ashi, Mizushima, and Shinohara (1967) it ,

placement at Loysville was felt to be successful, lustrates the hierarchial and overlapping
the above, with minor alterations was brought nature of the empirical clusters of problem
out. Loysville was felt to have its most success
with the less sophisticated, less agressive delin- behaviors. But in these various studies the
quent, having minor emotional problems, and following groups tend to be found again and
possessing. a relative degree of immaturity. The again:
youngster is pliable, passive, suggestible and
has the potential to develop relationships. On \(1) the situational delinquent (not, found in
the other hand, boys whose placeinent was felt some institutions that deal only with
to be unsuccessful could, be placed at the other severe" problems)
end of the scale. This youngster was seen as (2) the cultural identifier, socialized to a
being an aggressive, sophisticated, hard-core
delinquent with a poor attitude. Character dis- subculture
orders, security risks and boys with prior insti- (3) the neurotic
tutional experience were also seen as doing (4) the conformist
Poorly in the program at Loysville. (Erikson, ,(5) the manipulator or sociopath
mimeo) (6) the asocial
It is' quite likely, of course, that' the .hard-core For the present research we chose to use such
delinquent is seen as doing poorly in the pro- a schema, one promulgated by the Com-
grams of all kinds of institutions. Nevertheless, monwealth of Pennsylvania Juvenile Court
it is clear that, while there is a range of Judges' Commission (see appendix 10).
severities at Loysville, the problems are not so Not only has there been cogsiilerable work
severe on the average as at other institutions. diagnosis, there has also been considerable
attention to differential treatment and the
matching of youths and "helpers." Much of this
Study Design work appears tifto--. depend upon professional
judgment (e.g., Gerard, 1969; Palmer, 1967).
It is not possible to review here the immense One study compares outcomes for different
amount of research that has been done on the types of offenders in different types of settings
diagnosis and treatment of juvenile delinquen- (Warren, 1969). There are two studies where
cy. Still, some description of the state of the art youths were matched with helpers on a priori
is necessary as a setting for our own work. grounds and compared with youths randomly
 56- EXPLORING THE PSYCHO-SOCIAL THERAPIES

assigned (Palmer, 1973; Ingram, 1970). In both the severity of the, disoider in addition to
cases, the matched group showed a better out- diagnosis and improvement!
come. We could find no studies aimed St find- - To the problem of assigning youthi to
ing what characteristics of helpers are desir- helpers, a number of solutions were con-
able in the treatment of particular kinds of of- sidefed. For examplerserious consideration
fenders. The present study, then, modeled as, it was given to asking the youths themselves
is on the Whitehorn-Betz approach, appears to who' they felt had helped; them. It would re-
be breaking new ground in delinquency. quire considerable resources to, find these
In summary, our study design required a youths who came from all over the_ Common-
diagnosis for each youth, a measure of im- wealth and had already left Loysville. Because
provement, and the personality characteristics our resources werelimited, we chose to ask the
of his helper(s). The phrase, "of helper(s)," helpers whom they had worked with, i.e., to
conceals an enormous problem. In a setting assign youths according to helpers' reports of
where a group of clients is exposed to a zroup contact. Since helpers had some understanding
of helpers, how does one know who helped of the nature of the study, one might wonder if
whom? We,- tuKn now to the problem of they, would "claim" those youths who showed
therapeutic influence in a milieu setting. the greatest improvement. In fact, there is a
tendency for the opposite to occur. The less the
youth improved, the more the helpers reported
having a lot, of- personal contact with him.
It is in this area of assigning youths to,
The Problem of Therapeutic Influence helpers that this .present study is most ap-
in a Milieu Setting propriately viewed as a'pilot study. While our
methods produce findings which have face
validity; only in replication can their opera-
In any setting where two or more helpers tional validity be established. In the present
mingle with two or more clients there is a prob- findings there is replication, of the Phipps find-
' lem in assessing which personality affected ings for neurotics.
which outcome.-The problem is more common
than it appears to be at. first glance. It applies
not only to milieu-therapy, but to any therapy 'In the present study, the greater the severity of a pa-
involving more than one therapist. Moreover, tient's problem the less the improvement. It is not uncom-
even when the client_ is assigned to a particular mon to ..find greater. severity associated with greater im-
therapist, if there is a milieu, it is inap- proveMent. The reasons for. the varied findings are not
known; but it may be useful to speculate. Intuitively, when
propriate to assume that the assigned we view the range of human piloblems, the more severe
therapist is the therapist unless the assigned ones are' more intractable. Why then would more improve-
therapist actually does spend a great deal of ment be found among more severe cases? If we are dealing
time with the client (as at the Phipps Clinic). with a particular kind of problem (e.g., schizophrenia) and.
The processes in a milieu that determine our measure of severity is taken at a' nonrandom point in
the episode, the improvement might well be positively.
who interacts with whom are not well related to severity of symptoms. If our measure of severity
understood', but it is a reasonable assumption is taken at a time when the person presents himself, or is
that these processes are not random. This im- presented for treatment, then it is reasonable to assume
poses a further burden on understanding how that it was taken 4 a 'high point in an episode, on the
much help each client got from whom. As we average at least, and that some improvement from that
high point is likely simply by virtue of the episodic nature of
shall see, at Loysville there is a small but such disorders. If, however, we measure severity 4:ad-
significant tendency for those at higher levels ministrative" convenience (at the transfer betviten
in the organization to take on the tougher hospitals, or between jail and a prison) then we miglit:' ex-
cases. There is also a tendency for the tougher pect the measure to 'be less related to the: episodic
cases to show less improvement. Thus the phenomena. Such a measure would more likely reflect the
intuitive expectation that more severe problems show less
assignment processes are systematically biased. improvement. Admission to Loysville, 'is more 'ad-
We anticipated such confounding and decided ministrative tluin episodic and we:Tound that the greater
that we needed to know something about the severity the less the improvement.
 57
JUVENILE DELINQUENTS

tional problems as "Moderate,"

Youth Measure:. "Serious." :(The forms for the:judges included
A list of names was compiled of youths ad- only two cOluinns::"Moderate" and "Serious. ")
mitted 'to Loysville Y.D.C. between March 3, Later, judges and helpers were given a ltist
.1967, and March 26, 1974, who remained at the of those youths with whom he.had had a lot of
Center at least a month, and who had been personal contact and asked the following ques-
discharged the time the study began in May tion: When I saw this boy at digibarge I con-
1974. This list of 593 names was in order of ad- sidered his attitude, behayior And/or emotional
mission and indicated the date admitted and response as "Not ImProved," "SomeWhat
the date closed. The list was presented to each 'proved," "Highly Improved." They were also
of five "independent judges" (teachers in the asked to.make a diagnosis-.---.-.
Center School) and to each of 51 "helpers," The diagnostic system in the report of the
defined below. Judges and helpers were a-sked Juvenile Court Judges' CoMmission had only
to indicate for each youth whether 'their per- recently been released. Part of this report. is
sonal contact with this youth had been "very extracted in appendix '10. The helpers at
little," "some," or "a lot." They were asked to Loysville were not acquainted with the report
indicate for those youths with whom they had or the diagnostic system it contained. The
had some contact their answer to the following report was considered too bulky for their Use.
question: When we first became acquainted I Instead they: were furniihed the following
regarded this boy's behavioral and/or emo- diagnostic descriptions:

Neurotic This individual's offense was against property or person. His prior delin-
Delinquent: quent behavior was of the same nature. He constituted a possible threat
to self and/or community because of his own anxiety. His family structure
appeared as if the parents were childlike and he himself was expected to
be adult. His attitude-to his then current situation was that he denied the
self image of being a delinquent. He was anxious and confused. In regard
of school adjustment unusual behaviors were noted in the records. Re-
garding his peer group relationships he was probably a loner who may
have selected a few friends.
2. Situational. This individual's offense was- probably a delinquent act, an accident, or
Delinquent caused by the situation. He probably had no prior delinquent, behavior.
He' constituted no threat to self ,or community. His family structure was
probably acceptable. His attitude toward his then current situation was
realistic and regretful. His emotional adjustment and control were
probably sound. His school adjustment probably good, and his peer
group relationships probably acceptable.
3. Subcultural This individual's offense probably involved property not necessarily for
Identifier: Personal gain. His prior delinquent behaviors had all been' accepted by
his peer group. He may have constituted a'threat to others. As regards
family structure, his parents were probably suspicious and distrustful.
His attitude toward his then current' situation was that he believed delin-
quent behavior to be acceptable. There were no obvious problems as re-
gards emotional control and adjustment. In school adjustment, he was
identified as acting out and associating with "bad" youngsters. His-peer
group relationships were intense.
4. Antisocial This individual's offense involved a property offense he thought he could
Manipulator: get away with. His prior delinquent behavior involved property offenses
involvingTersonal gains. He-was a leader who could be a threat to others.
his parents were inconsistent in providing love and rejection. His atti-
 EXPLORING THE PSYCHO - SOCIAL THERAPIES . .

tude toward his then current situation was that he justified his behavior
as warranted and reasonable. As regards emotional adjustmea and con-
trol, he displayed emotional isolation. He had a school history of using,
classmates and being the instigator. In his 'peer group relationships, he
used others and saw them as suckers..
Conformist This individual's offense was a personal or property offense in hich the
client was a follower involved with others. His priordelinquent ehavior
involved being with others and being identified as a follower. H consti-
tuted a possible threat throUgh peer group pressure to himself or the
community. He had experienced weak parental discipline or indifference.
His attitude toward his then current situation was that he admitted re-
sponsibility for his behavior. His behavior was determined by a need for
self approval. In school adjustment, he had a history of attention-seek-
ing behavior but performed well when observed. As regards peer group
relationships, there were no close relationships but attempts were made
to establish them.
6. Asocial: This individual's offense was against person or property and probably
committed alone. His prior delinquent behavior was constituted of impul-
sive hostile acts which may have been violent. His explosive, behavior
made him a threat to self and/or community. He' had rejecting parents
With possible physical cruelty. His attitude toward his then current situa-
tion was that he denied responsibility and projeeted blame onto the
world. As regards emotional adjustment and control there was no im-
pulse control and there were infantile demands. In his school adjustments
he had a history of acting out of uncontrollable Page. In' his peer group
relationships, he was a loner and peers saw him as strange.

In table 21 it can be seen that for 289 youths, "moderate," and "serious" for severity, and to
there is no agreed diagnosis. In some of these "not improved," "somewhat improved," and
cases kthere are undoubtedly bona fide "highly improved.") For improvement, all
"diagnostic problems," e.g., youths 'with multi- ratings for a youth are averaged. For severity,
ple problems, or with problems not adequately the rating -format for judges and helpers was
covered by the diagnostic schema. It is likely, not identical (as noted on page 57) so we have
however, that if helpers could be trained in the two severity averages: helpers' and judges'.
schema, the number of disagreements could be
reduced.
'In fact, helpers did not rate all youths with whom they
"Agreed diagnosis" generally meant majori- hada lot of personal contact. To cut dovin the amount of
ty agreement among the diagnoses available rating, diegnpsis, etc. theY were asked to rate only asample
for each youth. HOwever, there is a tendency to of such boys. The result is that . the "contamination of
use "situational" as a residual diagnosis. Many criterion" is less thin would be eXpected. On the average, a
of the Loysville staff are conscious of the pe- helper's rating of if bciy he claimed is only about one-seventh
of, the youth's ratings. This contamination could be removed
jorative nature of diagnoses. When in doubt, but only at considerable cost. It ildoubtful that theresults
they chose the mildest label. Accordingly, we would be materially affected.
required that there be complete agreement for. In addition to the. improvement criterion, we had
the "situational" diagnosis to be applied. 'measures -of recidivism, a commonly, used objective
In addition to the diagnoses, the average vailable. What is needed is Aot recidivism but change in
youth has improvement and severity ratings recidivism, a much more difficult measurement-problem.
Presumably, a youth- who was originally apprehended for
from judges and improvement and severity violence, and who now is picked up for petty stealing, has
ratings from helpers.' (The numbers "1", "2", improved. We were not able to get satisfactory data foe
and "3" are assigned to the categories, "mild," change in 'recidivism.
 59
JUVENILE DELINQUENTS ".

Table 21. Diagnosis, Improvement, and Severity of

Youths Included In Analyses
No. of Average
Helpers Improvement Average
"Cipiming" Scores Severity
No. of These (Helpers Scores
Diagnosis Youths Youths and Judges) (Helpers)

2. Situational 88 , 37. 2.35 1.61'

3. Subcultural 46 41 1.92 2.10

5. Conformist 14 29 t93 2.21

4. Manipulator 42 45 1.72 2.20.

1. Neurotic 27 35 1.65 2.65

6. Asocial 18 29 1.56. 2.64

Subtotal 235 Youths with diagnoses

289 No agreed diagnosis

8 Readmissions

61 Youths not "claimed"

Subtotal 593 Youths on Rating Lists

353 Did not stay 30 days -

Total 946 Youths on the rolls, March 3, 1967

to March 26, 1974

Because of differences In the rating forms, severity ratings for judges and helpers cannot be pooled as they are for
-improvement. Judges' severity ratings show a similar negative correlation with improvement.

From these measures we derived "adjusted tracted from actual improvement (plus a con,
improvement." -Using multiple correlation, a stant to eliminate negative numbers) thus
least-squares equation was derived which yielding an "adjusted improvement" score, i.e.,
related improvement scores to judges' and improvement adjusted for ..severity. There is
helpers' severity ratings. Using this equation, no way of knowing whether such an adjust-
. predicted improvement was computed for each ment allows for the effects of 'nonrandom
youth. This predicted improvement was sub- assignment. However, it is fair to say that the

0
60 EXPLORING THE PSYCHO-SOCIAL THERAPIES

adjustment allows for the aspect of assignment provement was influenced by all the helpers
which appears to be most related to the youth's who said they had a lot of personal contact with
improvement. him.3
Finally, the process of,,adjusting improve- The improvement scores adjusted for severi-
ment was done seperately and within each ty, and averaged across all the youths, each .

diagnostic group. This was necessary since helper claimed sand across the youths each
most analyses arp -done for each' diagnosis claimed within each diagnoitic group, are used
separately. as performance criteria which can be related to
The diagnostic groups in table 21 are arrang- the helpers' personality measures.
ed roughly from least to most severe. The in- But before we get into personality Measures,
verse relation between improvement and let us loci& at the intercorrelation of the
severity can be seen clearly. Also it can be seen. helpers' imprbvement. scores. These are
that the more severe cases tend to be presented in table 22 and they address the
"claimed" by more helpers. question: To what extent i ert-Vho i sud-
The average length of stay at Loysville for cessful with one diagno is kr p'"a
the youths on the list was 462 days (std. dev: = cessful with another? The data bear on th
349). For those with agreed diagnoses, it was issue of the generalist vs. the differential
somewhat. longer, 526 days, (std. dev. = 255). hypothesis. Only one of the correlations is
. Length of stay declines substantially over, the significant, and that one .45 between ,Situa-
period- covered by this study (r = =.69, p tional and Subcultural, concerns the two 1.
Such declines are common in many
.001). mildest problem groups.' None of the others
institu ions during this time period and prob- are significant at evei) the .10 level. If we were
ably r flect administrative, not rehabilitative' to present adjisted improvement scores in
reaso s. Length of stay Is not related to im- table 22 the results would be essentially un-
prov ment (r = ,02) but it is correlated with changed. There is not much support for the
help rs' severity rating (r = .17, p < .01). generalist hypothesis here.
T average age of the youths is 15.3 years Relevant, perhaps, is our definition of
(std. v. = 1.4). There is a very slight tenden "helper": any member of the program staff be
cy fo o get boys to be considered more tween the years 1967 and 1974, who had served
disturbed, longer, and show less improve- for at least a year during that period and who
ment. Thiso Rive relationship between im- staled that he had had a lot of personal contact
proveme and age does_ not. hold for any with youths. The requirement that he stayed
diagnostic group except the neuroses. at 'least a year probably eliminated some staff
These various youth measures (averages for members-who were not suited to this kind of
each youth) will not be the subject of further work. If such are included in a study, the
direct analyses. Rather they will be ac- generalist hypothesis might get more support.
cumulated and averaged again for the youths Specifically, if there are some helpers in a
claimed by each helper.
'Presumably, if a helper who was successful with a par-
ticular diagnostic group consistently worked with the same
Helper Measures youths as another helper not so successful, both helpers' im-
provement scores are distorted in opposite directions, the
For each helper we averaged the improve- successful helpers' average being pulled down by the ef-
forts of the less successful, and the letters' scores being
ment scores of the youths with whom he claim- helped by the former. There is little evidence_that such con -.
-ed to have a lot of personal contact. Table 21 sistentpairings welrquent. Wit note that all such distor-
shows that many .youths are claimed by more tions are toward the ean, resulting in attenuation of the
than one helper. While this is to be expected, it criterion (see page 61), and lowered possibility of finding
presents some formidable statistical problems significant relationships. In other words, the error introduc-
alluded to in appendix 1. For purposes of the ed by this procedure, like most random error. is "conserv-
ative;"1 it does not produce invalid findings.
present analyses, if a youth is claimed by two `This correlation is strong enough that it remains signifi-
helpers, he is included in both-helpers' average cant after allowing for theact that we sorted over 15 cor-
improvement scores. It is assumed that his im-
-
relations to find it (Bonferoni's
 61
JUVENILE DELINQUENTS

Table 22, Intercorrelation of Helper Improvement Scores

for Six Diagnostic.Groups
(Pearsonian r's)

Average
Helper . for
Improvement Average Helper Improvement Score
3 Subcultural 5 Conformist 4,Manipulator 1 Neurotic 6 Asocial*
Score for

2 Situatignal 5 .25 - .22 .16 .07

3 Subcultural .24 ;22 .12 .10

a -I

5 Conformist
. 12 .20.
. 12 .17
4 Manipulator

1. Neurotic
.20
youths in a particular diagnostic group. The
The number of cases varies within this table since not all helpers claimed
33, beyond the .01 level).
bnly correlation significant beyond the .10 level is the first one (.45, N =

study who do not succeed with any of their correlation ma rices for the two groups are not
C----..
clients, we would find positive intercorrela- really very different and. the two groups are
tiOns for various diagnostic groups and this combined in the findings presented here.
would tend to be consistent with the generalist
hypothesis.
Statistically this phenomenon is called "at- Helpers' Personalities
tenuation of the criterion" and is illustrated
graphically in figure 2..In this figure there is an Helpers were asked to fill out the PTQ, the
overall correlation, but when those who left Strong Vocational Interest Blank (Form M, 400
the field before a year psed are excluded, the items), and the Quay Correctional Preference
correlation is lost. Figure 2 suggests the Survey. The first two provided continuity with
possibility that a correlation between criteria the studies in chapters III and' IV., The Quay'
will be found among inexperienced therapists Correctional Preference Survey was .the only
(D's and 0's) but not among experienced ones instrument we could find which was specifical-
(O's only).
ly designed to measure treatment 'ideologies
The 51 Loysville helpers include 20 no longer for delinquents.
With respect to the PTQ, it was our purpose
on the staff, but these 20 had been on the-staff
for at least a year during the period covered by to correlate the personality scales and A-B
the study. These, 20 former staff members are, predictors with helpers' improvement scores.
for the most part, either retired or working in We planned the SVIB as an item pool for item
other parts of the Pennsylvania correctional analysis just as Whitehorn and Betz had used
system. Most of them continue to live in the it. We also planned to use the Quay instrument
area surrounding Loysville. For these 20, data as an item pool for item analyses, these items
were collected by visiting them in their homes, being oriented toward specific treatments for
usually requiring several trips. For two of delinquents.
them the data were collected by mail. A The reasons for using the Quay instrument
number of analyses were conducted separately in this way may require some elaboration. The
for the 31 at Loysville and the 20 Who left. The Correctional Preference Survey consists of 15

41.
 EXPLORING THE d --
.

ir
.

HO- SOCIAL THERAPIES

Figure 2:
HYPOTHETIOLACORRELATION BETWEEN. IMPROVEMENT SCORES
FOR TWOlilfrAGNOSTIC GROUPS SHOWING POSSIBLE'EFFECT OF
'"ELIM MATING UNSUCCESSFUL HELPERS

ti
Success
with .

Diagnostic
Group A
0,
0
0
O 0 D.

0
0 D a.
O

0
O

El Stayed a Year

Left Before a Year

Success with Diagnostic Group B

items for each of four scales, the four scales be- appear to have been written on the bsSis of
ing designed to predict helpers' success: with clinical expeiience and judgement rather than
each of four Quay diagnostic groups: from any empirical test that they sttually
work for their intended groups. Our doneern
-.
(1) Inadequate-Immature
(2) Neurotic-Disturbed was with such an empirical test for each item,
(3) Unsocialized Psychopathic using .a slightly different diagnostic schema.
(4) Socialized-Subcultural (WO will also present findings for the four. Quay
Ingrani,1970, showed that matching
The last three of these would seem to corre titseil on the scales was indeed effective.) Our
spond with our Neurotic, Manipulator, and Sub- concern is to try to determine empirically the
cultural classifications, but the correspondence treatment philosophies of helpers effective
is by no means very exact. Moreover, the items with different kinds of delinquents.
 JUVENILE DEL INQUENrs

It table 21 it can be seen'that there were on- ings-in the one area where we eitpected them,
ly 14 'youtfie with an agreed diagnosis of "Con- namely: the Phipps Neurotic Predictors and
fbrmist" and 18; "Asocial." 'These numbers of Scales do predict the Loysville helpers success
cases seem inadequate on which to rest an with Neurotics. Underlining in table 24 is on
analysis. Bid we did conduct item analyses on the basis of two-tail tests. One-tail tests would
. the other four diagnostic -groups even though certainly be appropriate for. the _Neurotic
some might 'argue that .the 27 Neurotics are youths. On this reasoning, NSOLVE is. also
also an inkidequate base. significantly related to helpers! success rates.-
:For. the PTQ personality analyses we pre- Considering the differences in treatment set-
sent all six groups in table 23. There it :can. be tings, in r earth methods, in "patients," and
, seen that the numberof scales which...correlate in "therap ts" between the Lo YDC and
-significantly with'helpers' improvement scores the Phipps linic, replication co d hardly
is indeed related to the number of youths in the have been predictedi confidenc
base of those scores, there being no correlates" Details of the-item analysis are pr sented in
at all for the. Asocial group and only one each, table 25. The top half shows-item which dis-
Aot- the Ponformiat and Neurotic groups. The criminate success Ottlh .Neurotics at Loysville
tresting thing about these shigle correlates which are not incltided in the N Predictor
that
hat they will be seen by many clinicians as derived froth item analyses of the Phipps "per-
"reasonable." The. Conformist youth, who is cent of neurotics improved." It can be seen that
trying so hard for group acceptance .that' he in the two halves of the table there are only
will do anything he thinks will get it, is best three significant items in common
served by a helper who has a high regard. for ("Geography," "Algebra, " a d "conceited
individualism andstesists social pressure. The people") and that only two of th se are in the .N
Neurotic youth is best served by a helper who Predictor; However, "all of th items in 'the N
has low "Remorse" (items derived from Child's Predictor carry the approPri e sign in the
"Supergo" scale). One other "reasonable" rela- Loysvilli analysis, and they there s re "add up"
tionshiP should be noted. Most of the "Sub- to a significant predictor. While 'so e of the .
cultural" youths are black.. It is eminently NSOLVE items are weak 'in the Lo sville
reasonable that t bey are best served by analysis, "Calculus" certainly fits this d en-
helpers lowon theintidemocratic personality, -sion anctis-significanioier the Loysville hel ere.
most of these items being from the F Scale, an Thus NSOLVE scale as well, as th
/ indirect measure of ethnocentrism. P.edictOr eats -support in this item analysi
These "reasonable" findings, the "hoot- Tables 26, 27, and 20 present SVIB items
stripe of exploratiOn, tend to support the zthaNiscritdifiate ' success with Manip-
thht Our methoda have some.validity and that 'ers, and Situ
even with a small amount of data, meaningful 29 nd 30present
results emerge. s of the Quay. Correctional:
There are other meaningful results. Table 24 ,y. Interpretation of the
' presents the Phipps Clinic' A-B Predictors and next chapter which
Scales in relation- to Helper ImproveMent rin og tber :Ails findings in the last
Scores at Loysville. While there are few hree chapt
significant findings in table 24, .there are find-
 .\

:Table .23...CorrelatiOn of Personality Scales With Helpers' improvement

- Scores for Six Diagmistic 'Groups

Tursonlan r's)

Pationallty Verislie Diagnosis , liltutllontl Subcuiturer, Conlogolet blpulata Neurolla Atoclel

App, 4 Teble Hobe S1PRIT Mod . 14PRV Adisld 1.711T0171Wd HMV 5 Mild 611% 4 Mild P MPRV Adleld laPtillAdisld

31 Need lor Cloture -.036 -.024 -.073 -.017 -.060 .060 -.315 -103 .116 ;170 -.039 .007 999 =910
32 %Rolm in IM
Elervid of the Ego -.168 .019 .058 -.018 .101 .091 -,008 -.087 -.209 .199 .138.; -.006 .084

33 .Empathic Interest '163 , 212, - 008 -.036 .224 ,153 073 .050 939 -.054 .296 .239 -.012 .028
s..

34 Work Ethic ,108 .009 , .152 -.298 -.307 -,079 -.070 .154 , .244 130
1-
35 Nam' for Order ..082 .073 099 ' - .078'. -.141 -.079 . -,144 -.077 .167 197
'O6 Tailrace of the
Unreel Experience -.185 -.002 .032.. -023 245 ,228 .014 -.1:437 -290 -.339 .192 .182 '7499
,
-.187
31 Tolerance bf
Ambivalence =029 .158 .004 .016 .263 .181 .131 -.096 -.083 .255 .115 .181
38 Preference for
Decision Making

39 High Perini ,
-.196 '494' -.255 -.304 -.093 029 -.045 -251 f -,308 291. .294' 5.191 .92'5:

Slandards IV -.140 -.452 -.513 -033 -011 ,101 985 -.134 -187 .145 4070 *424 ; -.0(1

4q, Remorse
..
-.046 -919 -.342, -.369 .202 109 -.024
,
-.101 -.079 -.171 -.289 -.332 .144 .137

41 Antidemocratic -.039 -104 -on -.019 ', -,422 -.399 -,362 -.303 .188 .275. -.127, -097 .113 .) .096

42 individuellem -.043 008 041 7 7.002 .144 .193 .571 .438 -.185 .172 x,119

43 Deference. Anxiety -.130 -.140 -.058 -.083 2.232 ° -.326 -.289 -.076

44 Nurturence Anxiety. -.020 -.104 -.165 -.224 y -.049

.045 2100 *145 :

45 Science and
Morel Values -.126 -.007 .
-.042 .041 .027 .080 =9413 -.105. -.010 .,125
. 45 Preference for Intel.
lectual Challenge .001. 108 . .099 .080 -.131 .; .438 .278 -.101 -.114,

47, Extroversion .148 .036 .040 __-.1956 '1.080 .066 .280 .183 -.342 -176
Ow 1

-
.

(Inedequal6,1,1ure) '.218 938 .324 .321 . ,110. .092 -.016

s .054

Qt10Y,2 ,
Niurolialaiurbed) -151 .085. .567 .076 -.085 -.101 . .061' ..025. 7.291
Quay 3
(UnsoclalizedPsychopeth) 182 .195 094 .095 .' 046 ,,013 -.204 -,119 :161 .160 J44 .218

(SocitlizedsSobcullurell. -.017 -.104 -.321 -143 -.227 -.182 -.243 -.305 , -.193 -..113. :119 .032 .141
.

Organizational Level -.306 -.179 -.275 -.259 .123 .131 034i p.088 -.250 .

Number of Helpers 91 51 37 41 41

Single underlining indium that r 101110(10y dlltuent tram tiro twotill; double underlining it .01.
 Hilpers' Improvement Scores
Table 24.-CorrelatIon of MI Predictors With
for Six Dlagnostl.qroups

(Pearsonlan.r's)

Al
Pik loi
sod

Tsble timber
All Magoon

MPRV11/ Adistd.
.- Siuelloohl

MPRY 2 AdIsld
Subculual.

MPRV 3 .Adjstd
Conformist

WPM, ft 4, Adlild ,
Jisnlpulalor

SPRY 4 . Adjstd
NEICOR

WIN 1 AdJeld MPRV

Asocial

Adjstd

.118 7325 7220

7049 7019 025 -020 ,130
-319 ;-,141 -.189
-156 -175 -293 ,
I WI 22
-.225 .193
-As 721 -.237 -292 ,196 , IN -.146 -249 .066 '014 .

TOTES -.Ito 7128

-.158 .026 .025

.153 -166 7244. -.117
-.022 7.111 -..230 -.250 .201
31' r S -,129 -150 4

ir
...,
.201. ,182
.085 No ',IiI ..154 .120 ,

-.032 -.043 -A03 -.416

044 ; .' 921
3? SXPRSA .085 ....._.
t -900 -957 -036 7114 -952
022 211 .101 7918 .

029 -.053 7007 7070 900

31 SXPRSS
\s.
7,229 -229" .027 .094
.155 .213 7141 -,1131
1
-.270 7219 7153 -215 , -992 -079
31 SOUEST

7038 7114 271 ,229 -A18 7351 ti

.076 135 .

9 7023. 7128' -202 1-145 '-.128 -.193.

30
r :: .....
.-.388 , -281
r
178 .134
,054 _033 :isr 228 -.070 -.158
OTLD -095 -.118 -.161 -.224
A
.
.128 .105 -.184 -.024
.295 -.063 -.148 ,

,265 ,266. LIJ

32 7101 -049 -,119 -.174 ,

-991 903 -930 -929 .082

105 .007 ,108 -.043
32 013INNS 77127 7105 , -.1N -232 .107 .

7084 -.122 -.240 -.241

-,060 -.055 -067
II; -.033 -.152 -.110
30. ON M
-.114 305 259 -.294
-.178 931 -.038
005 ,7111 7180 7060 -098 c'
TOIL N

( 7053 -.149 .285 326 -.285 -.249

-.098 -.144,' -,009
7001 -,t59 7231 -.043
33

-008 330 307 -280 -263

-,066 -122 074 .022.
-.207 ..,047
33 NSOLVE -023 -,188

;279 .073 .042

7,115 -.231 .336
.046 -918 7095 ,

-.018 7184 .063

33 NOESTI 694

-.240 7:235 -.042 -.068

-.258 7189 .080 .129
,079 ,173
962 .087
33 NOEST2 ',048

35 35 29 29
29 45 45
41 41 . 29
31
No. of Helpers 8t..: 31

Variables In the stub are

'underlin Ing, at .01, Variables,In the heading are described on pagers 59410.
Singleunderliningindicalesthat r Is signilicantli all I eon! from zero al .05, tw o tall: double
. 1

described on pages 3540

 EXPLORING THE PSYCHO-SOCIAL THERAPIES

Tabie 25.-$V1B !terns Correlated With Helpers Adjusted IrnprOvement

Scores for Youths Who Are Neurotics
=35
- Item Number
SVIB PTO Item Gamma
5 Artist .42 .04
72 Politician .47 .01
85 Secretary, Chamber of Commerce .44 .03
86 Secret Sellivice Man = .48 .02
92 Stock Broker .49 .01
97 Undertaker .54 .02
107 Calculus --.36 ' .05
113 Geography" .44 .05
.
143 Boxing .47 ,01
163 Art galleries .43 . .03
182 EduCational vi s -.43 .05
232 Looking at a co ction-of r4Le laces .48 .05
292 Steadiness artd permanenceZirwork : -.39 .04
366 , Am quite sure of myself :61 .01
v

"N" Predidtor Items (see table.33) .

21 3-33 Cashier in a bank NQEST1 -.10 .62

101 3-35 Algebra NSOLVE -.39 .05
115. 3-57 GeomItry NSOLVE -.28 .14
119 3-58 kiteratiure NQEST1 .24. .24
148 3-61L Solving mechanical puz2les NSOLVE .36
149 3-162 Performing sleight -of - hand tricks NSIDLVE -.12 .56
173 3-66 Detective stories NSOLVE' 00 .99
185 3-68 Making a radio set NSOLVE -.28 .12
186 3-69 Repairing a clock NSOLVE -.23 .20
269 4-5 People who talk, very slowly NQEST2 .11 .60
353 2-35 Reading a book vs. going to movies (B) 00 .99
359 3-23 Jealous vs. conceited people (A) NQEST1 -.41. .03
386 3-24 Smooth out tangleS aryd& disagreements
betwq n people -.10 .76
(A) The "A" therapist the second cholce.
(B) The "B" therapist prefers the second choice. .

This item,was also predictive of success with neurotics at the Phipps Clinic and could have been included In the N
Pre it had been in uded In the PTO.

41
 67
JUVENILE DELINQUENTS

Helpers' Adjusted Improvement

Table 26.-SVIEI Items Carrelated WithAre
Scores for Youth Who Manipulators
N = 45

SVIB Gamma
No. ltern
5 Artist
-.34'. ..05
9 Author of novel -.70 .0002
27 College Professor
-.34 .05
.

Editor
-.45 .01.
31 -.43 .01
40 Foreign Correspondent

42 Hotel Keeper or Manager -.46. .01

57. Magazine Writer -.49 .01

62 Musician
-.36 .05
65 Office Manager
-.38 ,-.02

66 ~ Orchestra Conductor
-.50 .01

69 Physician ,
-.36 .04
=.35 .03
78 Reporter, general -.40 /1)2
93 Surgeon
112 English Compositionek -.43 .01

119 Literature
t,--c,4-
''' ' -.38 3

Music
--r.,
-.41
124 -.43
126 Philosophy -.40
129 Psychology
137 Golf
-.36
Excursions
-.63' .008
155
.007
156 Smokers -.45 .01'
167 Symphony Conc rt
"New Republic'
-.37 .04
176 -.40 .01
181 "Atlantic Monthl " .05
184 Social problem movie
.40 .04
190 Operating Mach .46 .02
191 Handlinghorses -.50 .004
195 Arguments -.42 .01
`199 Making' a speech
..200 / Organizing a play -.35 .04.

Teaching Adults
-.44- .03
Drilling §oldiers -.35 .05
Writing igports
-.451" .01
Progressive people
-.63 .002
256 .Sick people

Table continued on next page.

68
EXPLORING THE PSYCHO-SOCIAL THERAPIES"

Table 26.(cont'c) SVIB Items Correlated With Helpers' Adjusted Improvement

Scores for Youth Who Are Manipulators
N = 45

SVIB
No. item Gamma p
259 Cripples .43 .05
284 Determine the cost of operation of the machine .58 .001
299 Freedom in working out one's own methods of doing work .39 .03
402 Enrico Caruso, singer -.38 .03
304 Henry Ford, manufacturer - .48 .01

315 Chairman, Arrangement Committee' .35 .05 .,

320 Chairman, Publicity Committee .55' .002
331 Deal with things vs. deal with people (B) '.40 .03
343 Change from place to place vs. working in one location (A) .38 .03
344 Great variety of work vs. similarity in work (A) , .50 .03
345 _Physical activity vat mental activity (B) \ .39 .03
355
369
Few intimate friends vs. many acquaintances .50 .005
Have' more than my share of novel ideas .02
377 Get "rattled" easily .50 .05
378 Can write a concise, well-organized report .46 .01

(A) The 'LA" therapist prefers the second chciice.

(B) The "B" therapist prefers the second ch ice.

"1 7,
 69
JUVENILE DELINQUENTS

Table 2 . SVIB Items Correlated With Helpers/ Adjusted Improvement Scor4

for Youths Who Are Subcultural Identifiers .

N = 41

SVIe Gamma
No. Item

12 Auto Racer r , .34 -- .05

.38 .03
13 Auto Repairman .36 .03
32 Electrical Engineer
53 Librarian -
-.38 .05
, .40 .02
111 ,Economics

165 Vaudeville
-.45 .02
171 Sporting pages
-.64 .002
196 , Interviewing, men for a job ,- -.54 .005
.41 .04
276 Independents in politics .04
.40
278 Men who use perfume
. .450" .47 .007
294 Courteo& treatment from superiors .03
.41
349 Listening toaa story vs. Telling a story (B)
363 Win friends easily -.43 .03
- Usually liven up the group on a dun day
-.38 .05
365 . .39 .03
368 Have mechanical ingenuity (inventiveness)
.39 .05
381 Followu-P subordinates effectively .43 .04
396 Borrow occasionally vs..Practically never borrow (B) .4Q .05
398 My advice sought by many vs._Practically never asked (B)

:(B)The "B"lherapist prefers the second choice.

70
EXPLORING THEAYCHO-SOCIAL THERAPIES. '

Table 213:-SliiElptems Correlated With Helpers Adjusted Improvement Scores

for Youths Who Are Situotionals
N 37

SVIB
No.. Item Gamma
28 Consul .45 . .02
54 , Life Insurance Salesman .47 .02
65 Office Manager. -.51 .004
99 Wholesaler .
) .35' :04
-tog Agriculture .37 .04.
131 Public Speaking 4 --.45 .03
147 Observing birds (nature study) .40 .02
.168 Pet canaries .38 :04
169 Pet monkeys . 43 .02
190 Operating machinery .45 .02
191 Handlihg horses .49 .01
193 Raising flowers and vegetables o ..42 ..04
198- Interviewing clients - -.52 .03
199 Making a speech .04
208 Meeting new situations -.76 .01

217 Bargaining ("swapping") .43 .03

255 k, Foreigners -:39 .05
304 Henry Ford, manufacturer . 39 .05
308 William H. Taft, jurist -.39 .04
310 John Wanamaker, merchant . 36 :05
320 Chairman, Publicity Committee -.39 ,03
339 Selling article,-quoted.10%- beipw compatttor vs.
Selling article, quoted 10% above (B) . 48 .01
351 Amusement where there is a crowd vs.
Amusement alone or with one or two others (B) . 35 .05
354 *Belonging to many societies vs.
Belonging to few societies (A) -.53 .03
381. Followup subordinates effectively .53 .01
386. Smooth out tangles'and disagreements between people A5 .05

(B) The "B" therapist'preters the second Chbice.

'

efi
 71
JUVENILE. DELINQUENTS

Table 29.Quay Items Correlated With He !pert! Adjusted Improvement Scores

I for Youths Who Are-Neurotics and Subculture Is

. Item and Quay Behavioral Classification* Gamma

Su Ccess NEUROTICS

32. I would rather work in-a living unit wherelhe structure .54 .004
is clear and the limits are very tight. (BC3-A) '

pt. Workinbwith boys who like me one day 'and are mad at me
the next doesn't bother me at-all. (BC2-A)
.55 .01
50. It bothers me to have to work with boys wtlo act much
younger than their age. (BC1-D)
.41 .02
When I was a boy I had quitI a few delinquent kids as
acquaintances. (BC4-A)
.02
20. When a_-boy makes a decision,, l am willing to take a chdice
and go along with him. (BC2-A)
.03
9. Most delinquent kids would be o.k. if they had the work
skills to compete in the outside world. (61C4-A)
.68 .03
10 I don't mind admitting to a boy that I can make mistakes.
(BC2-A) .
.40 .05
44. Most delinquents are not much different from other boys.
, (lifC4-A)

Success with SUBCULTURALS

5. I do not like a living unit where the boys are not good in .44 .01
competitive activities. (BC1-D)

I find it is'very difficult to work with the kind of boys --.40 .02
29.
who are always demanding attention, (BC1-D) .
. . .:

20. When a 66y Makes a decision, I am willing to take a chance .42 ,0.04
and go along with him. (BC2-A)
.

should agree. D stands for disagreement. QUAY 1 Is im-

E3C37A" means QUAY3, sociopathic, and successful helpers
mature. QUAY 2 is neurotic. QUAY 4 is subcultural.
72 EXPLODING THE PSYCHO-SOCIAL THERAPIES

Table 30.Quay Items Correlated With Helpers' Adjuste Improvement Scores

for Youths Who Are Manipulators and Situationals
Item and Quay Behavioral. Classification* G

Success With MANIPULATORS

t

Most-delinquents are not much different from A9 .003

other boys. (BC4-A)

56.. When most delinquent kidrs learn right frorn wrong r. 3 .0

they will be o.k. (BC4-A)
It

54. ' Officers orccounselors should rarely It the boys know that .43 .04
they (the staff) are wrong-or have made a mistake: (BC3-A) tt

12. I feel more comfortable working with boys who are not easily .36 .04
bothered by thingS- than with those who are wore easily
..upset and tend to,show their feelings (BC4-A)

21. The idea that boys are delinquent because of forces. beyond .32 .05.
their control is generally nonsense. (BC2-Dy
,.
.36
, ,

41. I would rather work in a living unit where the boys are given .05
respohsibility to make decisions
)
about the rules within the
unit. (BC2-A)

,Success With SITUATIONALS

_is
It is particularly difficult for me to work with the kind of boys .40 .02
who are always. testing institutional procedures.(BC3-D)

Disciplinary actions for rulviolations are best handled- .35 .04

without a lot of botherabout what may have led up.to the
situation. (BC3-A)

31 Working, with boys who like mene day and are mad at m .37 .04
the next dOesn't bother me at al (BC2-A)

Most delinquents are hol much different from other boys. .36 .05
(BC4-A)

3. I like to challenge a boy to make a real committment to a .36

nondelinquent way f life. (BC4-A)

43. Scared and unhappJ boys need achance to express them-. .05
selves to an adult willing to listen. (BC2-A)

"BM:A" means QUAY, sociopathic; and successful helpers should agree. ,D stands for disagreement. QUAY.1 is:irn-
s
Menke. QUAY 2 is neurotic. QUA4.4 is subcultural:
Chapter VI
Clinical Interpretation
Valid clinical interpretation of empiriCal Neurotics
data involves two steps: (1) IioW sound is the
finding? (2) What does,it mean for clinics) prac- This brings us to the most rChustof our find-
tice? The second of. these questionsIs neces- ings, that the N Predictor derived from the
sarily 'inferential if the finding concerns theT success of Phipps Clinic residents with
personality of the. therapist. However, same neurotics, 'actually predicts the success of
criteria are available for assessing the first Loysville staff with neurotic delinquents.
question:' the soUndness of findings: (a) The What stronger evidence could there be of the
strongest findings, we, have are those where a importance of the personality of the therapist
relationship is observed in more than one sam- than that the same personality measure is
ple. Such a relationship is more dependable va id for the highly educated, highly trained,
and generalizable than One for whieliwe have a highly .selected psychiatric residents of the
single sample. (b) The. weakest findings' to be Phipps Clinic and for, the Loysville staff, most
interpreted here Are deriVed from of whom have no college education 'at all, but
'chaining" eorrelations. (cbrkelates of cor- have had years of practical experience? (tables
relates); such findings mayay be interpretable if 13, 24, 25). It is this replication that gives us
there are consistent patterns, but 'specific cor- some reason to claim, that all our 'change
relations ought, to be treated as suggiative agents" are "therapists?'
hints.for farther research. BetWeen these two, Of more interest perhaps is the meaning of
the strong and the weak findings, are, those the N Predictor and its significance for
which derive' directly from a criterion of suc- therapy. The N Predictor, contains at least
ceas (not. Chained) but for which we ha_ Ve no three clusters, the mo t important of,, hich is
replication lott/y one sample). 'NSOLVE: But let's l first at the weaker
At some expense to orderly content I Will clusters, NQEST1 an QES (tables 10 and
present findings in order Di their strength as 13).' The. PTQ sheds little lig t on these two
derived from data presented 'in this volume. enigmas (table la NQEST2 is particulariY
Findings from istudies being nducted by enigmatic and `-4t, fails to predict for the
marsoO be available an may Confirm Loysville sample (table 24). OnShe other hand?
the findings presented. here.-4But-,---for now, we NQEST1 contains the item "Literature."
have only what is here: Therapists successful with -neurotics like
I will use.the word "client". to stand for any literature. Careful comparison of the Loysville
patient or juvenile delinquent included in the data (table 25) with, Phipps data (table 13),
studies reported here. I will pie. the word *Wing those items significant between the
"therapist" to stand for' all the,therapkts and .05 and .10 level, reveals a number et items in-:
.hers in these studies. While some might ob- dicating that therapists effective with
-jeit to my calling a house parent a "therapist," ° neurotics like literature-and art
they' would prOhablyrbe equally unhappy with What does this ttention to literature and
ply calling a psychiatrist a "helper." Our find- art Inean' ih the :treatment of neuroses? .

ings reveal a-certain unity among these "change Perhaps it reflects int ive skills, or the us4of
i gents?' . the way one experienc s an object in relating
74 EXPLORING THE PSYCHO-SOCIAL THERAPIES

to . that object. Betz observes that many be pertinacious in searching for and in convey- -

neurotics like to 'live fanciful fairy stories. ing understanding to tthe patient at the expense:"
They act them out at an-adult lever. Moreover, of observing what is going on in the patient.
There is fpquently no therapeutic advantage in
neurotics have to work their way out of these doing so. 'As -Freud said, "The psychoanalyst's
patterns themselves. They cannot be pushed. ,,job is' to help the patient, not to demonstrate
This latter is the point of. our major N cluster, how clever the doctor is." (Fromm-Reichmann,
,NSOLVE. 1950, p. 19)
Most of the items in NSOLVE ircvolve solv:
. . the patient must be permitted to work on
ing a problem or puzzle (table 13). Therapists his own problem& at his own, speed without in-
who.are successful with neurotics reject these terference or pushing from the therapist.
items. The significance of this is probably best (Strupp, 1960, p. 209)
described in the following:
4. The seizing on every behavior pattern of the
In the early days ofinalytic technique it is patient (by the schizoid therapist) ... can 'be fol.
trie that we regarded the matter intellectually, the patient a pitiless opezation which he per.,
aad set a high, value on the patient's knowledge ceives #s detective-like All tge mkire so since
of: that which had been forgotten, .so that we the analysis takes place in, the described cooly
hardlr:made a distinction between Our Inowl- distant atmosphere. (Riemann, 1968, p.'719)
:./. edge and his in these matters. We 'accounted it
specially fortunatelf it were possible to obtain ommon theme.
Through these 4uOtes{.runs a common'
information of. the forgotten traumas of child:.
hood from external sources, from parents or The detective, the invatigator, the research
nurses, for instance, or froM the seducer him- er, the problginrsolver runs roughshod over the,
self, as occurred occasionally; and we hastened' clients' feelings. 'Another .SVIB _-item that
to convey the information and proofs of its. cor- discriminates success with neurotics is
rectness to the patient, in the certain expecte- "Usually ignore others' feelings vs. usually
tion of bringing the neurosis and the treatment
to a .rapid end by this .theans, Was'a bitter consider others' feelings. Therapists suc-
disappointment when the expected shccess.was cessful with neurotics chose the' second 'pole
not forthcoMing: How could it happen that the .(table 12). Moreover, this item correlates .38 (p
patient, who now had the knowledge of his .05) _with NSOLVE (rejects problem solv-
traumatic experience, still behaved in spite of it vine. There seems little doubt that these find-
as if he khew no more than -before? .Not e-ten
would thp rbcollection oi tiie repressed traana ings, taken `..together, confirm the cited:
come to mind after it had been told and de- literature.
scribed to him .
With of these pronouncements in the
In one particular case the mother of an hys- literature, what is so new about NSOLVE?.
terical, girl had confided to me the homosexual s Theodore NeWcomb says that the role of
experience Whi &had greatli influenced the!
fixatiop of the attack.ist; The mother herself had science is not so much to confirm common
come suddenly upon the scene:and had been a sense, but to define the conditions under which'
witness, of it; the girl, however, had totally . common . ense happens to be true. While it is -
forgotten it, although it had occurred not long -clear.th Freud is discussing the neuroses, the
- before ,puberty. Thereupon I inade a most in- context f, the ;other quotei is 'not so' clear.
structive observation. Every time that I re-'
peated the mother's story t&the girl she reacted Therapists' -problem. solving is particularly a
to it with an hysterical attack, after which, tbe problem'in ihe treatment of neuroses, ai;idnot
story was agaih_forgotteh. There-was no doubt n other disorders, probably because of' the
that the patient was ex, presiing a violent resis-
tance against the knowledge which was being
forced upon her; at last she simulated imbecility.
.and total loss of memory in order to defend her-
material. /
distress associated with wicovering repressed
,
The findings illusti:at thert relationships
self against 'what I told her. (Freud, 1959, p. amonglhe concepts of t hnique, personality,
362-30) and outcome. Specifically; they -suggest that
:the therapist must be careful in reacting, to'
Thii .... should rithind the psychatrist of two neurotic-Tefoblems. Freud sat behind-.Vie pa--
important fa 4s.. First, his interest in research e,..tient in illSidec'toNgive himself over to his un-
should be seeondary to his eagerness in dis
coOreg. data. strictly.pettment to his psycho- conscious tholighti. Freud's ice was in ex-
etitic obligatiOns..Second, sfiould not pressive face::Perhafila; a "poker face" need not
-..
 75
CLINICAL INTERPRETATION

be sccareful. Inany event Benedek is probably cessful therapist agreeing (vial' all '-of them:
correct that the therapist eicpressas his or her "Able 'to meet emergencies quickly and effec-
personality no matter what. But Freud is also "Stimulate the ambition' of ; my
right that the therapist should try ItOt to ex- associates," and "E3 pressing judgment's
press himself in.certain therapeutic situations. publicly regardless of criticism ". (table, 11).
The issue is not one of personality vs. tech- These are alhtzts-of leadership, but of a special
nique, but rather what light does -perionality kind. Whitehorn (1961) was much cancernad
shed on technique, and to What extent can we with the therapists' leadership role 'in the
modify our chaiacterJstic predispositions when treatment of schiz hrenia. He described this .1
theraReutic situations demand that we do so? role as "consultative" and "evocative?! These
Part of the, significance o the present per items and o s such as "Acceptjust criticism
sonality findings is that they define the kinds without g tting ime-I'suggest that an ,active, 4'

of patieiit for which certain prescribed tech- participati involvement is necessary to

niquei are,app priate, g reduce the s izophrepic's feelings of aliens:
PsyChoanal is is particularly associated tiou (Ricks, 1 74). Moreover, the active role is
with the tre tment of tbe neuroses. While . emphasize by the fact that: the- S predictor
many analys Would insist that it is useful for and SXP SA are positive)f correlated with .
a wide range 'of disorders, still many would in- the Preference for Decision Making (tables 38,
sist that traditional analytic methods need con- 16).
-

siderable modification when-applied to other The personal involvement of the therapist

thanthe-neuroses. If we assume a kind of cor- with the schizophrenic is not of 'the kind -pstakA_._
respondence:betiveen 'psychoanalysis and the or ly caned "sociability" or "extroveraion." There
neuroie's. certain- common-sense nOlions are- are a large' number of "social" items in the
difficult to ignore. Of all the mental healtk SVIB that do not discrizniate success:with ;
literature, the psychoanalytic literatuie schizophrenics(Full-dress affairs, Interviewing
most literary. The offices and homes. of men for. a job, Interviewing prospects in selh,.._
:analysts are 'adorned with art. Finally, ing, Interviewing clients, Opening conversa; -

'therapists higji in NSOMV'E reject the idea tion with a stranger, Meeting and directing .

that science Should_ have much to *say abqut

,a`r
people, Writing personal letters, Fraiiing
moral values' (table 16). Thus, the distrust of money fok a charity, Conge;rdal co-workers,' ..
Dealing wit people vs. things, Tell jokes well, -
science and systematic research (Colby, 1960, intimate friends,.
p. 52-57) so common among analysts may have Many' acquaintances vs: few
some functional utility in the treatment of Win friends easily, Many societies vs. few
neuroses. _ societies) The st 'two items db distinguish
°I,hose Alverapi ts who p scribe, drugs _f _grr :
sphikophrenics (table dig he relationship-th0
Schizophrenics helps schizophrenics is m re active and invorv-
ed than is' implied by `'sociability:"
Foe. schizophrenics we have one finding The S predictor and SXPRSA are not
which is repliceted. The, S predictOr is derived significantly related4.39 and .30 respectively)
,from the percent bf patients improved for to thPther4pists' success with schizophrenics = -`"
those ...Phipps therapists who. Prescribed no foy whom drugs were pre-scribed (table 18). Ittis
-. inferences from the null -,
drugs (table 11). This predictor and one:of its usually risky to ril
clusters, SXPRSA-,"are valid for another group hypothesis., and it -particularly risky when
of Phipps therapists,' those who prescribe samples are small- (N 3 12), but the findings
drugs, liu,oniY far those patients who did not suggest that this participative leadership May r
not be so crucial if drugs are preS ed. There
getrarirgs 78c. S and SXPRSA fall Short gs that sug-
significanee in predicting success when - is, 'hbcvever,- another setof fin
gest that ether aspe is bf the py with schizo-
drugs are ptescrribed. phrenics are some hat similar' regardless of-
!' Clinical interpretation 'of these findings whether drugs are prescribed. First, though, .4-
depends primarily on the meaning of SXPRSA.
Thii3 cluster consists of three items,..the sue - we consider the or gin of these other findings._
-t -,.
....
.0 ,. t . .. .
76 EXPLORING THE PSYCHO-SOCIAL THERAPIES
. .

So far in this chapter we hive been concern- person. This emphasizes anew the demanding
ed with replicated -predictors and their inter- requirements of psychotherapy with schizo-
pretation. We turn now to predictors not phrenics.
directly interpretable and the personality cor-
relates of these predictors which make inter-
pretation possible: While these "correlates of
correlates" are not as dependable as direct in- Personality Disorders
terpretation of predictors, there is one set of 4
correlates which is peculiarly replicated ih a Our findings with respect to the personality
fashion that gives considerable confidence disorders .are derived from the study at. the
If we compare the correlates of the .TOTL S Loysville YOuth Development Center in which
predictor derived from the drug-free therapy. we developed methods for dealing with some
at Phipps with the correlates of A-BDRUG knotty problems of therapeutic influence in a
derived from drug-related therapy, there is a milieu. While these methods yield reasonable
-hry similar Pattern. Moe precisely, if we ap- results, they cannot be considered as valid as
ply the drug-free predictor to therapists not in the findings from Phipps Clinic where each
a hospital setting (as discussed on pages 30-31, therapist was assigned to, and Spent; con- ,

most of them don't prefer drugs as much) siderable time with his patients. Moreover, for
and apply the predictor for drug-related some 'diagnostic groups at Loysville,_ the
therapy to therapists in a hospital setting, the number of cases is small. We will concentrate,
patterns of correlations are_ very similar (table therefore, on patterns of findinis tallier than
19). These patterns indiCate that therapists on specifics.
who are effective with schizophrenics are We have already noted above the similarity
tolerant of schizophrenic symptoms and they in personalities effective with neurotics at the .

are interested in .how other people feel Phipps Clinic and at L_ oysville. The other five
(Mosher, 1974). In addition to this "replication," diagnostic groups at Loysville do not. corre-
there is still another reason for 'accepting these spond with any Phipps data currently avail-
findings: they are completely consistent with able. Of The five (situationils, subcultural Wen-
Whritehorn. and Betz findings derived from , - tifiers, manipulators,, conformists, and asocials)
clinical records (p-age 16 above that success- all but the subcultural identifiers (page 78) could
ful therapists are, concerned with understand problbly be labeled personality disorders,
ing the patint, 'not with reducing his symp- Among the patterns of findings, the most im-
toms. , portant is, the absence of pattern across the
While we must use caution about sp cific diagnostic groups, This is true whether we
correlates, it seems appropriate to me tion look at' all six groups .or only the four personali-
Tolerance for Ambivalence because the finding ty disorders. In other words "juvenile delin-
is consistent with the "need-fear dilemma" of quency" and "personality disorder" stand for
quency"
.Burnham, Gladstone, and Gibson (1969). 'groups which' are much more heterogeneous
Thus the patterh,of traits for the 'treatment than the schizophrenias, the depressions, or
of schizophrenia involves, on-the one hand, ac- the neuroses.
tive, involved; structuring leadership, Moreover, if all of the groups are combined
SXPRSA, while, on the other,. tolerance and there are 'practically no findings at all (tables
understanding, S.D. While these traits are not 23, 24). The interpretation is clear. On the basis
completely inconsistent, they are empirically of present findings no treatment prescriptions
independent (tables 14, 16). are possible for the "personality disorders" as
It is possible, though-by no means clear, that a whole. But there are meaningful findings for
these two aspects of the treatment of schizo- particular personality disorders.
phrenics bear a relation to two general .aspects Of the four personality disorders, the .

of leadership in ,groups: the task function and numbers of cases are -very small for two: -the
the supportive function. Daniel Katz (1973) 'conformists; {and the asocials. The other two,
reviews the findings and concludes that rarely the manipulators and the situationals, provide
can the two functions be carried by the same more valid findings. <
 CLINICAL INIERIIILPTA,TION,77. ci
'See

40104re .

EFFETIVENESSPATTERNS FOR VARIOUS DISORDERS

Neurotics

Neurotics
CS
Schitophrenici Personality Disorders-,
(Manipulators ancid.
Situationals)
Depressives-

EffeCtiveness factors tnit.id be similar.

Effectiveness factors tend to-be reversqc1.

If we compare effectiveness factors for the If we compare factors t effective with

various diagnostic groups at the Phipps Clinic manipulators and those effeCtive with
and at Loysville, we find that the factors for schizophrenics, we find .several tel;ersals:.the
the two personality disorders are frequently personality, scales, Tolerance of .Urirealistic
reversed from those for neuroties,. schizo- Experience, Preference for Decision Making,
4ihre.iiies 'and depressives. This pattern is and Nurturance Anxiety (tables 16, 23), as well
presented schematically in figure 3. While we as two specific items,'author of a novel and
Would not necessarily expect similarities, it is social problem movies (tables 11, 26). There are
curious that-there are so many opposites. no congruences. There is even one. reversal
For example, there are a substantial number . comparing factors for manipulators and,.
of findings for effectiveness with "manipula- . depressives qperating machinery (tables 12,
,. 'tars: This group is probably nearest to what is 26).,
usually$alled "psychopathic" or "sociopathic.". 'While it cannot be asserted that the treat-
While *therapists who are successful with' ment of manipulators is directly opposed to the
neurotics tend to like liteiature and art (tables .treatment of neurotics,- schizophrenics, and
13;-25),--therapists who are effective With depreasives, there certainly is little_in common-_ _
manipulators dislike at least 10 SVII3. items inn in these findings. The emphs,sis for
the areas of writing,, literature, art, and music inanipulaiOrs is on no.nonsense controls and
(table 26). Still another reversal is for the Quay learning right' from wrong (table 30). The
. item, "Most delinquents-are not much different ,therapiit is himself not very sociable (tables
from other boyS." This is agreed:to by those el-. 23, 29). c
festive with neurotics and disagreed to by For :the situationals we havettle in the way
those effective 'with manipulatOrs (tables 29, ?..of descri4tion except that the 4tffective
30). therapist is not too demanding of himself (table.

96. 1
 EXPLORING THE PSYCHO-SOCIAL THERAPIES

2?). The original "A-B Scale" of Whiteliori:i and . samples, there are few:therapistS froth
Betz-is negative for this group, and-this will be outpatient clinics. However; aroful,stuAy of
discussed further in the next section (table 24). therapists in other outpaijerit clinics'indicates
The socialized subcultural identifiers that the correlational patterns, are
lincludiog. many , blackS) are not really that for t e original, Whitehoin-Betz
"disordered." They are socialized to deviint therapis t quite reasonable that eutpa-
subcultures. They are not well served by the tient clinics have a Psycho-socia' orientation;
'.'kind active involvement which helps
Of similar to.that of the Meyer-WhitehOtn Phipprs
I
schizOithrenics (table 24). The prime'.require- ,Clinic. Forqhfs reason; it is doubtful that.the
ment-is for' the therapist to be -free of prej- McNair-rev.ersal is a1seinantic reveisal.
udice, dficl'not. hurdened with remorse (table Phipps patients tend to be about Vul*cent
23). hbe inccesiful- therapist feels cotafortable female. The V.A. patients are overwhelmingly
with Subordinates (tables 20, 29). male. There is little ionforina.tiopiebOut how sex
In .summary,. the . labels "juvenile delin- roles influence the faetors.ettep4ive intherapy.
F and "personality disorders," cover a We can only note that Vh4e the, Phipps
_variety of pi.Oblem groups which have little in neurotics are also .prdominatelY female, the,
Common in these findings. For "situational'''; Loysville neurotics are all males. Yet some of
and "manipulators," more often than not, effec- the same therapist ylephdn'ality fadtorstseem to
' tiveness is gharaCteriied by the opposite of one be at work for'nenitics in' both settings. .

or another aspect of effectiveness in treating, Consider now Oa hipothesis Most/common-

the' /neuroses, the schizophtenias, and the ly held and, testediibat A's are more effective
- depressiOn& with schizophrenms 41$1e B's are more .effec-
tive with n4uiodes. This is usually/ a test -for
the significance °tan interaction. An interac-
The A ..B.Functional Reveisal tion will'he signidean-tg one group is relatively
more effective thait'tfie Other.,The A's- Could be
The original Whitehorn-Betz "A-B Scare" dd- more effective for both diagnostic groups, and
ferentiited the. A's, more effective with the Interaction. 'Would be significant
- if they
schizo hrenics, from the I-s, wit. so effective. Nwere reiative Cy more -effective with
When McNair et al. (1962) used the scale in -Schizephrenici.: In point of fact, there have
V.A. outpatient clinics, the reverse was found: been stUdieswiiere. the ,A's were more effec-
B's wereMbre 'effective than' A's. tive with schizOhrerlics and ,,equally effective
Although the diagnoses of the outpatients is rntreating neurotics (e.g., Berzins,
not known; these clinics generally halm about et al.,1972). Thit type of finding is consistent
80 percent neurotics, and the findings usually with some findinFs presented here: the original
have been interpreted that A's are more effec- "A-B Scale" (for ,§chizophrenics) is correlated
with schitophrenics While B'sare more ef- only .22 with *:4Percent,improved, neurotic"
fective with neurotics. In, fact, there are a (table 14). Thus there Should 'not be any dif-
number, of possible explanations for the ference between A's and B's -in: their success
McNair reversal: 'the A-B semantic reversal, w4th neurotics. ; .

female patients vs. male patients, There, is, hoWever, another interpretation:
schipphrenics vs. neurotics, schizophrenics vs. the 'Super-A? hypothesis is that A's 'can do
personality, disorders, and severe vs. mild better with schOophrenics .and at least as well
disorders.- with other diagniistiCgrolipd-beaddeorsom-e
With respect to theSemantic reversal, it was inherent chara'eteristics of A therapists (Char-
shown that. the original }'*A -B Scale," and in- tier and. Weisi,- 1974). This hypothesis could
deed other A-B clusters such as SD and also_be' supportd by data presented here. It
DBUSNS, tend to reverse their meaning wen could be argOd that since the A's show
we compare therapists who have high em- greater tolerance (table 16), they, are more flex-
pathic interest with those not so high, or when ible-, haVe great!er "plasticity" in Reich's terms,
we compare hospital therapists with and ,can adapt io a wider' variety of patients'.
nonhospital ,ones (tab" 3, 9). In the present There are some studies which are suggestive
 CLINICAL INTERPRETATION 79
I I . ,

Of. gi' ater plasticity/ among, A therapists different groups of therapists (tables 16, 58).
(Schuler and Wagner, 1975; _Smith, 1972; For interpretation we are therefore dependent
Haiezifu , Ma.rtindarerand Kaplan, 1975). uPoni the "correrates ecorrelates" wider cir-
ypotheais most consistent with the: cumstances where 'we cannot be certain that
dings, presented here is t t the McNair °'-we are usin'i an appropriate 'population,, i.e.,
everaal is somehow/ related to he presence of one similar, to the 35 Phipps Clink residents
personality disorders in the V. . sample, or to from yillonc the" cluster is .derived.
the mildness of the disorders in that- 'sample.. Granted all of these- reservations, ihe clinical,
Findings discuised in the list section above interpretation of SD is that the therapist effec-
point to 'a reversal of success; factors between tiye with 'psychotics is tolerant of 'regression
personality, disorders and the schizoPhrenicS,. and interested in others' feelings (table 16). He
it is. possible, for/example, that "plaiticity" is is probably high in nurturance anxiety, sug-
not helpful in treating manipulators. gesting that (in spite or his personal involve-
I-Towever; for :only one of the personality ment with schizophrenics) .is wary of al-
disorders is there, a negative relation with the lowing others, to become dependent upon him
original 4`A-B Sale" the situationals (table (table 16).-
24) and these happen to be the mildest of the , These inferences are much in need of replica-
.
disorders studied here. It is therefore difficult tion.
to say Whether the 'McNair reversal is related
to diagnosis or to severity, or both. t.
In any:.event, it should be kept firmly before The Depressives
'us that the findings presented here are for
long-term treatment of institutionalized In :addition to SD, there are two other
clients mlize severely disturbed than most out- clusters predictive of success with
patients. Further studies are needed to deter depressives. Not only do these therapists who
mine - low thebe findings apply to mildly -are effective with depressives_ reject 'manual
disturbed outpatients. Evidence in the present and mechanical interests, they also reject
studies indicates that effective treatment' of. buSiness interests (DBUSNS, table 16).
,' schizophrenics is different from that, for In fact, of the three depressive clusters (SD,
neurotics, but it is not the opposite, unless the DBUSNS, DCIVIC); DBUSNS is the dominant
neurotics are in fact personality disorders, Or one (table 14). Unfortunately, it is multisernan-
mild situationals. tic (tables 16, 58) and requires the same reser-
vation a§ for SD.
Our best guess is that this rejection of
The Functional Psychoses business interests reflects a high degree of in-
diviaualism (rejecting of social influence) (table.
The SD cluster is .of particular interest for, 16). Thii appears at first glance to unlike-
two reasons. The dimension it contains, rejec- ly findinosince business men _hink of
tion of manual'Cand mechanical, interests, themselves as individualistic.' However, if 'one
dominates the .original Whitehorn-Betz "A-B
Scale" (table 14). Mdreover, since it is preidic- It it necessary for many of the findings in these studies
tive of success with both- schizophrenics and to distinguish betireen ideologies, on the one hand, and-
behaVioilhat 4nay relate to' ideologies, Ou the,other. The
depressives, it is possible that it stands for distinction is illustrated in several ways among the
traits required in the' treatment of the func- variables included in the PTQ. Frequently both aspects
tional psychoses..1 were included toteSi which 'seemed, to be the relevant one
Unfortunately, the cluster is not directly in- for the psycho - social therapies. .

terpretable in psychotherapy, and, worse still, For example, what is the. significance of the rejection of
manual and mechanical interests? Is it possibly rejection of
It is multisemantic it changes its meaning for work as an activity, or possibly rejection of the work ethic?
It appears that the latter is primarily involved (table 16/.
'At. the Phipps Clinic, "depressives" included both Deference Anxiety is concerned with accepting authority
psychotic and neurotic depressions. We not be certain in a concrete interpersonal sense. Authoritarianism '(An-.
whether our findings .apply to one, or the &her, or both. Continued on next page
 1
EXPLORING THE PSYCHO-SOCIAL 'THERAPIES
.

Oinks of busin4S'S men fashions and other dividual resistant to, social influence; wary of
,aspects of socials 6infor-rnity in business, it.is others becoming dependent upon'him, and anx-
probably qu e reasonable. The rejection. of ious about having to defer to others. On the
businessint rests also probably reflects anxi- other hand, he probably has a broad social con-
etY abOut having to defer to callers (table 16).' cern. He is, not.authdritarian or ethnocentric.
liejectionA of- business interests is also He' probably views persons who have mental
associated .with the end of the an- disorders as similar to other people.
tidemocratiC. (anthoritatian). F Scale (table 16). - .

These ther4pists' are probably low in ethnocen-

. trism and "pre ,not, likely to have negptiye at-
titudes ftward serious Mental disorders The Generalist Approach and the
Wilbert aid Levinson, 1956).-Finally, rejection Differential Hypothesis
of business 'interests is associated with :47^,

tolerance: of the unrealistic expekence (table In these studies there is little evidence to
16). support the genecalist notion that-there are
The third /luster has been labeled pelvic. good therapiits aXid .bad ones regardless of
It l consists df liking civic's, disliking J P. what kinds of problems are being dealt with.
Morgan, :financier, and ,,preferring ajob i On the other h-and.the,re are significant, mean-
large corporation rather than being an inginr Jindings to support the differential
dependent business man. this cluster is .cor- hypothesis that specific interests and
relate& with a liking for being a labor_ ar- predispositions in therapists are related to suc-
bitrator (tableS 12, 14), cess with particular diagnostic groups.
With:DCIVIC we are trying to interpret col': The fixs question to be asked is whether
relates' (rather than correlates of correlates). therapists ,who are effective with a particular
Still it-may be somewhat of a flight of fancy to diagnostic group are alsd effective With/other
assert that these four items represent a com- diagnoses:An affirmative answer to this clues-
mitment to a broader social concern. Barbara tion would.imply a substantial positiVe correla-
Betz suggests that such a person. may provide tion among the measures of therapist effec-
a model of broad social interest for a person ' tiveness for the various diagnostic groups.
who is overly dependent up\ontoneother person Among the Phipps-- residents treating
or only a. few other persOns. While the schizophrenics, depressives, and neurotics,
schizophrenic is not at , all attached, the while the intercorrelations are pOsitive, they
depressive is too, attached to` others (personal are low and only one of them approalhes
Communication). Dependency in depressives is significance, the one relating ,.effectiVieness
_

summarized by Chodoff (1973)... with schizophrenics to that for depressive's

In ,summary, our inferences for the treat- (table 14, Triangle' 1). This may .reftect the
ment' of - depressives must necessarily be ten- preSence of a psychotic factor iri thOse two
,
ta.tive because of the nature of our' data. They gaups.
suggest that the therapist who is effective Ainong the Loysville staff,there is only one
with- depressives is, on the one hand, an in- positive correlation. There is a positive rela-
tion for the two mildest groups
,
the Situa-
tionals and the subcultural identifibrs, both of
Continued from Kevious page which some would assert to be "normal" (table
22). .
.

tidemocratic) is a broad ideplogy that tends to define rela- :

tions with almost everyone. peerS included. The correlates Thus although there is no support across all
of these two' are not so very different (table 16). diagnostic grodps, it is possible that the
.

Finally, Individualism isa personal resistance to social in- -' --generalist-notion-will-receive-some-support in

fluence (akin td Independence of Judgement). Rugged in- those settings which-are concerned primarily
dividualism, the opposite of what we are inferring in with mild probleMs.: As pathology becomes
DCIVIC, is a. broad ideology. It is suggested that
busineSsmen are high in rugged individualism but low inIn- more severe, it may become more differen-
dividualism, and that this explains the findings with respect tiated, and require more differentiated treat-
to the depressives {tables 14, 16). ment.

4--1
 CLINICAL INTER PRETATION
81

A second c9nsideration is that the generalist show such a factor (Garfield and Bergin; 1971a).
notion maY/r*ive support in those groups of They found their successful therapists did not.
.

therapista!yfio are in training or who have lit- have elevated scores on

on certain MMP scales.
tle exPerp,fice. In effect, there are in such There is widespread acceptance of t e idea
group,s/4,4Me persons Oho don't belong (not that therapists who themselves are po rly ad-
gooisit;''any kin& of therapy) and who will justed are not likely to be helpful to of ers. 'In
leave; or go into related fields such as research the present studies there is no measur of per-
or. administration. Were I not concerned for sonal adjustment. (This was by desi n. See
the anonymity of. my respondents, it would be appendix 1). Therefore,/the present studies
possible to point to some anecdotal evidence in provide no test of the niost plausible eneral
support of this' idea. There is no systematic factor in the literature. On the other h nd, the
evidence in the present' studies to support it, present studies would go beyond the adjust-
but the idea is discussed (page 61). ment -hypothesis and assert that perrnal. ad;
There is a further reason why the generalist justment is not enough to be effective in
hypothesis gets acceptance. If ,a particular psychotherapy.
-clinic populatioli is dominated by a particular What are the clinical implications of the dif-
diagnostic group; factors which. are really ferential hypothesis?' It places an f ,pormous
burden on diagnostics. (Alternatively, the
specific to that group appear to be general, for
want of -a wider reference population. t
findings reported above are all th more' im-
As for the present evidence bearing On the pressive, recognizing the frequentl reported
generalik approach " and the differential unreliability 4 diagnoses.) If the therapist- is
hyPotheses, second: kind of question can be reasonably well suited to his specialty, there is
asked: are there trait that all good therapists little that can be said about appropriate treat-
have in common? Ain ng the Phipps residents', ment unless we know something about the
a iow Need tor Closur is found correlated with client's problem (Horwitz, 1974). I
the predictors for all three diagnostic grOups: . 'When we consider the implications of the dif-
the 'schizophrenics, the depressives, and the ferential hypothesis for milieu !therapy, it is
neurotics (table 36). the notion is beguiling: all clear that, while individual therapists may be
mental disorders are complicated and incon- able to adapt their styles to their clients'.
stant; closure can never be achieved_inreatink needs, milieus are not easily c angeable, nor
them. can they easily discriminate among their
While such an ideal should net be rejected, members. Consider the following Quay items
there are two reservations. The apparent (tables 29, .30):
generalitS, of this factor at Phipps could be an I don't mind admitting to a boy that I can make
artifact of the way the PTQ was constructed, mistakes.
certain idlosyncratic items being excluded (pp. Officers or counselors should rarely let the boys
41-42). The second reservation is that the know that they (the staff) are wrong or have
need for closure fails to predict success with made a mistake.
neurotics at Loysville, or indeed, with any
group at LOysville (table 23): Therapists effective with Neurotics say "true".
The most interesting evidence in support of to the first of these items. Therapists effective
the differential hypothesis, and rejecting the with Manipulators say "true" to the second.
idea of "the good therapist," is that, while This suggests that clients should be assigned
.),- 'specific .therapist personality factors predict to cottages or nursing units with varying
success with specific` diagnositic groups at milieus designed to match varying diagnoses.
Loysville, there are 'practically no factors This idea has sometimes been deliberately im-
--which-predict, therapists' success with all of plemented (e.g., Gerard, 1969; Palmer,. 1974).
their youths (tables 23, .24 John Williams, Director at Loysville, feels the
One final reservation 'must be made about idea has merit provided the institution has con-
the failure of the present,;Studies to show a trol over intake and can exclude those who do
7 general personality factor in effective not fit- any of its treatment capabilities (per-
therapists: One study might .Eeinterpreted to sonal communication). In any event, the dif-
82 EXPLORING THE PSYCHO-SOCIAL
6
THERAPIES

ferential hypothesis has many implications for . for the most part, exprefiing their own pert-
the design of therapeutic services. sonalities, then.the implications of the differen-
Further implications foi one-to-one therapy tial hypothesis are far-reaChing; they extend to
depend upon future studies that tell .us the design of services, training programs and
something about therapisti' abilities to adapt . core curricula, selection for such services and
to individual clients with varying diagnoses. If programs, professional services review, aid
therapists can adapt, given an understanding perhaps even to licensure specialties. The im-
of wl\at is needed by each client, there_islittle plications are too far-reathink for reasonable
reason, for long-run concern.. If, however, inference from our presently available
clients are best served by therapists who are, knowledge.
 Chapter VII
Implications for Research
The. primary research implication of the I am not suggesting that we be less rigorous.
studies. reported here is as follows; relatively' Indeed, in the last part of the chapter, where
simple conceptualizations, and naturalistic ore detailed implications are presented, I will
methodologies have. yielded meaningful, con, Cj plain about the lack, of a particular kind of
sistent, and, in some cases, replicated' findings. gor in most 'of today'S behavioral research.
As noted in the' Overview', such findings. are Ifs is the erroneous assumption that all
not the- rule for research in the psyc cial hu an beings. are alike (spawned also, in
therapies. The first part of thisiecha ter c biology). There is a general failure to adequate-
diders by simple concepts and Metho s are ap- ly define human, subpopulations, to sample
propriate to this field at this time. them properly, and to attempt to include all
The4 is little doubt that researchers are at- who are sampled.
tracted to and feel pressures toward certain This chapter is concerned with research
prestigious . approaches, multivariate strategies in exploration, the limitations of the
models and experimental Methods. In general, biomethodology, and the specific implications*
the prestigious approaches are those of "ad- of the studies presented above. Throughout
vanced" fields where there is a large body of the chapter the emphasis is not on determining
knowledge and complex .conceptual systems, which approaches are "right" or "wrong" but
.
and.where the task is to prove or disprove cer- rather which are optimal in. the sense that they ,'
tain deViations froin, the existing .store 'of maximize returns and Minimize err.ors- and
knowledge. Such is not the cafe for the psycho- biases in this research area at its present singe
- social therapies; the needs are for exploration. of development.
Generally speaking, referee systems tend tO
penalize researchers who chOoseto explore, far
the referees tend to start with the prestigious' Research Strategic in Explora#on
approaches as ideal. Moreover, it is the
prestigiOus. approaches which are taught. In the year 1500, Colninbus ,had already
There are many journals in which it is impos disiovered a way, to the. East by sailing west,
sible. to present exploratory findings. Ins but most people,_and indeed most scientists
ploration, the methods develop with the data' believed that the learth Was flat and that the
they cannot honestly. be detailed first. Thus, surf and the planets reirolved about* /Coper-
the prevalent scientific biases lead editors to icus found this conception unsatisfactory.
prefer an experimental study with null worked for several decades en the probleni,
findings to an exploratory one which presents - inally sending the last of hi's :De 1e31)021,-
-some systematic findings and some loose ends. t ntbus, to print When he was on his deathbed
This, then, is the bias against exploration. in 1543. He claimihino *oof It would, be nearly
But the psycho-social researches suffer from seven more decades before Galileo, using the ..
another bias too. The methods used are those newly discovered telescope, Could provide the
developed in the physical and biological proOf, and yet another seven decades before
sciences; they are not appropriate to studies of Newton could describe with precision the force
,
social behavior. This is discussed in the second of gravitation that explained the movement of
part of the chapter \ the planets, of the tides, and of objects on earthy
 r '
EXPLORI THE PSYCMO-SOCIAL THERAPIES

The qt(estion COpernicus.asked did the'sun Reeognizing.that each of the blanks actually
and the planets kevolve around .the earth; dr stands for a number of Variables, not a single::
did the _earth-and the planets revolve akiand, one,,it is appa nt that the number of/variables
the sun Was
--- not an easy, question, but it/Was a //if is legiOn:.,In thi situation vv do not seek proof,
telatively oircurnscribed one. Note thathe I We explore re a Coperbic.uSdid. We Measure a
not ask:.Are the orbits circles or.elypses?; This .laige, number of variableS as cheaply as pos-
was Kepler's problem. 70, ye!tra later: He did Sible,,sO that we may' di,Ver thoSe Which' con-
not ask whether. and the planetS 'rotate tribute most' to thr.yarianee.
on thdir axes, ,This was Neither' But exploration, is only a partial answer to
could .have. been asked the ,f question; There are too many variables
Until:that:of Copernicus was ansWered. eveir for expIoxitiOn. We, must find some Way
, -

This illustratetWo propositions: to ' simplify the blankety-blank question.

Exploration should not be bUtdened with Ijnagine,how many patients'and therapists we
prOof. When we demand proof,' We inhibit would need to; deal with this question!) -;
systematic exploration. The literature of the .. Thq task is.tO OptiniizL In the IntiochiefiOn
psycho-social' therapies is to to polarize (chapter:' II)) ,T.:MentiOnect..one kind of nonop-
toward case reports on the one hand; and con- ' timizing ;pOurOg; :enOrniOus energies into:
trolled experiments on the other, ,,measuring,the OntcOine. while the independent ,

2. Ceitain questions, are, prior questions. Un, vatiable'iS taken largely (for granted. Another 1.

til the prior questions have been 6alt with in. aspect of bptiznizingiS in, the ch,iee of critical
'at least a pieliminary way, the niceties of the . yariables., Certain: ''existinA" studies suggest
model cannot even' be approached. For ex,- that certain of '.'the blanks , are critical while
ample, when '/'Suggests that abet* can be, 07rie of the critical
munication is Subliminal, he may be ones
,
is the type of prOblenit
.,,,
the di4noiis. In
precipitating a Prior 'qUestion (page above)., -those studies ,iy;nere .the type of therapy was'
If communication in therapy is largely .Measured', and where diagnosiW4s included in
subliminal,; certain kinds of Manifest proCess the stlidi, the ',differential hypothesis received
, analyses are not likely to yield useful findings. support ;Thus ; "type: of problerh" (diagnosis) is
In the psyCho-social gierapies Wellave not yet a critical vdiriable. .
Ai
identified the relevant variabies.` Are there Not,1:Only: can noncritical blanks be post-
Any gbidelines that can be Set up ttat would pOixed; a:-Critical blankcan be poStponed if it is
help us to proceed in an optiml fashion? highly correlated with a critical blank which is
Consider now the following paradigm: Given. iii*Itrded. When we incinde the type of problem
BLANK time limftatiOns, and* BLANK in a study, we have specified to 'a degree' the
organizational climate, to, achieve 'a BLANK type of OUteome. Inifact, some of the' specific,
objectiVe, with a BLANK type 4f patient,.who phobias; for example, have a very limited im-
has a, BLANK type of PrOblenn; at a BLANK plied outcome; and some of the most successful
stage Of development,a BLANK type of. therapies address themselves-only to that out-
t`erapist, should use a BLANK type of tech- come: relief from'the,phobia. This is' not to say
niquei in coMbination with BLANK othertypes that outcome pis akyay or, -even usually in
of treatment, under BLANK conditions, at a terms of reduced symptoms, but rather that
BLANK stage of therapy. The task is to' fill in outcome is implied in the' problem. It may be
the BLANKS. While slightly overdrawn, this nice, and even elegant to measure social func-.
paradigm is not really too different from some tioning, or ,income, or selfiesteem before and
you will find in the current literature. Having after, but these outcomes are differentially in-
defined the problem as imposSibly complex,
,

volired in' differentproblems.

each investigator is free to pick two or three .: For example, it is frequently asked whether
blanks he or she, likes, and to, emerge some the Outcome, of therapy should be judged by
,

time later, as often as not, with null findings.' -the patient, by his or her family, or by/SoCiety.
Unless Ale resea:rcher picks the two or three fact, neurotics generally come to treatment
which have a critical role in the variance, null because of Their own feelings pf distress;
findings are almost certain. psyChotics are frequently brought to treat-
 IMPLICATIONS FOR RiSEARCH

Ment 'by .their. families; and sociopaths' are

,
founder on the, ele.gance, of du5 :methodology.
brought to treatment by society.. there be 'y "approved'.. 'methodology s that ,which
a general answerto this question that does rtot proVides "Proof.' The "apfiroved1'; method-
consider the nature of the problem? The cor olOgy assumes a body of knOviledge, the task
relation between type of problem and type of being to "Prove" a deviation from that body of
outcente is fuither discussed by Kiesler (1971). knowledge The inet.hodOlogy. therefore makes
Our first task in the paycho-social therapies is the saMe.MegalomanicrrorS that our
to disCover what ameliorates the problem. The . tualizatiods made when aPplied4O reSearch\in
fringes come later. the psycho-social theilapies.
There is another correlation in the blankety- The taak'cif science is. that Wibartteire7
.blankuestion to Which we have already given be suite that We hair!) achieved finaN:Potif..-- NO
conaiderable Eaten ion; type .of . 'therapist by matter how. elegant` or .perfect the stucfie''s'Ana
type of technique.. It is not uncommon',for our plications, some Young. upstart will 'de" roy
MethodOlogists to Worry, about how i!*e them, with anew-yariahle previously
. ever separate the therapist froin his or her(', = There is suchi-a thing as 'relative pro4f;:sOnie
'technique. They are thus implicitly recdgnizing things are better "known".than Otheri: But if
'findings that- the therapiSt's we 'do. not knOw what should be or .

scinality determines', technique. T view this con- manipulated, therela no sense totolirocif. If we
foanding,not as a; worry. but as a boon; it per- ' do not understand What dirnensionkare baisic3
Mits us to simplify the paradigm by :there can be no relative proof..I.Ptherk'is ho
stibstituting personality kir technique as a first body of .knowledge;
approxiniatiOn.Later,,we can separate them in In Qne recent experi nt' thliTe5w,pre three
specific studies designed for this. purpose. groupsNn.experimerital group*id4O'Control SSA

.', Another critical variable is "other types of groups, NplaCebo contecil, and:MitjWcontrol.
I
treatrnent." If;. for example, drug therapy has The research" protocols Con--
been shown/to' affect t behavior of patients siderable resources were expended. in nkasur-
(for better or for worse), this variable must be ing outcomes and other characteristics ht pa-
included. Then there is, "milieu therapy" tients, in all three gro4s:.-If the experiments...c
more often than not called a "control group." had provided. instead 'for three different kinds
Howtcan we hold the'expectittion that contact of psychOtherapy, and had pleasured the kindis
with a psyclioltherapist, in many, cases no more of. Psychotherapy, the increase in:expenditilie
than 5 or 5 hours a week, is critical, :wile the would not have been great, but we wouldledOw
score of hours with an aide is nothing? In'the a great deal more about thglpSycho-Social
chaiter on juvenile delinquencY, George Furse therapies: three valuated ,ipStead of one.
and I tried to deVeloP the Methodology that While the depen nt wete, -very
shill permit us to study one-to-one relationships. ;. carefully meaSured;t he .11-Kleptiricieni
within the milieu. f
therapy, was pot ineaorectat. all, So there still .
was no "proof."
,

reading of the current literature is that ;

the other blinks have not yet ..shown W are easilyibegniled 'by :the; niceties'. of
themselves to "be critical. The critical ones ap- m d, but; as Of;;: these
pear to be: (1)the type of problem niceties are ''keally.
(2) some ineasurg of chan4-With -respecif.d. the say that Wegian
problem, (3) the therapistlas a measure of the about methods;' we Cannot. concern
nature of the psfyCho-Social therapy, and (4) ourselves with..biaic,iirinciPres.!:(e.g.;.:tandoM.
other treatments, aSsignMent, or some way;
But the desig4 must be' deveropek for each
problem, not taken, troth., a textbook:. in .ex-
Limitations the °Bion*thodology" perimental desi
Not only has he biomettiodology steered',.uS
Let us assume that we can do a rational job toward see proof. It has-failed to4 definWiin.
.

of simplifying our. conceptualizations, that we Portant m hddblogical prinCiples that are needed
can learn to ask the. right questions. Still we . for the human Sciences. Pasically
. /
 EXPLORING THE psycHo-soc

ologies are those appropriate to, the evalua- lsome-of the expernnenta! ,xnampullttions,.
tion of hybrid corn, ,or -tip subdivision 6f The fart that operant conditioning Worked was
amoeba, or the growth of tumors in rats. T,hese ,iloOnew at Hs.wthOne.: What:404.10W Was:the
methodologies do not allow for the effect§ of fillet' 'that.. the eiperlinentafmaniPniations,
Iranian expectations, with one exceptiOn: the operant :conditiOnIng;slid not explain
`"placebo",is designed to allow for the all of ithe:effeCO.:These: added-eifects 'Seemed
that "treatment will help." The -phicetio-
,
to be due: tOeipeefationsilor-the most part
makes no allowance for negatii.re expectations, shared eXpeetationigioUt environment. In
.nor does it make allowance for the power of some rases =these expeetationatilted. in in
shared expectations in a social grOup. Picture. ceased :prodnrtiwity,: In ;others /the effect, was
youkse about to start group therapy with the negative ;*-=`,t
followi group: theY have just arrived to_sit Parsons: :acknowledges, the importance. of
-wait fo; the end of the, world whichis' to Sha701:::eXpeclations;1.bUt:..he, chpoies to call
'""happen at midnight. Migtit they not be dif- ihenf`'CiinaitiononOt-"Canses.7 There is much
ferent from your usual group?' 6.6 of , this in the .hunian sciences: "My
In order to see the power of expectations, Variables are causes; .yours /are conditions, or.
is necessary to -have a' clear understanding of circumstances; :
the '''Hawthorne Effect." Mountains of material Not: only dees the j.bioinethodologY, fail to .

have been written ,ahout the Hawthorne ex-,- :handle the c6M-Ple_exPectatiOnsof patients, it
periments, and sstiil the misconceptiiins'. fails to .4eal ::'With. the:..expectations of
abound. It would perhaps be appropriate to therapists. We-Might get ,i.w4y, with ignoring
Say: Never have so .many been so _confused° the therapist in the:dirug Oierapies, but not'
about s uch. At a recent large sympbsium on .the psycho-social therapies 'Where relation-
evaluation, he chairman was heard to say: ships are so important.] Oe'Con this below.'
"The Hawthorne effect :.was an experimenter I have stated that the biOinetliodoIogY steers
effect, and experimenter effects wear off." us toward proof instead of tOwarditheexplora-
What is properly called' the Hawthorne effect tion that is need4d;'::fU#her,?the biOmeth-
was-a very special.kind of experimentation ef- OdologY is unabletki/j*iildle, ind?in fact may ox-
fect that did not 'wear 'off. In same instances acerabate, the cOMPIek'effeCtSof hueinan ewer-
the effect persisted for years and was ended . tations. Before'.illiiitrating,'these:'problems in
only' by the dissolution of the experimental set- more detail, I wilt .mention yet'. another prob-
ting. For example, ,in order to vary the il- lem with the bionietliadolOgies, This burden;
lumination, the experimenters moved some the most troubleioMeOeCaii§e itIstarids in the
employees to a separate room. Their findings way of our learningiisthat the biOmethodology-
indicate that the move to the separate room distorts the reporting:6f findings These human
had permanent effects whith were far more im- reactions are too Oft0 .vievi.od'iis "frailties" of
portant than'tlie changes in illumination. Thus, method, which 11 known will consign the
in humans,' experimental manipulations can research proje, to a second-class journal. .

create unintended expectations. When expec These pseudoscientific coicerns interfere with
tations are shared by a. face-to-face group, they the develOpriient of a' true methodology for
can be very, powerful indeed. As Norman it. F. behavioral science. A truly. scientific .

t Maier put it: "It was not an experimenter ef- behavioral' science must deal with these .

feet that was discovered at Hav horne; it was - frailties, not deny and ignore them.
the social system." One stiicly. has I been:. roundly critici;ed
Much of -the misunderstanding and over because seme members of the control group
simplification of 'Hawthorne is cleared up by got themselves. some therapy. Why can't they
Parsons (1974) in his concise, precise behaVe like corn stalks? The issue here is not
reanalysis. But even he seems to me to have' the ethical one of withholding treatment,
missed the punch line: His revelation that some although the ethical one is the controlling One-.
of the effeets could be explained by operant The issue is a scientific one. How can we
conditioning is useful but not surprising. assume that ,their desire for 'treatment, or the -.
Operant conditioning was indeedthe rationale lack .of it, does not affect our findings? In short,
 IMPLICATIONS 'POlt RESEARCH

the ethical, issue is requiring nsto be more, not expectations and resentments. When we coni--
less(as some seem to suppose) scientific. pare institutions and 'programs; we frequently
Por more than '30- -years I have observed cannot use random asSiknthent. In this case we
eValnatiOns,,Conducted them; talked informally-- should attempt to equate,pOpulations on Other
. with investigators. .and with their subjects. I measures, .hopefully, including expectations.
have ',found that, in many experiments there Caine and,.Small ,(196il.present an interesting
Were human reactions ,which could have had comparison -,of institutions:
'important effects on the outcome and which While comparisons of diagnosis, severity, ex-
were not reported. In some -cases, the human / 'peetations, etc., do not assure us of equivalence
, interaction's are viewed as "dirty linen"; it (Kraus, 1959); we must always be aware that
would'be "bid taste" to report them. In other ra4dom assignment also does not assure Us of
cases the experimenters themselves were equivalence, and that the latter has some
,traumatized, thrown out of the research set- disadvantages of its own, including not only
ting, sometimes scapegoated by competing placebo effects but ,also piobably a greater
groups, or by resentful admifiistrators. There selective loss of subjects. The question is not
are many too many of these experiments whether bias -can 4be eliminated; rather we
where the people 't want to -el..: bottt it" must decide hoW it can be minimized-
or have come to',d t ms were en- There are situations where the regular ad-
, countered. ministrative procedures !Jesuit in nearly' ran-
There are 'exce a tions of course. Joan Rit- dom assignment (without having an ex-
tenhouse's detailed report of experimental perimenter stirring the place with a stick). If,
troubles will be disc' ssed belbw. For now, let for example,.assigninent to wards is Chiefly.on
us illustrate the problems of the biometh: the basis, of empty beds, then comPariurrnitc..,
odology by considering two important issues: come of these wards may result in less biased
random' assignment and control groups. findings than would a controlled experiment.
Random assign_ ment 'is a most useful prin- In the 'Phipps Clinic; Where Whitehorn and
ciple. However, (frit. does not in itself assure Betz'. did their research, assignment to
equivalence in groups being compare js.9431tual- .1 . therapists was chiefly on tie basis of patient
ly, experimenters go beyond and compare their load: Occasionally, a therapist might ask for a
groups on, background factors. Grossman (1952). particular kind of patient because he had not
did this and found no significant differences. had an opportunity to work with that type for
But when he went still further and compared awhile. Such requests tended to be honored. It
his "groups on their expectations regarding would be helpful if the 'basis for assigning each
therapy he found significant differences. Such Patient were a matter of record.
expectations he, reasoned are far more impor- We 'turn new to' the second issue: the "con-
tant in the outcome of therapy than age, sex, trol Noup.". While the .problems of the bio-
and so forth. methodology undoubtedly extend to both ex-
Rarely .do investigators attempt to measure perimental and control groupS,It is particular-
patients' expectations, although soMetimes it ly with respect to the latter that the scanty
is done retiospectivelY. Sloane eta p.l. (1975) evidence available is most daMaging. In 1970,
-found some interesting explanations for treat- Fiske et' al,. (Donald Fiske, Howard Hunt,
ment failures by asking about expectations. In Lester Lubersky,Martin Orne, Morris Parloff,
brief, the expectations had been-vielated. Morton Reiser, atid .Hussitin Ttima, a' truly im-
When,random assignment is explained to the . pressive grotip)-Wrote: "It is impossible to Con-
patient, it may have negative placebo effects ceive of a true control groUP..." (1970, p. 24) In
by raising doubts,as to whether he is receiving view of the .nurnber'of investigators and critic's
'appropriate therapy "even though we tell him who are continuing to demand control groups,,
.

that we do not know which is *ore effective. this judgment seems. hardly, to have been no-
This last explanation itself can have negative ticed:It is my hope here to be more explicit and
placebo effects. detailed, and thus bring the issue to greater
We should always attempt random assign- discussion.
ment if it is possible to do it without creating 1 Some of my complaint has to .do with the

1U.
 EXPLORING THE PSYCHO-SOCIAL THERAPIES

name "control group." Frequently the com- Control groups in outpatient settings ate
parison -between the experimental group and plagued with the problem that controls seek
the "other" group is of interest; it is substan- help elsewhere. Even if these contacts are
tial and worthwhile. If we could accurately known there is '.a problem in defining which of
describe what, was done to both gtoups of-pa- the contacts are to be considered therapeutic,
tients, part of the probleM would go away. If a which controls ,are disqualified, and whether
.particular type of therapy is compared with the remaining controls are still repregentative.
"control groups" in two institutional settings, It is not my intent to suggest that coirtro).
one being a high staff-to-patient private groups are not possible in behavioral research.
hospital and the other a back ward in a State In some educational interventions, control
hoipital, it should be clear that the com- A groups have been used successfully. In mental
parisons will not "add up": the control groups health settings, a control, groups is feasible
are not equivalent, and the comparigons are where the clinician has advertised for his "pa-
not equivalent. We have no scientific reason tients," or where he seeks out his client
for calling these groups "control' groups." (Massimo and Shore, 1963). However, when the
However, when we give up the name, we are in patient has a presenting problem, it is doubtful
effect Comparing two or more treatments, as that he can be "controlled."
Fiske et al. make clear we should. Turning now to the therapists and how they
But this solution does not end our troubles. feel about experiments, we find ourselves in
It is' not uncommon in studies of inpatients to; the middle between two fervently held and
draw both the experimental patients and the conflicting faiths, the, one insisting upon ex-
"control" patients from the same wards. This is perimental control while tilt other insists upon
done in the name of, and proof for, "experimen- optimum care (see Colby, 11060). It is not s'ur-.
tal control." It is also not uncommon in such prising that some of the best experimental
studies for the "control" group to show a evaluations are those where the researcher \.
deterioration over the course of the experi -. and the clinician are combined in the same
ment (Buckey, Muench, anti Sjoberg, 1970; person. _ .

Kraus, 1959; Peyman, 1956; Spear, 1960). Of The conflicts rarely surface in print
these investigators,' only Kraus discusses the although they are frequently encountered i
possibility that control group patients feel re- informal dis ussion of projects. Sloane .et al
jected. It is certainly reasonable that a patient, (1975, p. 55- devote several paragraphs to
seeing others getting treatment that-4as not the conflict and th bruises, but do not indicate
given to him' , should feel that the hospital had that it acfeeted the esults in any way. One can-
given up on him. If alternate treatments are of- , not help but der whether "psychether7
fered, this rejection might be avoided. Still it apists" might not feel more 'bruised by a con-
would be useful to know how the treatments trolled experiment than..!`behavior therapiits."
are perceived. Gunderson, Schultz, and Feinsilver (1975) and
Control groups' can cause damage to ex- -Grinspoon, Ewalt, and Shatter (1972) touch on .

perimental groups which are on the same some problems in relationships between,.
ward. "One problem for which no satisfactory, researchers, therapists, and administrators. a(
solution was pound lay in the tendency of some By far the most'detailed description of the
patients in the psychotherapy. groups to listen difficulties in a field experiment is th of Rit-
attentively to some of the more intimate y.z,nhou. (1970). She provides a real-life
revelations of the other patients and to scurry description of the problems of establis ing an
back to the wards and broltdcast them to other controlling the independent variable, of ran-
patients not in the group. This at times led to dom assignment, of control groups, and of the
some embarrassment, and encouraged some expectations of therapists;' patients, and their
members of the two psychotherapy groups t families. In . some instances she is able to in-
deal only With superficial issues, or else to dicate how the problems might have damaged
become significantly less talkative during later the results. (She does not, however, provide
sessions" (Peyman, 1956, p. 39). the systematic treatment of patient expecta-

u
 IMPLICATIONS FOR RESEARCH . 89

tions which is found in-Sloane et 1975.) Con- These are the chief concerns about the
sidering the-difficulty of the project op which biomethodology:
Rittenhouse reports, all will agree*Oatit was 1 The blomethpdology assumes a body of
carried off very well. Moreover, the significant knowledge abouttwhich we seek to prove or
findings cannot possibly be explained in terms disprove certain tleviations. When no body of
of the methodologic difficulties. It is not my in- knowledge is available, the biomethodology
tention to summarize .either the project or its does not provide:a framework for exploration.
problems. However, soineisspeets of the study Seeking proof wlien we should be exploring is a
illustrate the need to minimize biases. waste of research resources.
The -rules of experimental design dictate 2. The biomethodology does not provide an
that one first define the population, and then adequate framework for understanding effects
randomly assign each member to* a treatment of human expectations which are uninten-
or control group. For human populations this tionally created by the experimentation. Such
usually means that one must determine not expectations, Of patients or of therapists, can
Only who is eligible, but also who is willing to have effects 'which are both powerful and
he assigned to the experimental treatment. Ac- lasting. .1 .
cordingly, it-was explained to each patient and 3. The +methodology discourages the
his family that he might be assigned either to a honest repoifting of difficulties encountered in
hospital treatment _team, or to a team con- conduct' of experiments. It thereby con-
ducting family therapy at home. To prevent ceals the e ects of human expectations, andit
selective losses, they were told that if they did prevents o r learning how to deal with these
not accept the assigned treatment, they would adifficulti0." It discourages the development
not be admitted to the other treatment. of a truly scientific social science.
The researchers encountered resentments 4. A "control group" is generally desirable,
from patients and families who wanted the but it m not be. possible or desirable in the
other treatment, and a number of losses for the evaluati of the psycho-social therapies.
same reason, and for the reason that the clini- a. If ere is a presenting problem, and
cians asserted their right to transfer patients the e usually is, it may be difficult to
between the treatments for clinical reasons. prevent_ patients from getting treatment.
Not all of the losses and resentments could be I ;these days of pastoral counselors (to
avoided in any case. But if the random assign- sa nothing of bartenders), cassette
ment had been made on admission, and .pa- courses, etc., how does the experimenter
bents told only of thek assigned treatment, prevent the patient from getting help?
some false expectations and negative placebo How do we evaluate whether he has got-
effects would have been avoided, and the ten lielKwhether he is still a control?
biases might have been less than in the un- b. More important scientifically, our at:,
critical conformity.to the biomethodology. This tempts to control the subjects Mind us
is presented as an illustration of a research to the fact that the patient's expectations
choice where biases will accrue in any case and about treatment, and how they are met,
where the task is to choose that method which are critical variables in the outcome. We
will minimize bias. will never understand the psycho-social _
Had 'the experimental program been treated therapies so long as we ignore these
as a "new service," which it was, instead of an expectations.
"experiment," further expectations and c. Contrbl groups can create negative placebo
resentments might' have been avoided. (The effects that are not adequately, handled
number of natural experiments which we have in most current evaluation deSigns.
failed to evaluate is legion.) Finally, it is to be 1:1) The most frequent. design is actually a
noted that the procedure used is quite "ar- comparison of an experimental treatment
tificial." It introduced forces into treatment. with the "usual" treatment called "con-
that would not be present in a setting'where trol group." The term "control group"
family,'home treatment is a standard pro- blinds us to the fact that the "usual" treat-
cedure. ment is quite different in different set-

06'
 EXPLORING THE PSY&10-SOCIAL tHERAPIES .5

tings. Such comparisons nnot be ex 2. All measurements either preceded the

peeled to "add up" acr s studies. . therapy or followed after the therapy had
e. Given the limilations of our present ended. The measurements themselves
knowledge, the Scientific yield would probably did not affect the therapy. This
be /greater if, instead, of comparing an is discussed further below.
erimental and a control group, we
ere to compare twok psycho-social Because of long experience with evaluatio*
.

therapies, provided the two therapies studies, I have been sensitized to these issues.
are defined sufficiently precisely (e.g., I wondered, for example, whether the Phipps
therapists' personalities) that comparison therapists would not discover 'that they had
across studies .is facilitated. It is not been labeled "A" or "B" and whether this
enough to study "psychoanalysis" or might produce resentments. I have been
"behavior therapies." unable to find such resentments, although I.
5; Certain scientific principles inherent, in have talked with a number of persons who par-
the controlled eXperiment must not be ticilAted in this research. By contrast, I have
"thrown Out with the bath water." Never- encountered resentments in several ex-
theless, it is probably not true that a principle periinents. A number of techniques have been
like rand-om assignment can best be achieved used to bar my from contact with persons who
always in a controlled experiment Under cer- participated in some ex,perimental studies., In
tain conditions it might better 'be achieved in a some cases WhWeThid-Contact, given:the
natural experiment. text of our discussions, persons who par-
6. The natural disaffection and distrust be-. ticipated should have volunteered their par-
ween researchers and clinicians are so great ticipation, and they did not. In other cases the
we should come to expect that tliose who con- complaints were 'openly expressed. I cannot
uct controlled experiments in clinical settings emphasize too much that these concerns are
will give us detailed.meaSurements of the clini- important.
cians' attitudes toward the experiment.
7. There are no absolute answers in the
choice of methodologic tools, the textbooks not- Specific Issues
withstanding. We can hope only to minimize
biases, not eliminate them. In existing studies, This section is concerned primarily with the.
the various kinds of biases have usually not specific implieations of the present studies for
been measured and/or reported. Among these research in psycitotherapy..Appendix '1 also
biases are those in the selection of patients and discusses a number Of specific'' issues of
therapists, and those that result from ex- . statistics and psychometrics.

perimenteManip.ulations. It is indeed likely First, with respect to psychotherapy, there

that the "tighter" the experiment, and the is a clear need to 'define researh poPula-
more it has intruded upon the clinical setting, tiOns-- both of clients and of 'therapists. With
the greater are the biases of selection and respect to clientsparticularly institutionalized
manipulation. elientSodiagnosis is critical. In the Loysville
Obviously, the studies reported in this data there was an OPOortunitylo try a variety
monograph do not proye, or even illustrate all ot. approaches. For example, a number Of ef-
of these propositions. I mean only to suggest forts "were made to combine diagnOStic groups
that the- simple methods of Whitehorn and Betz in order to achieve greater numbers of cases.'
avoided many of these problems in that: Without eXception such coitIbinationsresulted,
in "lost findings." There was nothing in these
e ' analyses to suggest that the six diagnostic'
1. All measurements 4- nd "manipulations" , groups could be reduced in number, and in-
were part of the esta shed clinical routine. deed, the ultimate reduction "all clients"
They were not part of an experiment. 'No 'resulted in practically,no. findings at all:
expectations were created for, either There are other aspects of the Loysville
therapists or patients. diagnoses which are of interest. For' the most
U
 IMPLICATIONS FOR ,RESEARCH 91

part they were made by persona.with no train- \ (mimeob report that the contribution of the
ing in the diagnostic system. While there is no therapist is critical even when he is not sup,
doubt that diagnoses could have been im- posed to be doing' psychotherapy,- when he is
proved by suChAzaining, there is little support supposed to be merely dispensingdrugs7 If the
here for the .position that diagnoses .must be "therapist variable" is as important as it ap-
made by skilled experts. In fact,, it is quite pears to be, then we have no huSiness doing
possible that overly skilled experts- may pro-, pSychotherapy research using only a . few
duce consistent but nonreproducible findings. therapists,. and we must somehow define the
More important than expertness for research therapists such that some kinds .of statistical
purposes is consensus. controls can 'be utilized to define our indepen-
Because of the need to form subgroups of dent variables (aipects of therapy),
clients homogeneous as to their problems, The need to -define the population of
large pool of clients is needed. The Phipps data therapists can be shown in several ways. As
redts on at least 1,300 patients. In the Loysville between .hoSpital and nonhospital psychi-
study we started with nearly 600. The tempta- atrists, the "A713 Scale" changes its meaning.
tion in clinical research is to pick out a couple Presumably, treatment ideologies are not in-
of major groups and combine the smaller dependent 'of other aspects of personality: In
groups. For example, among institutionaliZed any event, the "A -B Scale" cannot be expected
patients, one frequently finds a number of to be valid in most hospital settings. We need.a
schizophrenics and depressives, and only a few better definition gf treatmentorientatiw-bet--
. manics, or neurotics, or personality disorders. ter than hospitarvs. nonhospital. The issue is
The fact that there are only a few of these lat- discussed further in appendix 8.
ter does, not justify, lumping them together. In In the therapy;analogue studies we are fre- .
a number of clinical studies neurotics and per- qUently told that the "pseudo-therapists were
sonality disorders have been cOmbbined. Some from undergraduate psychology classes." If, as
findings presented here suggest that when suggested 'by the present findings, the "A-13
such combinations are made, positiVe correla- Scale" Changes Its meaning between
tions may cancel out negative correlations and psychology majors and premedical students
all is. lost. (there are undoubtedly existing' data to test
But we should not" give up hope about this hypothesis further), we, need much more
numbers of oases. Sometimes, we ask for too information about who the pseudo-therapists
much in this regard. For example, in the are and how they are recruited. Do they know
Phipps data, unless a therapist had at leas) that it' is a study of psychotherapy ?' Do they
four., patients of the designated diagnosis, his volunteer? Do they really. volunteer, or is-the
success rate was not computed and he was not semester ending and they need the lab credit?
Included in analyses. In the Loysville data, I ex: Is it an advanced- class in personality, or a
peripented with a variety of forMats begin- beginning class in physiological psychology? It
ning with as minimum of five per therapist and is likely, that, with respect to the psycho-social
running on down to one per therapist. The area, psychology researchers have not been .
number of, therapists that can be included in- well served by, a too-ready access to human
creases, of course, but more important there organisms assumed to be representative of: all
was a steady increase in the number Of signifi- human organisms. While the present findings
Cant , findings as the number of clints per suggest that psychology majors will give us
therapist' decreased. The statistical theory is more understanding of the "A-B Scale" than
considered in appendix 1. It is sufficient to note premedical students will, still the flight. from
here that; while we need a large pool of clients psychology major to psychotherapist is
to begin with, there, is much to be learned by probably too great.
including data from comparatively small In therapy studies. and in therapy analogue
.subcells. studies, it is essential that the researcher think
Turning now to the therapists, "there is an about and try to define how his population' of
. equal if not greater need to define our research therapists might differ from a more gener-
populations than for clients. Turns et al. alized population of therapists. In this connec-
 EXPLORING THE PSYCHO- SOCIAL THERAPIES

we probably need, a great, deal more snakes," "silde-ishoW : ifreaks,," "DCIVIC,"

understanding of such, aggregations as "geography "). Moreover ;,_ many of the ceitral
"p§ychoanalyst;" "behavior therapist," etc. variables are na adequately measured (e3g.,
than we now have.. empa,thic interest, tolerance for ambivalence)
Variations in response rates among various a thou h their test-retest': reliability may be
portions of a population ay reveal that the uch igher .than their , internal consistency.
,population is really not hiomniogeneoui. In one n th one hand; we ntiusenot be carried away'
study reported here, it was such variation y su 'h provoking loose 'ends, some of which
which suggested the differences between ay be dead ends. We Must view them with
hospital and nonhospital therapists. aution without disparaging them, thus
While I have no, evidence, I have discussed lemonstrating our own low need for closure.
the possible differences, not just in outcome, It is useful to take yet another look)it the
but also in observed ,relationships among .Phipps Clinic data, not so.much from the stand-
variables, between experienced and inex- point of what, we have learned.from it as from
perienced therapist§:this issue deserves .con- what we have yet to learni-The purpose of this
-,,siderable attention'. '',\ exercise is not so much to; define a research
Although the reversaY the meaning of. the prOject as it is to illustrate, significant con-
"A-B- kale" for various populations ithe / siderations in any naturalistic Clinical 're:
semantic reversal) helps -to 'explain the con- i search. r . 4
flicting findings in the litera tire; it actually --,--,1t- is likely-that-the-great-Atrehgth-of-the-
raises a number of que§tio for clinical . Phipps data is in the recording. of outcome,
research and for personality research in- systematically, over many doktides. Some have
general. It makes sense that pers'Ons with high complained about the subjective nature of the
'empathic interest would- view the' Manual oc judgment "improved" even ;.though consider-
cupations in a quite different Context from able objective data were available to those
those with low empathic interest. But what are making, the judgment'. .There is also concern'
these contexts, and why does their view of over the fact that .the therapist himself par-
theseoccupations seem to have opposite nian ticipated in the judgment. ,

ings? How frequent in persohality research are There are available, at 'the Phipps Clinic,
such reversals and whit implications do they, nurses' behavior charts. What findings would
have for mathematiCal models like factor el:merge if criteria derived from these charts
analysis which, in effect, bury them? Were used instead of the global improvement
The functional reversal of the "A-B Scale" Measure? More-over, it is lilcely that these
also remains mysterious in spite of new. light charts could yield nuances of i 'provement not
shed upon it. The various possibilities were available, from the overall j dgment..fir.,the
outlined in the last chapter and will not 'be later years, 'for example, // depressives
repeated here. improved.
Finally, with respect to' therapists, "the It would be very valuable tO compote ther-
therapist variable" is not a variable. It is a col- apists'--succesi rates with finer 'Classifications
lection of:v.ariables, and the collection probably of clients: (1) male vs. female, within diagnostic
varies as we pass from one diagnostic group to groups, (2) for schizophrenics, process vs. reac-
another. There is good reason to believe that tives (Betz, 1963), (3) for deptesfves, neurotic
empirically derived predictots of success vs.. psychotic, and so. forth. Can anything be
should not be homogeneous or internally con- learned &mt. the relatively few manics and
sistent. Each of the diagnostic groups for personality disorders that were identified?
which we have adequate data turns up more More attention needs to be paid to combina-
than one successifactor; and there are probably tions of therapies: psychotherapy, insulin,
many more, if only these additional factors had ECT, and drugs (e.g., Whitehorn rand Betv
been included in our' instruments. 1957). There may be other resources like tter"
As is always' the case with exploratory Phipps Clinic that are ripe for exploration 7r;:
research, there are many thought-provoking However, unless they contain systematic d-ata
loose ends in the chapters above (e.g., on outcome, they ar% unlikely to be useful.,.
I
.
 .93
IMPLICATIONS'nFOR RESEARCH

The presen research leaves unanswered the and sociopaths ciii.the courts: The dependent
question of hether the personality of the `variable is, therefore, in interaction with a con--
therapist is imply a convenient, inexpensive trol variable:
approach to hat is important in therapeutic .Finally, there is a further complication that
processes, o whether understanding the per- is almost universally ignor4d, although it
sonality of the therapist is essential to would be easy to present anecdotal evidence
understanding therapeutic processes.' Are that it should not b&, namely: he' impact of the .
..there subli inal communications which are dif- dependent variable on the indeWndent ones: It
ficult forte researcher ,to observe? In one is not uncommon in evaluation studieS to find
study tha did include both personality that the helpers are adapting their interven-
variables a d behavioral ones, the personality tions to the criteria of evaluation. Teachers
variables appear to be more potent (Alexander ...teach things that are likely to be on achieve-
et al., 197). This issue may be critical to ment tests. If in a large experiment it is e
research strategies in the psycho-soCial cided that Symptoms are the major criteri of
therapies. Change, homed° we know that this does no con,
One issue has been avoided in the present centrate the therapists' attention on sy ton's.
research. By using "improvement" as the and their. reduction, rather than on under-
dependent variable, we have not dealt with the standing the problems they deal wit . f; in-
problem that ..various diagnoses may require deed, this were the case, then the evaluation
different specific outcome measures. design itself would contribute to deterioraion
Moreover, different vantage points may be in; among schizophrenics. We must watch,
volved: while neurotics:_ usually presen0,-,, therefore, for' interactions among the depen-'
themselves foi treatment,' psychotics are fre- dent, independent, and control variableS.
quently brought to treatment by their families,
:

r.
I

a
 Chapter VIII
Summary
4
These studies began with the relatively sim- By definition, a milieu cannot adapt to dif-
ple, naturalistic methodology used by ferent kinds of patients. And indeed, there is
Whitehorn: and- Betz in their studies of some evidence that- opposite kinds of milieus
psychotherapy effective with schizophrenics. are neesled for some' kinds of disorders (e :g.,
When extended to a variety of disorders,. the schizophrenics vs: sociopaths). It seems 4ikely
methods yield meaningful,consistent, and that milieus will have to specialize in particular
replicable findings. Moreover, we have extend- disorders.
ed this methodology to milieu therapy. The There are many specific findings in, these
personality characteristics of effective studies. They range from some which are
`therapists can be used to define therapeutic replicated to others which are best regarded as
relationships and the resulting findings yield good hypotheses for further research. AmOng
implications for,; what it is that, effective the replicated findings is onethat neurotics are
therapists do: not well served by therapists, who like to solve
However, there- is little that defines the ef- problems. Evidence is presented that 'this
feetive therapist in a general sense in these finding supports Freud's contention that the
studies. Instead, specific personal, tendencies 'neurotic must solve his own problems. The
and interests define effectiveness with par- therapist should, do little more than ask ap-
ticular types of mental or behavioral problems. propriate, not too leading, questions.
There is therefore no support"for the "gener- This replication is found for two very dif-.
alist approach," but there is a great deal of skip- ferent groups of neurotics served by two very
.
port for the differential hypothesis that dif- different groups of therapists. One group is of
ferent disorders require different treatments. severe neurotic adults' -hospitalized at the
(The present studies do not include a measure Phipps Clinic of the Johns Hopkins Hospital
. of therapists' personal adjustment. It is pos- and served by' highly selected psychiatric
sible that personal adjustment is a general residents. The other group is cj neurotic delin-
'trait of effective 'therapists.) quents in a Youth Development Center at
If future studies continue to confirm the dif- Loysville, Pennsylvania, and served by staff
ferential hypothesis, the implications of the members with considerable'experience, but for
finding are widespread. For one thing, the most part without even a college education.
diagnosis is a critical variable. Evaluation of One the one hand, this replication emphasizes
services. is impossible- unless consideration is that personality factors in the therapist trans-
given to. the type of problem the client cend his education and the. setting in which he
presents. works. On the other hadd, the -replication em-
As we learn more and more about the phasizes anew the large overlap between
specific needs,of particular kinds of clients, we "mental health services" and "corrections."
mat also learn the extent to which therapists A second replicated finding suggeits that
can and will adapt .to different kinds of clients. schizophrenics are best served by a therapist
If therapists adapt, there is little probleni If who is active and personally involved. with the
they do not, then we must understand how patient. It is..possible that this active involve-
clients get to theraPists and whether theSe ment is not so important when drugs are
paths are optimal. prescribed. But regardless of whether. drugs

94
70-

1,
I Q'
 c, a,
SUMMARY 95
-r.
are prescribed, sclizophrenics must be treated stated another way, if an item can be inter-
with tolerance and understanding. These find- preted in a wide variety.of contexts, it may not
ings are consistent with the original find- yield findings which are reproducible over dif-.
ingkof Whitehorn. and Betz. ferent groups of therapists. Specifically, the
Otker detailed findings of a substantive Whitehorn-Betz "A-B Scale" is found to
nature are .presented in chapter VI. reverse its meaning as 13etween hospital and
The specificity of the findings suggests that nonhogpital psychiatOsts. It is not surprising,
the generalized techniques which are lauded then, that the "A-B Scale's has failed to produce
and taught are not all that general; but rat er completely consistent findings across all the
specific to specific problems. The proscripti n studies in which it has been used. It will not
of problem solving appears to be specific to th yield consistent' findings unless. certain
.

neuroses. We need to define specific aspects of characteristics of .the therapists are controlled
the psychO-soCia therapies, 'and the specific in selection or analysis.
uses of these aspects. The studies underline . Thus the studies emphasiz the need to
the need for comparative studies .of psycho- define the population being studied, both of pa-
.social approaches to .various inentar disorders. tients and of therapists. Populations must be
In the measurement of personality, the defined, methods of sampling must be made
studies indicate that if a personality scale con- clear, and the proportion of the sampled
tains items. of which the Meaning is not population that.actuatfiparticipated should be
manifest, the .items -and -the -scale may not have reported.
the same meaning for cliffeient subjects. Or,

4,

.f
 References _J

Adorn°, T.W.; 'Frenkel-Br nswik, E.; Levinson, J Effective Psychbtherapy with the Schizo..
_ D.J.; and. Sanford,, -R.N. he Authoritarian Per- phienic Patient. New York: Aronson,: 1975,
sonality. New York: Harp 1950. Chapter 1, 3-21.
Alexander, James F.; Barto Cole; Schiavo, R. Betz, Barbara J. A study oftactics for resolving the
Steven; and Parsons, Br ce* V. Systems- autistic, 'barrier in the psychotherapy of- the-,
behaVioral irtterventiOn with milies delin schizophrenic personality: American Journal'of
quents: Therapist characteristics, family Psychiatry. 104:267-273, 1947; RePrinted to
behavior, and outcome. Journal of Consulting Whitehorn, John C., and -Betz,. Barbara J..
and Clinical Psychology, 44:§56-664, 1976. Effective Psychothenspy with the Schizophrenic
Asch, Solomon E. Opinipns and social pressure, Patient, NeW York: Aronson, 1975, Chapter 2,
Scientific American, 193:31-35, 1955:' 22-37.
Barron, F. Some personality correlates of Indepen- Betz, -BarbaraJ. Experiences in research in
dence of Judgement. Journal of Personality,
-, psycho- therapy with -schizophrenic patients.
-

21:287-297, 1953. .
In Strupp,H'ans H., and Luborsky, Lester (Eds.)
Bednar, Richard L, and. Mobley, Max J. A-Erthera- 'Research in Psychotherapy, Vol. 11, Washington:
431st perceptions and preferences for American Psychological Association; 1962,
schizophrenic and pkychoneurotic clients. Jour- 41-60. .

nal of Abnormal Psychology, 18:192-197, 1971. Betz,: Barbara J. Differential success rates of
Benedek, Therese. Dynamics of the counter-trans:- psychotherapists with 'process'` and 'non-
ference. Bulletin of the Menninger. Clinic, process' schizophYenic patients. American
17:201-208, 1953. Journal of Psychiatry, 119:1090-1091, 1963.
Bergin, Allen E. The effects of psychotherapy: Reprinted in Whitehorn, John C., and Betz,
.
Negative results revisited. Journal of Counsel- ?Barbara J. Effective Psyctotherapy with the.
ing Psychology, 10:244-255, 1963. Schizophrenic Patient. Ntw York:, Aronson,
Bergin, Allen E., and Suinn, Richard M. Individual .1975; Chapter 15, 244-245.
psychotherapy and behavior therapy. Annual Re- Betz, Barbara J. The problem solving approach
view of Psychology, 26f509.555, 1975. . and therapeutic effectiveness. American Jour-
Berzins, Juris I. Matching patients with therapists: nal of Psychotherapy, 20:45.56, 1966.
Conceptual, empirical, and pragmatic persp Betz, Barbara J. Studies of the therapist's role
tives. Paper presented at the 5th annual meetin in the treatment of the schizophrenic patient.
of the Society for Psychotherapy, Research, American Journal of Psychiatry, 123:963-971,
Denver, Colorado. June 1974. 1967. ,
Berzins, Juris I. Therapist-patient matching: Betz, .Barbara J. A-B therapist variable: 'Com-
Gurman, ASS., and Razin, A.M. (Eds.) The Ther- mentary on 'Bowden's paper. The Journal of
apist's Contribution to Effective psychotherapy: Nervous and Mental Disease, 154:287, 1972.
An Empirical Assessment. New York:Pergampn, Beutler, Larry E. Psychotherapy: When what works
in press. with whom. Paper presented at the 7th annual
Berzins, Juris I.; Bednar, Richard L.; and Severy, meeting of the -Society for Psychotherapy Re-
Lawrence J. The problem of intersource con- search, San Diego, June 1976.
sensus in measuring therapeutic outcomes: .
Bowden, Charles L.; .Endicott,Jean; and Spitzer,
New data and multivariate perspectives. Jour- Robert L. A-B therapist variable and pbyc.ho-
nal of Abnormal Psychology, 84:10.19, 1975. therapeutic outcome. Journal ,of Nervous and
'Berzins, Juris I.; Ross, Wesley F.; and Friedman, Mental Disease, 154:276-286,1972.
William H. A-B therapist distinction, patient Buckey, Harold M.:. Muench, George. A.; and
diagnosis, and outcome of brief psychotherapy Sjoberg, Bernard M., Effects of a college
in a college clinic. Journal of Consulting and visitation program on a' group of chronic
Clinical Psychology, 38:231-237, 1972. schizophrenics. Journal of -Abnormal _Psycho,-
Betz, Barbara J. Somatology of the schizophrenic -1 oqy, 75:242-244, 19-70.
patient. Human Biology, 14:21-47, 192.234, 1942. Burnham, Donald L.; Gladstone, Arthur I.; and
Betz, Barbara J. Experiences in the psychother- .
Gibson, Robert W. Scphrenia and the Need-
apy of obsessive-schizophrenic personalities, Fear Dilemma. New 'Y` rk: International Uni-
r
.

Southern Medical Journal, 39:249-257, 1946. Re- versities Press, 1969.

printed in Whitehorn, John C., and, Betz, Barbara Caine, T.M. : and Small, D.J. The Treatment of
,
 REFERENCES 97

ness of psychotherapeutic: programs: A multi-

.

Mental illness: Science, Faith and the Thera-

peutic Personality. New York: international Uni- criteria comparison of .four programs, for three
versities Press, 1969. different patient groups. Psychological Mono-
Campbell, David P.; Stephens, Joseph H.; Uhien-- graphs: General and Applied,. 74:1 -26, 1960.
huth, E.H.; and Johansson, C.B. An extension Whole No. 492. 4 _

of the Whitehorn-Betz A-I3 Scale; Journal of Fiske, D:W.; Hunt, .1-1.R; Luborsky; L; Orne, M.T.;
. pervous,and Mental Disease, 146:417.421,
1968. Parloff, M.B.; Reiser, M.F.; and Tuma,- A.H.
Campbell, Donald T. Research in psyChotherapy, Planning of research on. 'effectiveness of psy-.
A.P.A. Monitor, 7(1):p 2, 1976. chotherapy. American Psychologist, 25:727-737;
Carson,R.C. A and B therapist "types": A possible 1970. Also Archives of General Psychiatry,
critical variable in psychotherapy. The Jobrnal 22:22-32, 1970.
of Nervous and Mental Disease, 144:47-54, 1967. Fox, Harvey, and Lefkowitz, Joel. Differential
Chartier, George M. A-B therapist variable: Real validity: Ethnic group as a ,moderator in pre:
or imagined?. Psychological Bulletin, 75:22 -33; 1.MM% 'dieting job performance. Personnel Peychology,..-
11371.
27:209-2230 974: . .

Chartier, George M., and Weiss, Lillie. A-B Thera- Freud, Sigmund. On the 'beginning of treatment.
.pists and clinical perception: Support for 'Collected Papais,.'ll; 'New York: Basic Books,
'"Super- A" hypothesis.. Journal of. Consulting . 1959, p.342-365,4Iso in. Strachey, James (Ed.)
and Clinical PsycVlogy, 42:312, 1974.0 The Standard' Edition of the 'Complete Psy-
Cheney, Warren I/ peurosis-spedific- psycho- cpologital Works of Sigmund Freud,: Vol. .12,
'analytic techniques and personality 'theory 'of . London: Hogarth, 1958; p. 123-144Noriginally 4.1=4!
Frit Riemann. The AmeriCan JOurnalof Psy- --published_191.3).,
cho nolysfs; 26:20-28, 1966 Fromm-Reichmann, 'Frieda. PrincipleoThif nsIve
;. esthetic Psychotherapy. 'Chicago: ;University of Ci cago
I.L. Personality correlates k
ju ement in college students.. Journal of Prets;1950.
.Personality,.33:476-511,1965.. ° Garfield, Sol .L., and Bergin,. Allen' E. ersonal
Child, I.L.; Cooperman, K.; and Wolowitz; H.M. therapy, outcome, and some therapitt v riables.
Esthetic preference and 'Whisk correlates of Psychotherapy: Theory, Researckan Practice,
active versus passive food preference. Jour - 8:251-253, 1971a.
nal of Personality and SOcial Psychology,11:. Garfield, Sol L., .and Bergin, Allen E. Therapeutic
75-84, 1969. .
conditions and outcome. Journal of Abnormal
Psychology, 71:108-114, 1971b.
Chc:loff, Paul.. The depressive persenality: A Gerard, Roy, et al. Differential Treatment.
.critical review. ;international Journal of Psy-
chiatry, 11:196-217, 1973. gantown, W.Va.: Robert F. Kennedy Youth
Colby, 'Kenneth/Mark. An Introduction to Psy- Center, 1969.
choanalytic .Research. New.York: Basic Books, 'Gilbert, Doris C., and Levinson, Daniel J. Ideology,
personality and institutional policy in the men-
1960:
Dent, .jarnefis-K. A Bibllographid)ndex of Evalua- tal hospital. Journal of Abnormal and Social
.:tion in' :Marital Health, P.N.S. Publication No. Psychology, 53:263-71, 1956. ReVised version in
. 1545. Bethesda: N.I.H. 1966.
Greenblatt, et al. The Patient and the Mental
Kwiatkowska, H.Y. Aesthetic Pref- Hospital, Glencoe, Free Press,-1957, 20-35.
Dent, Goldstein,, Arnold. P., and Stein, Norman. PreScrip-
erences of young adults for pictures drawn
.by mental patients 'and by members of their tive Psychotherapies. New 'York: Pergamon,
.

immediate fimilies. Sciences d Art 7-Sci- 1976.

entific Aesthetics, 7:43-54, 1970. .
Gomes, Beverly, and' Armatrong, Stephen H. De-
Draper, Franklin. M. The dcictor's personality and terioration effects in piychotherapy: Once more
social recovery of schizophrenics. Archives : revisited. Paper presented at the 7th' annual
meeting' of, the Society for Psychotherapy Re-
of General Psychlabv, 16:633-639, 1967. searcK San Diego, California, June. 1976.
,Dublin, James .E.; Elton, Charles F.; and Berzins,
Goodman,. L., and, Kruskal, W. Measures of as-
Juris I. SOme personality and aptitudinal cor- sociation' for cross-classification, I. Journal
relateS of the "A-B" therapist' scale. Jour of American Statistical Association, 49:732764,
nal .Of Cons6iting and Clinical -Psychology, .
33:739-745, 1969. 1954.
Erikson, Robert V. LoysvilleA six year evalua- Greenblatt, Milton; Levinson, Daniel J.:. and
tion. Loysville, Pa.: Youth Development Center, Williams, Richard H. The Patient and the Men-
tal Hospital. Glencoe, Ill.: Free Press, 1957.
(mimeo) .

Grinspoon, Lester; Ewalt, Jack' R.; and Shader,

Eysenck, Hans J. The effects of psychotherapy;
an evaluation. journal of ulting Psychol- Richard I. Schizophrenia: Pharmacotherapy and
Psychotherapy./Baltimpre: Williams & Wilkins,
ogy, 16:319.324, 1952.
Fairweather,' G.W.; Simon, Cebhard, M.E.; 1972.
; Sanders,. R.;
Grossman, David. An experimental investigation
Weingarten, E.; Holland,.
Stone, G.B.; and Reahl, J.E. 'Relative' effective
 98 EXPLORING THE PSYCHO - SOCIAL. THERAPIES

/ of a psychotherapeutic technique. ,Journal of therapy' and Behavior Change. New'York: Wiley, .

Consulting Psychology, 16:325-331, 1952. 1971, 36-74. , .

Gunderson, John G.; Schulz, Clarence G.; and Kish, Leslie: 'Survey, Sampling. New ,York: WileY,
Feinsilver, David B. Matching theripists with 1965. , .

schizophrenic patients. in: Gunderson, John Klerman, Gerald L. The relationships between per
and Mosher, Loren R. (Eds.) Psychotherapy sonalitif and 'the clinical ,depression$:
of Schizophrenia. Newc.York: Aronson, 1975; coming the obstacles 'to verifying ..pSych&
343-359. dynamic theories.. International Journal, of Psy-
Hasenf us, Nancy; Martindale, Colin; and Kaplan, chiatry,' 11:227-233.; 1973. .

Cynthia. Alpha amplitude and hemispheric KlugmanSamuel F. Comparison of total Interest

asymetry in A and .B type clinical and experi- profiles,. of a psychotic and a normal group. ,-
mental psychologists. Paper ',presented at the na.1. of Counseling psychology, 7:283-.280,
6th annual meeting of the Society for Psy-
chotherapy Research, Boeton, Mass., June Kn ee; GAievieve; Jacksoh,, Doh D..; Krieger,
1975. tW
_Gorge. Perdonality differenCes between more
Hays, William L. Statistics. New York: Hoff. 1963: endless'oortipetent psychotherapists as'a func .

Henry, William E.; Sims, John H.; and Spray, S. Lee. tion crit ia of competence. Journal 'of
.The Fifth. Profession. San Francisco: Jossey- NervOLis and ntal Disease,.129:375-384, 1959.
Bass, 1971. Kobayashi, S efUrn1; .Mizushima,, Kelichi; and
Hoilingshead, A.B., and Redlich,.F.C. Social Class Shinohara, utsuharu. Clinical groupings of
and MentarllineSs: A Community Study: New' problem' c ren based on symptorna and
York: Wiley, 1958. behavior.. at n I of Sobial Psy-
Holt, R.R, and Luborsky,' L. PeriOnality Patterns chiatry,- 13:206'. 7196 _

of Psychiatrists. New York:-Basic Books, 1958. Kraus; 'Anthony R. Experimental study of .

Horwitz, (.. 'Clinical Prediction in Psychotherapy. effect of group :psychotherapy with chronic
New York: Aronson, 1974. psychotic. patients. 'international Journk of .
ingrarn, Gilbert L. Matching volunteers andAmve Group,Psychotherapy, 9:293-302, 1959:
nile offenders. Paper presented at a workshop Kullback; S. Mforrnation Theory and. Stafisfics,
at the University of Arkansas "Institute Awith New York: Dover, 1968. .

Volunteers in Juvenile Delinquency," May ir970. Lambert, MichaeLJ.; Bergin, Allen E.; and Coti.ins,,.
Jenkins, R.L. Diagnosis, dynamics and treatMent John L. Therapist-induced deterioration in psy
in child psychiatry. Psychiatric Research Flepfirt
.

chotherapy. In:' Gurman, A.S., and_ Razin, A.M.

18. American Psychiatric Association, 91;120;1 - (Ed s.)_The Therapis0 Contributions tol'ffective
1964: Treatment.. An. Empirical_ -AsSessMeht,;'1,New
Jenkins, R.., and Hewitt, I.E..-Types of personality - Ydrk:Pergamon; in dread.' .
structure encountered in child guidance clinicS. Lambert, MichaaJ, and De Julio, SteVen S.Ther,
American Journal of OrthopSychiatry, 14:84-94, apist interpersonal Process, outcome,,
1944. methodological considerations and recOm,
Jung, Carl G. The state of p6i9hotherapy today. mendationS for. fulure research. Paper read :at .'
Collected Works, Vol. 10, -CIV)11zation in Trans!. .
the Seventh Annual Meeting of: the Society for.,
Hon. Princeton University Prees, pp. 157-173, PsybhOtherapy Research", San Diego, June 1976.
.
1964 (originally published in 1934). .
Lewin, Kurt. The conflict between Aristotelian
Katz, Daniel. Pattern,s of 'leadership. In: Knutson, and Galileian modes of thOught-in contemporary
,
Jeanne N. (Ed.) Handbook of PolitiCal Psycho!: psychology:.. Journal of Gerieral Psychology,
ogy. San Francisco: Jossey-Bass, 1973, 203-233. 5:141-177, '1931. Reprinted in Lewin,' Kurt.
Kemp, David E. Personality and behavior in psy- A 'Dynamic' Theory of Personality. New. York:
chotherapeutic relatipnships: Correlates of a McGraw-Hill, 1935; Chapter 1, 1-42.
scale of therapeutic effectiveness. Unpublished Lionells., Marylou J. Themes' of therapy: The
doctoral 'dissertation, Duke University, 1963.. relationship of personality and ideology to
Kemp, David E., and Stephens, Joseph H. Which therapeutic style. Doctoral :dissertation, The
AB Scale? A comparative analysis of several University of Chicago, 1967.
versions. Journal of Nervous and Mental Dis- Luborsky, Lester; -Singer, Barton; and Luborsky;
ease, 152:23-30, 1971. Lise. Comparative studies of psychotherapies.
Kendall, M:G. Rank Correlation Methods, 2nd Ed. Archives of General Psychiatry, 32:995-1008,
London: Griffin, 1955. 1975.
Kerlinger, Peed, and Rokeach, Milton. The factorial Manosevitz, Martin. Item analyses of the MMPI Mf
nature of the' F and D Scales. Journal of Per- scale using homosexual and heterosexual
sonality and Social Psychology, 4:391 - 399,196.6: males. Journal of Counseling and Clinical
Kiesler, Donald J. ExPerimental designs in Psychology, 35:395-399, 1970. '
psychotherapy research. In: Bergin, Allen E., Massimo, Joseph and Shore, Milton F. A
and Garfield, Sol L. (Eds.) Handbook of Psycho att comprehensive, vocationally oriented psycho-
.

i Itt'y
 ..99.
REFERENCES
. .

therapeutic' program for delinquent boys. The Quay, Herbert C., and- parsoria; to. I B. The Dif-
American Journal ,OffiDrthopSychiatry, 33:634- ferential. Behavior ClassificaAidn of thejetveriile
642, 1963. Offender, j2nd ed. Morgantawn; W.Va. U.S.
May, Philip Treatment of Schizophrenia: Depart-het-It of JUsjice, Bureau :of Prisons,
a Comparative Study of Five Treatment Meth-. Robert F. Kennedy Youth Center, September
1971 (mimeo). -.
ods. New York;;SCidnCe House, 1968. ,

May; Philip. R.A. ; and Turna, A. Hussain. The Raiin, Andrew M. A-B variable, in psychotherapy,:
effect of psychotherapyand stelazine on length A critical review: PsychologicaVulletin, 75:1-21; -.

of hospital stay;'release rate and supplemental 1971.

treatment of 'schizophrenic patrents..: Journa/ Razin, Andrew M. The A-B variable:_ Still promising
of NerVouS and Mental Disease, 139:362.369, after twenty, years? 'In:IGurman, A.S., and pazin,
1964. A.M. (Eds.) The Therapist's Contribu'tiop to'
May P.O.A. For better or :for worse ?. Psycho7
Effective Psychotherapy: An Empirical Assess-
therapy and n variance change: A critical review ment. New York: Pergamon, in ptess.' -,
of the literature. The e-Journal Of Nervous and Reich, Wilhelm. Character Analysis, ;3rd ed. New
. : York; Orgone Institute Press.1949 ,
Mental Disease, 152:184-192,i 1971.
McGuigan, Sharon, and! Seidman, Edward. A-B Ridks, David. F. Supershrinii: -Methods, of a ther-
"type" status; s41%-ccindept, and social com- apist judged successful on the basis of adult
pietenCO:Vourne, of Clin014"sychology, 27:493- outcomes of adoleacent, patients. liet:AAfoks,
494, 1971. .
David F:; Thomas, Alexander, Land:Roff, MOrrill.
:.McNair, Douglas M.; Callanan,-Daniel M. and Lorr-r- (Eds..), Life History Research in PsYp-hopathol-'
Maurice. Therapist ',type" and patient response - ogy, Vol. 3. Minneapol)s: :University 'of. Min-.
nesota Press, .1974, 275-2971. -
to psycnotOerapy. Journal of Consulting Psy-
chology, 26:425-429, 1962: Rierriatin, Fritz. The personality structure of the
Menninger,, Karl A. What are the goals of psy- analyst and its influence on' the course o,f9,
chiatric:edircation?-Bulletto of the Menninger treatment. The American Journal of Psych&
Clinic,16:153-158, 1952. analysis, 28:69,79, 1968.
Hdrbert L., and Garrett, Jamesa, The Prot Rittenhouse, Joan D. Without hospitalization: An:.
estant ethic as a personality variable. Joignal experimental study of -psychiatric care. in the
of Consulting and Clinical Psychology,'- home, 1970 (mimeo) ; 7
.

1971. Rogers, Carl R. Client-Centered Therapy: Its'

Mosher, Loren R. Psychiatric heretics and the CurrentPractice, Implications and Theory. New
: extra-medical treatment of schizophrenia. In: York: Houghton-Mifflin, 19.51. ,

-CanCto, R.; Fox, N.; and Shapiro, L. (Eds.)

Rogers, Carl R. Introduction, in Rog rs, Carl Ft.,
'Sfrativic Intervention In Schizophrenia:. Current and Dymond, Rosalind F. (Eds.) Ps chotherapy
°
Developments an- Treatment. New York: Be- and Personality Change. Chicago: niversity of
havioral PuOlications, 1974, 279-302. Chicago Press, 1954, 4-11; '
Newcomb, Theodore M. SOcial Psychology. New Rogers, C.R.; Gendlin, ,E.T.; Kiesler, D.J.; and
York: Dryden, 1950. Truax, C.B. *The: Therapeutic Relationship and ..
Nunnally, Jum C. Psychometric. Theory. New Its linpact:: A Study of Psychotherapy with
SchizophreniCs. Madison: University of Wiscon-
York: MOGraw-Hill, 1967.
Palmer, Ted B. Relationships between needs of sin Press, 1967.
youngsters and characteristics of treaters: Rokeach, Milton. The Open and Closed Mind.
illuStrations of factbrs whioh bear upon match- New York: Basic Books, 1960. 7 .
ing. Paper presented at the National Meeting Schubert, Daniel S.P., and Wagner, Mazid'Earle.
of the Big .Brothers of America, Anaheim, "A" therapists as creative and personally in-
California, June 1967. yo ved with other people. JoUrnal of Consulting
Palmer, Ted B. Matching worker and client in a CI' Ica/ Psychology, 43:266, 1975.
corredtions. Social Wbrk, 18:95-103,,1973. Sega ,, etnard. Illustrations of therapeutic inter-
Palmer, Ted B. The youth authority's community vention of A and B therapists. Psychotherapy:
treatment project,. Federal Probation, 38:3.14, Theory, Research and Practice, 8:273-275, 1971.
1974.
Seidman, Edward; Golding,- Stephen L.; Hogan,
Terrence P.; and LeBow,_ Michael p. A multi-
:
Parsons, N.M. What happened at Hawthorne? dimensional Interpretation and comparison of
Science, 183:922-932, 1974.
Peyman, D.A.R. An investigation of the effects of the k13- tValea: Jou'rnalr of Consulting and
group psYrAhOtherapy on chronic schizophrenic Clinical Psychology, 42:10-20,1974.
patients. Group Psychotherapy, 9:35-39, 1956. Shader, ichard I.; Grinspoon, Laster; Harrnat?,
Prinzhorn,' Hans. Psychotherapy: Ste Nature, As- Jerold :; and Ewalt, tfack fl.,,The therapist
sumptions, Limitations. Translated. by A. Eiloart.. variable, American Journal of Psychiatry,
London: Jonathan Cape, 1932 (originally pub= 127:1009-'1012, 1971. ''

lished in 1929). Sharer, M. ., and Levinson, D.J. Patterns of

 EXPLORING TIDE PSYCHO-SOCIAL THERAPIES°

Idehlogyand role dennition among psychiatsric T,uma, A. Hussain; iViay,..Piiilip R.A.; Yale, Coralee;
residents.:, in The Patient and lhe?'Mental and Forsythe, Alan B., Therapist characteristics
, .
lloSpitaL _Greenblatt, M.; Levinson, ;FLP.J.;. and
.

and the outcome, or treatment in schlzh-:

°
,Williamir'R.H. (Eds.), Glencoe, ilit:'1,he ,Free
,
4
phrenia. NIMH, prepubitpatiOn mimeo.
,Prese,,:1957. 263-285. Vandenbos, Gary R., and 1 Karon; Bertram P.
310ane, R.: arucel Staples, Fred R.; drit9:1, Allen Pathogenesia: ' A 'new therapist : personality
,; 114 YhricstOn; Neirrrrqi Whipple, 1< Psycho; dimension related tatherapputic effecnveriesa.
therep3(..vesus Behavior Therapy. Cambr!dge, Journal of Personality Assessment, 35:252-260,
; Mass: :' Harvard UniVersity Press, 1975. 1971.. : . -1 .,, ,,,
Smith,.rtdWarcr O dsturak and gestural coat
i

Vinitsity, Michael., A farty-year. follow-up .on the

,

muni9atori.,ii
' A and -tleraPlettYpes";during vocational interests Of Psychologists and their
dyadic ,in'terviews., journal at and relationship, fo,tareer development. Alnerican
Clinical. Psychology, .39:29.36; 102. Psychologi#,.g8:100c1,1009, 1973.
°Sainers, Robert, H. A. nevi .6SpY(netric measure Warren, Marguerite a Interpersonal Maturity
-,af.association for Orilinal,_Variables, American Level Classlficatidn: Juvenile. Diagnosis- and
SociologicatReView, 27T3.9-8.11;.,190g- Treatment Of 'Low, Middle and High Maturity
SO*, F.G. Deterioratiortio schizoPhrepid cant DelinqUenta. Sacramento: California Youth
*ups. British. Journalrof Medical PsychOl Au hoilty,.1966. ,
13143-148, 1960., Warr n, Marguerite Q. The case for dint:Kat-ill&
Stansfield;Ronald,G:;7he.nejiir thecilogy? The. ti: tment of delinquents. Annals 'of the Ameri-
Or-the-!drippin_g' tap, or -Students' demotion can Acay00-Y7thPd.liticatArntt:adclal Science,
and the forces Whicti',:mould ,them: Paper rtd, 30; 4r-159,. 1969,'- ' 7
j?resented at tf*, Association for the
.

WaskOI:iy; Irene, 'E., and P I . .,:141c1 rris, EilEdd.)

Advancement of;$clence, September 1975. This Ptychdtherapy Cha O.! ;- ''' a....;..,.floCkyille,
paper is summari/edby.Bernard Dixon, MUsirigs Marylarkp N4MH,li9r. il ., UblicatIon No. ;E
qn, ;the trained incaOacities of splentists, APA,, . ,'Iki,M, 714-'120: .U.S.G.P.O. NO,-1724-Qc1397y
, ittfonitor;.7(3; May 197¢.
. ,

,, Wexler, Milton.:' Cornmeiit on the five treatment :

Steiijberg;*:The'rEiltItIon ofNocational Preference ,coMparative study. in Gunderson,- Jahn G.,.. .
tb" ''emotional adjustment: EduCational and. and. Mosher, Loren-- gds.): Psychotherapy of a..!
Peychological Measurement, 12:96-104, 1:952:: _- Schizophrenia. Mew York: Aronson, 1975, i
Stephens, JoSeph H,, and_ Astrup, Christian. . 431-433.
Treatmenf outcorne7 in "process" and "non; Whitehorn, John C. Guide to interviewing and
process' schizophrenics treated' by "A" and clinical personality study. Archives of Neurol-
"E3'.! types of therapists. The Journal of Nervous ogy and Psychiatry, 52:197-216, 1944..
and' Merital Disease, 140:449-456, 1905. Whiteharn, John C. The concepts of "'meaning"
Stephens, Joseph H.; . Shaffer, John W.; and, and "cause" in psychodynamics: American
Zlotowitz; Howard I. An optimum A-B scale of Journal of Psychiatry, 104:289-292; 1047.
psychotherapist effectiveness. The Journal of Whitehorn, John C. The acquiring and imparting
Nervous and Mental Disease, 160:267-281,1975. of mental health. Bulletin of the New York
Strauss, A.; Schatzman, L.; Bucher, Rue; Ehrlich, Academy of Medicine, 30:601-607, 1954.
Danuta; and Sabshin, M. Psychiatric ideologies Whitehorn, John C. Alienation and leadership.
and Institutions. Glencoe, The Free Press, Psychiatry (sUpplemeng, 24:1-6, 1961. 'Re-
1964. - ' printed in Whitehorn, John C., and Betz, Barbara
Strupp, H.H. Psychotherapeutic technique, profes- J. /Effective PSychotherapy With the Schizo-
sional affiliation and experience level. Journal phrenic Patient. New York: Aronson, 197, Chap-
of Consulting kychology, 19:97-102, 1955. .ter 5, 75-84.
Strupp, HMI Psychotherapists in Action. New Whitehorn, John C. A.B. therapist variable::
York: Grupe & Stratton, 1960. Commentary on Bowden's paper. Journal. -01/
Sullivan, Harry Stack. Personal Psychopathology: Nervous and Mental Disease, 154:289, 1972: .,-;
Early Formulations. New York: Norton 1972. Whitehorn, John C., arid Betz, Barbara J. 'A study
Sundberg, Norman D., and Tyler, Leona E. Clinical of :psychotherapeutic relationships betWeen
Psychology: An Introduction to Research arrd :physicians and schizophrenic patients. Amer- -
Practice. New York: Appleton, 1962. ican _Journal of Psychiatry, 111:321-.331,-j 54.
Sundiand, D:M., and Barker, E.N. The Orientatiohs Reprinted In Whitehorn, John C-.,, anc1.13etz,
of. psychotherapists. Journal of Consulting Barbara J. Effective Psychothera& with the
Psychology, 26:201-212, 1962. Schizophrefil9:' Patient.' New York: Aronson;'
Tourney, Garfield; Bloom, Victor; Lowinger; Paul 1975, Chapter 8, 107-12if., , ;

L.; Scharer, Calvin; Auld, Frank; and Grissell; Whitehorn, John C.,. and Betz, Barbara,;J, A' adm- '
James. A study of psychotherapeutic process Parison of -psychotherapeutic relationships
variables in p4choneurotic and schizophrenic between physicians and schizophrenic patients
patients. American Journal of Psychotherapy, when insulin is combined With psychotherapy
20:112-124, 1966. and when psychotherapy is used alone. Amer- .

119
 . : ,:111EFERENCES

!can Journal of Psychiatry, 113:901-910, 1957. in the -outorne cif treatment 'with schizo,
Reprinted in Whitehorn; John. C., and Betz, phreniq'ifpatients. American Journat'of Psy-
Barbara J. Effective Psychotherapy with the chiatrw, '117:215-223, 1960. Reprinted- in White-
Schizophrenic. Patient, 'New York: Aronson, horn; John q, and Betz, .Barbara :Effettive
1975; Chapter-19, 129-147...' PsYchotherapy with the SchizOph,ranft Patient.
Whitehorn:,John C. and Betz, Barbara ..1°.:Further New' York; Aronson-, 1975, Chapter41, 183-197.
studies of the doctor, as a crtictat:yariable ,

. 4 .

4
 a

Appendix 1
Statistical and Psychometric
Co erations
There is a perennial problem in applied should be 'read by anyone with an interest
research, namely: How can the practitioner the area. But the, tiook does have a viewpoint,
-land the methodologist communicate with each which means that it must be taken with a grain
"'other? Both are too busy to become expert in of salt or so bile 'purposes.) Nunnally writes:
.
the other's field, and so communicate they "There .adrinttbrrect ways to. construct
must. All too frequently the communication tests: one is to select. items according to their
consists of "little rules" laid down by the Correlations with a criterion...." (p. 245, em-
methodologist. Sales should -be such and; phasis in the original} Under this rule,
such; reliabilities should be such and such; and WhitehOrn and Betz should never have pro-
so forth. The problem is that while the little duced the "A -A cale," and:indeed many of the
rules may be valid in some general sense, all analyseS reported here are "incorrect." Nun-,
too frequently they do not fit the purposes at . nally is concerned with-"meaning." One should
hand. Specifically, the little rules usually serve start with content and aim hopefully at even-
researchers who are doing basic research in a tually validating a construct, chiefly through
highly developed content area far better than empirical correlations.' .

they serve those who are trying to explore,an I, too, am concerned with meaning. But I am
area where little is_known. Moreover, the little quite willing to recognize the impoitance .of
rules of psychometrics serve the area in which certain criteria, to start with them, and' to seek
they were developed, ability testing, far better meaning, and theoretical constructs, through
than they serve the area of personality testing. empirical correlations. The goal is not in
There are a number of these little rules that dispute. The dispute centers on, what is the
may fail to maximize returns in exploration: most efficient way of getting there when there
test construction is best saved by is little existing knowledge to start with..
multivariate linear models resting on the Pear- It would be very easy to assign a co on
sonian r; scales should be long, i.e., consisting meaning ("manual") to the following th ee
of many items; scales should be used in their en- items: "cabinetmaking," "repairing, a clock,
tirety, never piecemeal; standardized tests are and "machinist." But fiom the standpoint of
inherently superior to those not-yet standard . certain criteria, the cdirelations among them
ized; reliabilities should be established at being small, the three items 'fall into three did-. .
each step in the measurement process. All of ferent clusters: SQUEST, NSOLVE, and SD.
these rules'are functional under certain condi- "Cabinetnlaking" is not so much "manual" as it
tions. Frequently, they are not functional for is "lonely" (nonextrovert) 'correlated positively "\-
exploration. with "marine engineet," and negatively with
"conventions." "Repairing 'a clock" is not so
much "manual" as it is "problem solving."
Criterion-Based Tests This highlights the.problems with most, items
in. personality tests (including projectives as:
Very fundamental is one of Nunnally's little Well as questionnaires). Whereas, a column of
rules. (Jum Nunnally's book, Psychometric figures with a " + " sign is likely to be seen as
Theory, 1967, is undoubtedly a classic that arithmetic, ,most items in personality scales
102
 APPENDIX I 103
,

can be viewed in a variety of contexts. If is dif- It is podsible, then, to write a rule

ficult to know in advance what Context dif- diametrically opposed to Nunnally's, namely, if
ferent viewers will use. This may reflect the an area is being explored, then investigators
fact that personality tests are in their infancy should not be boundby existing psychological
by comparison with °ability tests (which Nun- constructs. They should regard standardize&
billy emphasizes). , ; tests as item pools to 'be run against a signifi-
But even for ability tests, where an ability is cant criterion in the area they are exploring.
:little understood, it may be well to start with a But note thel "if" clause; the only satisfactory
behavioral criterion and find its empirical cor- rule in research is that the tools and the way
relates..If the behavior is a phenotype, it like- they are used are determined by the objectives
ly will not generate correlates with n single of the research.
meaning (pages 11-12). In the case of the "Alt a

Scale" the wooden application-of rules about in-

teiynal consistency' and common .meaning Hypothesis Testing in Exploration
resulted in efforts to hoMogenize items tiat
should not be 'homogenized. As a result, the ac- There are two types of hypotheses:
tive expression dimension uncovered in' \statistical and scientific. The basic statistical
chapter III was missed completely for years: hypothesis is the null hypothesis. It is"set up
- When Nunnally insists as he does that items us ally with a hope that it can be destroyed by
should never be correlated with a.criterion of the data. If we can destroy it, we accept an
success, but always with each other or a known alternate hypothesis, one that MS scientific im-
psychological chnstruct, he is, in effect telling plications that, for example, A is greater than
Copernicus to start with Aristotle's concepts. B, or A is positively related to B, and so forth.
In the case of the present research, the We may, be forced to accept the null hypothesis
somewhat shaky, but replicated SXPRSA ,if, in fact, we are studying theWrong variables,
cluster, together with its nascent construct or we have failed to measure properly, or we
tentatively called "active social expression," is chose the wrong statistical test, or for any
not well measured in existing standardized number`of reasons. If slid' errors are nonran-
personality tests, most of which measure dom, they can lead either to acceptance or re-
broader, more inclusive aspects of sociability. jection of the null hypothesis. If, however, they
It is likely, though bY no means yet evident, are random, they contribute to acceptance of
that for the purpoies of exploration, existing the null hypothesis.
personality tests shchild be viewed as item It is particularly not impressive to accept
Pools for coirefatioin with various types of the dull hypothesis in an area where little is
criterion variables. : - known (random variables), or when methods
Indeed, concentrating on constructs in stan- are not established (random measurements),
dardized tests has been one of the liniitations unless there are other evidences that the
of many psychometric approaChes to the mean- various negative contingencies above probably
ing of the "A-B Scale." While the standardized do not apply. If there are substantive findings
tests have shed'a little light on the "A-B Scale" from the same data, then, accepting the null
there leas been little agreethent about that hypothesis has more significance than is the
light. What is needed is some hard thinking case when all the findings arenull. Specifically.,
about the needs of schizophrenics. When we do 'in the present studies, failure to find signifi-
this kind of *thinking, the standardized tests cant correlations among criterion variables
found waritin.g. (Other standard would be of no interest whatsoever were it not,
psychometric approaches that probably have for the other meaningful findings that emerge
not furthered our understanding of the "A-B using these criteria. . .

;Seale" include lengthening the scale by adding It is easy to assume that our knowledge is
correlated items, homogenizing it, and using greater than it is. If we assume that we know
ittiterns out of context. See pages 18, 108, 113, the salient variables, know how to measure
and 143.)
theni, and so forth, then we can attach undue
,
 104 EXPLORING THE PSYCHO-SOCIAr, THERAPIES

importance to a series of null findings. For ex- whether the findings as a whole are significant.
-
ample, Bergin and Suinn write: "Differential This is usually done by comparing the number
effectiveness of techniques is not well of observed. significant 'relationships with the
established" (1975, p. 525). This could easily be number that should be expected by chance:
interpreted as a suggestion that differential Such an approach burdens e4ploration with
techniques have been demonstrated to beinef-, proof. There can be no proof in exploration, nor
fective: In a field where very little has been should there be. None of the pisent studies
established, where the null hypothesis' is ac- purport to prove that the personality of the
cepted more often than not, such an interpreta- therapist 'is important. The Lionells study
tion is unwarranted. reported in chapter II is addressed to this ques-
The other danger of overinterpreting au ac- tion. Nor is it suggested here that the SVIB is
ceptance of the null hypothesis is in assuming the ideal instrument to measure the personali-
that others, who have not had to accept the null ty of the therapist. The question addressed
hypothesis must be wrong. The "A-B Scale" here is: If the.personality of the therapist is im-
has been used by literally hundreds ,of in- portant, how might it affect the outcome? Only
vestigators, many of them with meaningful in studies where 'the personality of the'
results. If an investigator uses the scale with therapist is studied in relation to other factors,
ppll findings, it is more reasonable for him to can its impor 'tance be assessed (e.g., Tuma, et
suggest reasons for the failure of his own data, mimeo).
than for him to' attack the research which pro- It is curious that the success criterion- for
duced the scale in the first place. Manipuifitors produces, of 400 items, 50 signifi-
In exploration, hypothesis testing has a dif- cant at .05 (table 26), while the success
ferent purpose than it does when we seek' to criterion for Subcultural Identifiers produces
confirm or deny a particular aspect of existing only 18 (table 27). The number of cases involved
knowledge. The investigator has a hunch that are similar, 46 youths and 41 helpers fOr the
some variable is important. The task in ex- 'latter, and '42 youths and 45 helpers for the
ploration is to assess which variables are im- former. If.We, had data only for Manipulators,
portant and to provide hunches about their we might be tempted to conclude that the
dynamics for further research. Variables might analysis is worth reporting' Conversely, if we
be considered important if they correlate had data for only the Subcultural Identifiers,
highly with a criterion. While this is -a useful we might be tempted to say that the data are a
principle, it breaks down in practice because of Waste of time. Neither -conclusion iis justified
W
the large number of measures of association since proof that the SVIB is a usefdl tool is not
that are available, each yielding a different the subject of investigation.
valuk.e, and because such things as the number There is yet another reason why the latter
of eases, and levels of measurement influence conclusion is unwarranted. In order to assess
how largc a correlation coefficient may be "chance" we must have appropriate tests of
Moreoveri,'some data must be approached with significance. For Most of the tables presented
statistical tests for which there is no measure in this report, no such tests are available. See
of association. Where different kinds of the section below, "Clustered Samples."
variable's with different distributions and dif- For clinical purposes, the question is rarely
ferent levels of measurement are being of proof, but what is the best evidence
studied, the only common comparison that' can available? For research purposes, when we are
be made among relationships is the probability exploring we aril not concerned with whether
that can be attached to the relationship. This is the findings as a whole are significant; whether
in stark contrast .to the situation where our we have "proven" something. Rather we are
question is whether to accept the alternate seeking systeht,atic evidence about what
hypothesis; in that situation we must establish variables might be important and under what
an a priori level and either accept or reject on conditions they might be important. At the
the basis of that level. same tirpe, "explorers" Must make plain the
Finally the question is frequently asked limited nature of their "proof."
 APPENDIX 1 105

Measures of Association Used in consistency, is related mathematically to r.

These Studies Another reason is that presently available
computer programs for r are much more' effi-
Two measures are used: the Pearsonian cient than are those for gamma. If there are- a
product-moment r, and gamma (Goodman and large number of intercorrelation,, to be com-
Krug kal, 1954; Smilers, 1962). Linder lying r are puted, gamma 'can become eicpenswe.
a number of statistical assumptions contino- Gamma has been used to relate items to
ity, normality, linearity, homoscedasticity and criteria. In personality tests, responses to
interval measurement. Underlying gamMa are items are 'frequently "U-shaped," not "hat-
only two assumptiOns: the variables consist of shaped," since people tend to polarize on some
ordered classes (or only two classes) and the kinds of issues. If r is used to relate a
:relations, if curvilinear, are at least monotonic "U-shaped" distribution to a skewed criterion,
(do not reverse). a few extreme cases can result in a significant r
The assumptions underlyingthe Pearsonian even though there is no relation at all for most
r and the resulting distribution of the statistic of the cases.
are such that only one thing needs be known to The differences between r and gamma can
test whether a given correlation is significant-1 be highlighted by considering item 236,
ly different from zero: the number of ea:sec/CT Energetic people, in table 13. The gamma ,is
such assumptions underlie gamma. An ac- large, .72, indicating that 72 percent of the
curate, significance test for gamma requires rankings do not reverse. But this is largely due
that we build into the test the marginal to many ties. Of the 32 therapists, all but 5 say
distributions (Kendall, 1955, p. 45-46; Hays, they like such people, and those 5 are indif-
1983, p. 654). The numerator of gamma is iden- ferent. The significance level for the gamma is
tical with that for Kendall's tau, and the therefore only .09. The Pearsonian r 'for these
significance test is likewise identical. The same data is .35, significant at .05. In other
.

denominator of gamma has been chosen such words, had r been used for screening items,
that the coefficient can be interpreted as the this item would have been accepted even
proportion of nonreversals in ranking on the though, because of its skewness, it contributes,
two: variables. 7 little to accurate measurement.
Most of the variables in the present. study 'Some psychomstricians will insist that the
are of three types: criterion should have been transformed to nor-
1. Criteria (therapists' success measures). Con- mality, so great is their commitment to the
tinuous variables, but sometimes. skewed. linear model. But how do you transform to nor-
.
2. Personality scales, predictors, and clusters. mality if nearly half the therapists are sue-
Built from at least three items, they gener- cessful with all theiw patients as is' the case for
ally consist of at least seven ordered classes. this particular criterion? Some will suggest
Only one of theM is so badly skewed (Work that the data be thrown out. Yet here are data
Activity) that r is possibly inappropriate. that are irreplaceable (pages 16, 20), very costly
The Work Activity correlations in table 16' to produce, and at hand. Do we still have the
fall short of significance when gamma is right to insist on the linear model? Such
substituted for vr. cavalier attitudes are being recognized and
3. Items. Responses consist of three or four resented by interested representatives of the
ordered classes: true/?/false; like/indiffer- public. Such attitudes do not serve our own
ent/dislike (PTQ, SVIB); agree very much/ long-run interests.
agree somewhat! ,disagree somewhat/ dis- In some tables presenting r's 'for scales
agree very much (Quay). (tables 9, 16, 19, 23) there are two variable's-
Generally speaking, in the present analyses, r Which are' single items (SCIENC and,
is used for relations among criteria, among STNDRD). For these two, gamma's have also
scales, and between criteria and scales. It is. been computed and- tend to be reasonably
also sometimes used aiming items, though not similar to the is shown in the tables. The
for this purpose, because coefficient significant gammas for STNDRD are: TOTAL
alpha, the most popular measure of internal S .45 at .09; S .61 at .02; SXPRSA .67 at .02, and
106 EXPLORING THE PSYCHO-SOCIAL THERAPIES

SXPRSS .83 at .005. The significant gammas 4ample in clusters. The appropriate theory is'
for SCIENC are: TOTL N .38 at .03, N. .34 that of cluster sampling where the degrees of
at .04, and NSOLVE .49 at .002 (Cf. table 16). freedom are reduced in effect by the intraclais
Finally, analyses were undertaken to deter- correlation within the cluSters (Kish, 1965).
mine whether "?" is On the same dimension as In one study of 10 therapists and 17 patients,
"true" and "false," i.e., whether it "lies be- the degrees of freedom are apparently as ..
tween them." The problem is similar for "indif- sumed to be approximately 200, since that
ferent" ih relation to "like" and "dislike." If "?" many segments of therapy .weee rated -1
and "indifferent" are on other dimensions, the (Tourney et al., 1966). Another study evaluated
monotonic assumption is inappropriate. The in- milieu therapy on 43 patients in two wards.
formation statistic (Kullback, 1968) makes no The number of cases is assumed to .be 43 (May
such assumption. If the significance level of the 1968).
information statistic is found to be more ex- Consider two statistical prOblems: In One
treme than that* for gamma from the same there are 35 therapists treating an,average of
data, there is reason to question the assump- three patients each. In the other there are 35
lion of monotonicity. Of the relationships so therapists treating an average of five patients
compared, less,than 1 percent showed a more each. Clearly, these two problems are not iden-
extreme level for the. information statistic. tical from a statistical viewpoint. It hardly
This is not proof of" .dimensionality and seems correct that the number of cases is iden-
monotonicity, but it suggests at least that the tical. On the other hand, it is doubtful that:the
assumptions are not violated by the data. difference in number of cases in these two
In scoring items, the numbers 0, 1, and 2 are studies is of the order 105 to 175.
assigned. to True, ?, and False (or False, ?, and The usefulness of cluster sampling theory in
True, depending upon how the item is keyed practice is limited by the fact that relatively
with respect to the scale). Scale values range few kinds of tests can be applied and computer -

from zero to a maximum which is twice the programs are not generally available. The
number of items. main usefulness of cluster sampling theory,
and of a frequent ally, stratified sampling
theory, is in reorienting us toward "error." In-
Clustered Satirples stead of insisting on certain levels of reliability
in measurement at each stage of measurement,.
Degrees-of freedom are a function of sample the entire enterprise is viewed as designed to
size. But what 4...the sample size? The confu- produce findings, i.e., error enters the
sion arises from the various ways that,the data- measuresin proportion to the cost of removing
can be conceptualized. In the present studies, it at the various stages considered all together:
we think of the analyses as concerned with the In the present studies, we are not really in-
personality of the therapists. The sample size terested in how effective each therapist is..
is the number of therapists. If, however, we Rather we wish to compare groups of
think, of therapeutic process as related to therapists who have- varying characteristics. If
"dyads" or "matches"between client and we accept a therapist who has only one client,
therapist or that certain clients`were exposed we may reduce the total .error more by in-
to psychotherapy, the . number of cases creasing our sample by one, and eliminating a
becomes the number of clients (Berzins, in source of bias through exclusion, .than we'
press). Consider, in table 11, item 221, Express- would lose by accepting the relatively
ing judgments publicly 'regardless of criticism. unreliable measure on that therapist.
If this pis seen as the responses of 35 therapists, The phrase, "bias through exclusion" in-
the gamma is .31, p. = .08. If it is seen as a troduces a different idea. Up to now we have
characteristic of 282 dyads, gamma is .25, p = .01. talking about random error. If we exclude
Most .statistical tests assume a simple ran- certain measures as unreliable, we may in fact
dom sample, i.e., that each unit was selected in- be excluding certain kinds of therapists. Thus
dependently of all others. This assumption is biases are introduced. There is no way for
rarely aChieved; rather, .units come into the statistical tests to allow for biases. In fact, the
 APPEND IX l 107

mOrelrequent the errors of bias, the more like- Harry Carimr used to say: "The error is made
ly that statistical tests will support erroneous very accurately."
findings of substance. Related to the notion of a "long test" is the
Beyond the conceptualizations remain some notion of "fullscale:" if a scale is taken, from the
difficult practical problems. In the present literature, it must be used in its entirety, not
studies we have taken the "conservative" just a part of it. Reproducibility demands that
Course: the number of clusters (therapists) is exactly the same measuring techniques, !be
the Aumber of cases. The Loysville_satnple' is used. In practice, the . idea breaks down
clustered not only around therapists, but also because the literature is not standardized.
in that some youths were helped by more than There 'are, for example; at least 20 different
one therapist. Thus therapist measures are not versions of the "F Scale" in the literature.
independent of each other. As noted above Beyond that, there are instances where ,using
(page 60n) this clustering is also conservative. the full scale is not only wasteful, but likely to
Criteria are biased toward the mean in' a way assure: that the results are not. coMparable.
that is likely to militate against achieving The "F Scale" contains an item that' homo-
significant findings. sexuals should be, treated As criminals. All .

mental health professionals reject this item; it

therefore contributesinothing to measurement.
Construction of the PTQ Worse still, they resent it; it is a rapport .

breaker; they wAnder whether the in-'

In.'Order to persuade a reasonably high pro- vestigator knows what he is doing. In this
-portion of 'mental health profeasionals to re- sense their responses cannot be comparable to
spond to the PTQ, it was decided that the those for whom the idea Of punishing homo-
length should be such that it could be answered sextials severely is not outlandish.
in half an hour. At the same time it was felt Even though an item is not rejected, it may
that the objectives required that the PTQ con- have 'a specialized meaning in a specialized
tain at least 30 scales. These objectives run population. Consider the following two' items:
afoul of two widely held psychometric notions: A man who truly loves a woman must regard
the "length of the test," and "full scales." her as the best in the world in every important
One of the first 'principles of psychometrics respect.
is that the reliability of measurement is pro- We may confidently expect that mankind will
'pOrtionate to the "length of the test," the someday attain a stable social order in which
marital'infidelity will be-unknown.
number of items'in the scale. Psychometricians
have been blessed (or cursed) with subjects Irvin L. Child found that, in a cross-sectional
who have : little -choice but to "cooperate" population, these two, items are positively in-
(students; patients, etc.): There can be no tercorrelated and -negatively correlated to a
assurance.that the subjects will actually read . scale that he calls "Tolerance for Complexity."
all the items, so if the test is very, long, itis A population of mental health professionals .is
necessary to add' still, more items which are rather More sophisticated about loVe and mar-
such as to reveal that the subject is simply riage and they view these two items qUite dif-
-making check marks. ferently. Responsea to the items are not cor-
Interview researchers are actively con- related ( - .02): and do net correlate 'With . the
cerned about rapport and the motivation of the total scale. They were therefore not included in'
respondent. In the impersonality of the testing the final scoring. (For a variety of reaSons;. a'
situation, these issues tend to be forgotten. number of items were not included; according-
The point here is that flagging motivation can ly, the scale was renamed "Need for Closure"
more than compensate for the increase in and inverted. See table 31;)
reliability of a longer test. Worse still, if the SOMetimes 'a scale is a composite of related
subject . can choose not to participate, scales. Child (1965) sometimes combines his
nonresponse can introduce biases which are Tolerance scales (Complexity, Ambivalence,
simply exacerbated by increased reliability. As Unrealistic Experience) int% a single scale.
.

the well-known mathematician and statistician Similarly, the Protestant Ethic of Mire's and
108 - EXPLORING THE PSYCHO-SOCIAL THERAPIES

Garrett (1971) contains several related 'dimen- items were from the F scale, thus giving a nine-
sions: work and leisure, spending, etc. At the item F scale.
time the PT_Q,was designed, it wisielt that the In revising the PTQ, more items were added
work aspect was the impatant one. While to some scales, and more :scales were intro-
subsequent analyses suggest that the spending duced. The composition of the revised PTQ is
one is also important, still it is Wear that so far presented in tables 6 and 31 to 50. (A, second
as the psycho-social therapies are' concerned, . revision of the PTQ is disctlised in appendix 8.)
the two aspect's are distinct. Had the whole A major concern was rapport with the
'
,
scale been included and scored as such, this in- respondent. Instead of, requiring a True-False
sight would, have been lost (see appendix 8). answer, the respondent could check a "?" if he
In short; long scales and full scales would wished. This probably enhanced measurement;
have been possible only by sacrificing more see page:106. The Kemp extension of the A713
than half of the dimensions investigated, or by predictor was dropped:for a variety of reasons
suffering4 greatly reduced response rate. Full including' the fact that it contains an item about
scales Would not have assured comparability; bowel movements which struck some
and might have masked the varying usefulness respondents as being out of place.
of subscales.. Any item which could be seen as measuring
it might well be asked: How can we achieve adjustment was eliminated. The therapist's
reprochicibility in personality measurement? It own adjustment is probably a critical variable
won't be easy, but it' seems more likely that we in therapy (see page 8), but measuring' per-
will achieve it if we think about the constructs sonal adjustment requires a great deal more'
and the populations we study than if we apply acceptance of the researcher than is possible in
some wooden rules. a mailed questionnaire. (It might possibly have
The PTQ was constructed item by item. been included for the Loysville study.) Mental
Selection of SVIB items is discussed in appen- health professionals appear to be' more willing
dix 5: For the other PTQ scales (the manifest to be evaluated psychologically than do most of
personality scales), items were included if they the public. But it is precisely those who might
appeared to relate in any way to how people be sensitive' that would be missed.
might react to a schizophrenic, success rates Even beyond this concern not to measure ad-
with depressives and neurotic patients having justment, was a concern not to include items,
become available too late to rethink the PTQ. which could'be interpreted as measuring social
'An item was excluded if there was any desirability. In fact, it was partly the neutral
evidence that.it might be answered in the same aspect of Child's items which attracted me to
way by' all, or nearly all respondents. This his work. This is considered in the next section.
evidence came largely from analyses of the .

first PTQ which had been used in the pilot in-'

vestigation reported in chapter III. That in- Social Desirability
strument contained, in. 'addition to a sentence
completion test, an 18-item A-B predictor, and It is not uncommon to find that much of the
an 4-item Kemp extension (Kemp' variance in tiersonality scales is taken "tip by a
Stephens, 1971): It also contained items from factor which 'represents the socially desirable
Barron's Independence of Judgement Scale responses to each question. This response et
(20), Child's Tolerance for Complexity (17), Am- is not only a characteristic of personality tests;
bivalence {10), Unrealistic 'Experien'ce (9), Am- it varies alb by individual. Frequently a social
biguit (3), Deference Anxiety (3), Preference desirability scale 'is built.into the test 'so that
for IntJlectual Challenge (3), Preference for allowance can be made for this set. .

Decisi Making (3), Preference for Work Ac- Another approach 'is to seek neutral items.
tivity (3),, Singer's 'Regression in' the Service of As noted 'above, Irvin L. Child's items tend to
the Ego (20),' and the author's Need for Order be descriptive rather than evaluative, and this
(3). Four of the Independence of J,udgement was my concern in choosing' items for the PTQ.
items and five of the Tolerance of Complexity A number of analyses': were undertaken: to

I2
 APPENDIX I 109
1

check whether the items chosen Were indeed This was done using the 133 mental health pro-
"neutral". fessionals- and professionals -in- training (table
Seven jUilges independently gave "the 7). Using this wider group- made it possible to
sociallidesiable response" to each item. Of compare covariance and correlation matrices
232 items, th* were only 41 for which at least of PTQ variableS for various subgroups of the
six of seven judges agreed. on the socially Iample, an important part of the analysis in
desirable response. These items were scored in hapter IV. The values of coefficient alpha for
several ways. The common characteristic of these scales are given in table 6.
the resulting measures is that they correlate For phe A-B clusters, it is not appropriate to
very much like the particular substantive scale expect internal :consistency for the 133 mental
which contributei the most items. In other health professionals and professionals-in-
words, "we failed to find any independent social training, since .nany' of these clusters clearly
desirability scale. have different meanings for different
Next, we put all 232 items into a single factor subgroups. For the clusters derived from the
analysis. Udnally, this process results in the 35 therapists who did not prescribe drugs, the
first factor being "social desirability" and ex- average gammas given on pages 43ff. tend to
plaining much of the variance. In the PTQ no overstate the internal consistency since these
single principle component explained -more gammas were used to form the clusters. For
than 6 percent of the variance: No rotated fac- the PTQ's from the nonhospital therapists,
tor explained more than 5 Rercent of the coefficients alpha for the A-B- clusters are as
variance: No component or faCtor could be follows: SD, .84; SXPRSA, .49; SXPRSS, .08;
identified that had as many as 14 of the 41 SQUEST, .17; DBUSNS; .72; NSOLVE, :68;
items on which there had been interjudge NQEST1, .15 and NQES7, .03. _
agreement as to thp soc*lly desirable A curious observation is that some scales
response. with very low alphas consistently generate cor-
The lettef to respondents contained the relations with other. variables. For Empathic
following paragraph: Interest, alpha fs t24; for Tolerance of Am-
There are no right or wrong answers to the biValence, it is .28NYet both scales generate
questions. No item has been included becatise meaningful correlates (tables 16 and 19). These
it measures adjustment, or mental health, or latter could easily be random. But the need to
intelligence. Rather, the items are concerned
with various ways of thinking or acting: think carefully about the assumption that
It is unlikely that all potential respondents ac- items in personality tests should generate
cepted these assurances. But it was gratifying large alphas. As noted above such items tend
to see one day, scrawled across the front of one
to tap, more than one 'domain, or have the
PTQ: "This is a' fun test."
potential to be interpreted in more Nan one
domain. For this reason, there may be a .low
correlation among personalitek items (there
Internal Consistency of Scales usually is) and still they may have a common
One of the most popular of the various don-fain. The, item, "Science should have as
measures of reliability, homogeneity, and/or much 'to say about moral values as religion
consistency is coefficient alpha (Niinnally, does" would seem to involve at least three do-
1967, p. 196). It has a variety of interpretations, mains: the worth of science, the need for moral
of which one of the easiest, is that it is the prescriptions, and the ,role of ,religion. De-
average of all possible split-half reliabilities. pending upon the viewpoint of the
There are, of course, many measures of respondents, and the way it is combined with
reliability,.and they sdo not all give the same other items, it probably could contribute to at
answer. .
least three genotypic variables. Barron
For the manifest personality scales in the probably had in mind-acceptance of religious
PTQ, items were excluded beginning with that prescriptions when he tried the item for his In-
one least correlated with the total score. If dependence of Judgement scale. Our
such exclusion resulted in an increase in coeffi- nonhospital psychiatrists apparently see the
cient alpha, the item remained out of the scale. item as concerned with the worth of science.
110 EXPLORING THE PSYCHO-SOCIAL THERAPIES

Some will argue that it is a bad item, that it There is no assurance that short simple items
should be purified: "Science should have more will be answered in a common context.
influence." Or: "Religion should have more in- And so this appendix ends as it began, ques-
fluence." What is missed in such criticism is tioning whether the usual 'statistical and psy-
the fact that these pure items are answered in chometric rules serve the purposes of explor-
some context, this context being supplied ,by* ing the personality of therdpists as well as they
the respondent instead of by the researcher. .serve the areas in which they were developed.
Appendix 2
Possible Personality Correlates
of the "A-B Scale"
(This paper was written in the fall of 1972 before the second study was begun. Its con-
tents are not necessarily consistent with the present material. It explains, in part, the
construction of the Personal Tendencies QuestionnaireJ

It is now. a dozen years since Whitehorn and teracted with make-believe schizophrenics and
Betz (1960) plablished, the A-B Scale.. Although neurotics with results that' almost always prit-
these years. have seen atleast 150 articleS- and duced an interaction among the Scale,
dissertations in which the A-B Scale played a diagnosis, and some dependent variable. Ito
central role, we seem to have' very little would be an interesting study in the sociology
understanding of what the scale means. The of science to explain why the research has
present paper is concerned with possible mean- developed in this fashion.
ings of the A-B Scale..There are a large number Recently, it has broken out of these bonds
of possible meanings; presently available -"Ivith results disappointing (Bednar and
evidence does not permit us to choose among Mobley, 1971, Bowden; Endicott, and Spitzer,
them. The purpose here is simply to present a 1972), possibly encouraging (Shader, Grin-
number of possible meanings. spoon, Harmatz, and Ewalt, 1971) and definite-
The reader is referred to two excellent ly encouraging (Berzins, Ross, and Friedman,
reviews of the. A-B literature (Razin, 1971; 1972). Where earlier studies were primarily
Chartier, 1971); and to articles cited below with therapists in training or with quasi -
which have, been published since 1969,.the cut- therapists,' the last mentioned is with ex-
off date for thOSe reviews. There is very little. perienced therapists. The interaction was con--
in this literature concerning A-B as a concept. firmed- suggesting that the personality at-
While there is some divergence in the ap- -tributes underlying the 'A -B Scale are not easily
praisals of the empirical usefUlness of the modified by the therapists' own experience
scale, when one considers (1) 'how little we with different types of disorders. Perhaps if we
understand it, and (2) 'how -unreliable are our can learn the meaning of the scale, therapists
diagnoses (and even our ideas about can learn to adapt.
diagnoses), one is impressed with the degree of The present paper speculates al3out the
°consistency that runs through these studies meaning of the scale and about the la ossible
that A -therapists are More effective with. puzzling contradictions it may contain. It
.schizophrenics, while. B's may be more effec- should not surprise us that the 'A -B Scale .is
tive with. neurotics.. If thiS Is the case, then puzzling. The schizophrenias are very puzzling
these items from the Strong Vocational In- disorders. It is possible that the puzzles await
terest Blank Stand for a very poWerful a common solution and that whatever sheds
variable, or set of variables, indeed. light on one of them will help with the others:
During most of its 12 years, the A-B Scale
has lived .primarily in the Henry Phipps Clinic
where it was developed, in the nearby Shep-: General Characteristics of the A-B Scale
pard Pratt Hospital, and in the psychological
laboratories of several universities. In these There are in fact a number of A.:B Scales
laboratories, make-believe therapists in- (Kemp and Stephens, 1971). Some are efforts to
111
U
 112 EXPLORING THE PSYCHO - SOCIAL THERAPIES

shorten the original scale, some are efforts to 13 items, representing the rejection of manual,
lengthen, it, while one is an effort to' adapt it to mechanical, and engineering interests may
the revised form of the Strong Vocational In- , reflect more the generhl content of the Strong
terest Blank (Campbell, Stephens, Uhlenhuth, . Vocational. Interest Blank than it does the
and Johansson, 1968). Of` interest. is the fact
. . needs of schizophrenics. The other 10 items
that the interaction tends to appear regardless may be just as important. Further, the items
of which form is used. I will concentrate on the together may not be exhaustive of the needed
original analysis of Whitehorn and I3etz (1960). traits.
They presented their results in two forms: as In presenting the possible meanings of the
career patterns, and as individual items. In A-B Scale below, I will present one at a time,
career patterns the A therapist4 (effective not because I believe there is a single cor-
with schizophrenies)'tended to be like lawyers relate, but because it is difficult to think about
and. C.P.A's, while' the B's tended to be like combinations of scales when we are not even
printers and mathematics and phy cal science sure that, one of them is relevant:What is
teachers. Using item analysis th found that needed, of course, is-a study of the correlates of
23 items differentiated the. A's the B's. It the subscales of the A-B Scale.
is this 20-item scale and- modifications of it that
have been used most often in studies of the ef-
fectiveness of therapy or quasi-therapy. Possible. Meanings of the A-B Siutbscales
Of the 23 items, 13 represent manual,
mechanical, and engineering interests. The A's Ideas about the meanings of the A4_i
say they dislike these activities:Psychometric Subseales may I be derived from several
efforts to make the scale unidimerisional result sources. There is the A-B literature. There is
in these 13 items being among the survivors, other literature of personality which may be
indicating that whatever its that these items relefant.. Finally, if we simply consider what
represent tends to dominate the -total scale. the schizophrenias are like, we might an-
The other 10 items don't sem,to hive much in ticipate that certain traits are relevant in
common and they usually do not correlate well treatment.
with the total scale. The emphasis here is on the schizophrenias,
Razin (1971) questions why the scale should not on the neuroses. Whitehorn and Betz (1954,
-be made unidimensional. I agree, and indeed it 1960) were concerned with characterizing the
n can be argued that making it unidimensional successful treatment of the schizophrenias. It-
may destroy its value. If we assume that the was later ,foUnd (McNair, Callahan, and Lorr,
'successful treatment of the schizophrenias in- 1962) that B's were more effective with outpa-
volves a rare. combination of traits,, we would tient neurotics.'
expect that within the 23 items there would be In thinking about the schizophrenias and
several dimensions, and.that these dimensions what it is that 'schizophrenics need from their
might well b_ e uncorrelated in a random sample therapists, one is immediately confronted with
of people, or even perhaps: negatively cor -.. conflicting hypotheses (see Razin, -1971., pp. 13
related. and 18): In its simplest form, the -question
. The A-B Scale may be not one test but two of becomes whether the therapist should be like
more. It may function like a battery of tests in the patient in order to understand him, or .

employee screening,which are designed to, say, whether the therapistshould betinlike the pa-
select machine operators who have high finger tient in order.to help him" out of whatever it is
dexterity and high grasping strength, tviro he is in. Therefore, with respect to most per-
characteristics which are. probably negatively sonality variables it,could be argued that A
correlated in a "randoiniample of individuals. therapists should be similar to schizophrenics,
The combination of traits required to treat or that they 'should be the opposite, corn-
schizophrenics may be very rare indeed.
Whitehorn and Betz did their analysis item 'More accurately, more effective with Veterans Ad-
by item, each entered individually. Itwas not a ministration outpatients, a finding that is frequently inter-
multivariate analysis. The fact that there are preted that B's are more effective with neurotics.

3i
 APP:p1DIX 2 113'

elementary. If the A-B Scale is multivariate, 31 which Manosevitz 11970) found to

both of, these may be the case, i.e., the discriminate homosexuals. Taken together,
therapist may be like a schizophrenic in some however, these MMPI studies suggest that
respects and unlike him in others. there are many items relating to femininity
In presenting possible`ineanings of the A-B and homosexuality- which are not correlated
Scale, I am not so much concerned with the with the A-B Scale. Additionally, of nine MMPI
direction of the relationship (positive or items which appear to correlate with the A-B
negative, like or unlike) as I am with the possi- Scale, five lie outside the domains of femininity
'ble relevance of the.dimension. The first task is and homosexuality.
to uncover relevant dimensions, and .secondly In the A-B Scale itself, the A therapist says
to discover, how these dhnensioni relate to the he prefers to be president to being a'conimittee
subscales of the A-B Scale., chairman, and. that he follows up subordinates
The, suggestions below are in no particular effectively. These items have a masculine tone.
order; no effat is made here to judge which of Thus although it is possible that the A
them (or which combination) is likely to be therapist is more feminine than the B, the
found valid inuture research. Some of the evidence available is not yet persuasive.
ideas presented may seem to the reader to be Less Social Competence. Schizophrenics
rather farfetched; how could such a variable generally have less social competence than
relate to psychotherapy with schizophrenics? neurotics. Medvigan and Seidman (1971) fOund
However, since communication of personal that of 65 students, the 12 nearest the "A" pole
values is very subtle, our present lack of were less- socially competent than the 12
knowledge requires us to keep an open mind. nearest the "B" pole. For the gioup as a whole
Femininity. Dublin, Elton, and Berzins (1969) the correlation was .19 (p < .20). The tendency
found that among volunteer freshmen, A's are is therefore for the A's 'to resemble
more fegninine than B's as measured by the schizophrenics more than they do the`,.-
Omnibus Personality Inventory. This could be neurotics, but the correlations are not strong
of interest' because of the general clinical con- enough to be persuasive.
sensus about homoSexUality 'among Rejection of Manual Occupations. Since A
schizophrenics. The difficulty with the Dublin therapists tend to reject the manual occupa-
et al.: finding is that the O.P.I., like most inven- tions, one wonders if this might also be true for
: tories, includes interests among their schizophrenics. Klugman (1960), ,using the
measures of masculinity-femininity. Of 56 Kuder Preference Record Vocational,ifound
items, 13 concern preferences for science or that schizophrenics as compared with normalS
the ,humanities. Conceivably, the rejection of -rejected meehanical pursuits, and showed
science could be akin to the rejection of tendencies to prefer literary and clerical ac-
mechanics in the SVIB. It would be interesting tivities. On the other hand; Steinberg (1952)
to know whether the A-B Scale correlates with reports similar findings.- comparing neurotics
a femininity score derived, from 'the 43 items with normals. It is not clear how schizo-
which are not: concerned with science or the phrenics _would compare with neurotics on
humanities. these interests. It
is surprising how few
Kemp (1963) found that 8 of 300 items in the studieg there are of vocational interests among
MMPI correlated with the A-B Scale. There is persons with mental disorders.
a ninth item (building contractor) in both tests. Rejection of the. Work Ethic. Mire's and Gar-
Four of the 9 are also among the 60 MMPI rett (1971) have devised a Protestant Ethic
items which measure femininity. Of these four, Seale. Most of the items are concerned with ac-
three (building contractor, mechanics ceptanee of the work ethic and rejection of
magazines, and fore\st ranger) represent con- leisure. The remainder extol sacrifice. and suf-
tent overlap, or possible overlap with . the fering in contrast to pleasure seeking. They
SVIB. This leaves only one item of femininity find that the SV-Ift- scale for math-science
picked up. This one item (bothered about not teacher (B therapidt) is positiVely related to
being better looking) along 'with building con this scale. The SVIB scales fqr lawyer and for
tractor and mechanics magazines, are among author-journalist (A therapist) are negatively
114 EXPLORING THE PSY.CHO-SOCIAL THERAPIES

related to the Protestant Elhic Scale. Thus determination rather than for obedience and
there is a possiblity that A therapists reject Conformity. On the other she presents the
the Protestant Ethic. Whether this might be.a -sChizophienics' need for firmness and limit-
rejection of work, or a rejection of moralism or- 'Setting. Possibly the dilemma can be resolved:
both, is certainly an issue worth pursuing. by the now classic differentiation of
Very few therapists would argue that one dernocracy, autocraCy, and laissez faire, where
ought to be moralistic with schizophrenics, but the B's are seen as alternating between"
probably even fewer would argue for moralism 'autocracy and laissez faire. If so, does this lat-
in the treatment of neurotics. e ter behavior help a neurotic? .

Rejection of Dogmatism and Author- Field Indej$endence. A therapistS,have been

itarianism The Protestant Ethic is positively found to be more field dependent than B
Correlated with Authoritarianism as measured therapists on the rod-and-frame :test (but not
by the F Scale 1Mirels and Garrett, '1971), -on embedded figures): Field dependent persons
which in turn is highly correlated with are thought to be less cold and distant; less in-
Dogmatism (Kerlinger and Rokeach, 1966). volved- in cognitive, intellectual, and
Betz (1962) sees the B doctor as a rigid, author Philosophical pursuits; less individualistic; and
itative instructor who sees things as right or more attentive to subtle social cues. Since
wropg. these atter characteristics have greater
Tolerance of Schizophrenic, SyMptomatol- relevance to psychOtherapy than the rod -and:
bgy. Authoritarianisth has been found frame test, it might be preferable to try to
repeatedly to be associated with negative at measure them directly.
titudes. 'toward mental disorders. The pro-
totype of mental disorder is psychosis, not.,
neurosis. If the A. therapist is less author- Significance of the Issues
itarian, .he might be more tolerant 4".Of
schizophrenic symptomatology. Whitehorn and Studies of in-therapy._ behavior se m to in-
Betz (1954)Jound that he was leSs concerned dicate that successful therapy reqiiir tha a
with reducing symptornS and more concerned therapist have accurate empathy and un di-
with helping the schizophrenic to resolve con- tional regard for his patient (Rogers, Gendlin,
flicts. He was more likely to confide in the pa- Kiesler, and Truax, 1967). But hoW is this to be
tient and to participate actively. Segal (1971) il- accomplished? Are these charactOisticof par-
lristrates how the A therapist becomes more titular therapists or does the A-B Scale sort
perSonally involved with his clients. All of out which therapists are capable of having
these findilgs suggest that the A therapist these kinds of relationships with. which.
may be less concerned with status and that patients?
. there may be less social distance between him . Still another line of reasoning is that the A-B
and his patients. Whether this Would mean Scale sorts out different the pies e.g., emo-
greater freedom for the patient is problem- tional involkrement with the chizOphrenic and
atical. We have been:considering here author- 'intellectual, involvement "th the neurotic.
itarianism as an .ideology, not as an interper- Much of what Betz (1966) writes. suggests that :-
sonal stance; as suggested: next below, the two a trusting confidential relationship is par-
May not be the same. , titularly important for schizophrenics.
Acceptance or Rejectio'n of Authority. How, Studies of in-therapy behavior tend to be
ideologies work out in actual patternS of in- :very expensive relative to correlational and 4)
.
terpersonal relations- may be quite complex.. analogue studies of the A-B Scale.. The latter
Betz (1962) makes much of the schizophreniCs' may give us hints as to what to look for in.
problems withiuthority: It would be easy to therapy. MoreoVer,, while the in-therapy .'...
assume that democratic A's avoid these studies tend- to concentrate on -specific '

problems. But among the 23 items'is one about behaviors and dyadic relations, studies of the
drilling in a military company. It is the A's who , personalities of therapists may explain the dif-
show less dislike for this item. On the one hand ferential meaning of these behaviors and rela-
:Betz talks about the A's respect for- self- tionships.
 APPENDIX 2 J 115

The same Overt behavior may haVe very dif- few.") Indeed, the A-B Scale has very differept
ferent -meaning in different therapeutic set- correlates for females than for males (Dublin,
tings. Consider for,example, the very different Elton,.and Berzins, 1969). If we could assume
megningS of touching for like-sex dyads as corn- that the phenotypic behaviors we observe in
pargd with opposite-sex dyad 'The overt therapy, have the same meaning when 'per-
behaviors have different meanings. milarly.a formed by male and by female therapists; it
common meaning: may express itsel 'n dif- might not be so important to discover what
ferential behavior. The specific beha personality dimensions underlie the A-B Scale.
which are consonant with a particular kind of Until we understand these dimerisioils, the A-B
therapeutic relationship may be qUite different Scale offers no guidanc or r-the many women
in individual therapy, grout, therapy, arkd who are psye gists. . .

milieu therapy. Understandilig personality at- Actually, as things stand now, the A-B Scale
tributes may help us to generalize relation- Offers no real guidance for male therapists. It
ships across various treatment modalities. could toe used as a selection device for male
The A-B Scale was validated on a sample of therapists-(which it apparently has not), .but it
therapists, nearly all of whom were male' Of really doesnt help us* understand. If we
the., various- items, hardly any could possibly understood the personality correlates of this
have the same meaning for a woman that it has scale, we would have significant leads toward
for a mail. ("I would like to be a carpenter." "I the effective treatment of .the schizoplirenias,
prefer ohaving many women friends to only- a and .possibly also of the neuroses.

4
Appendix 3
Personal Tendencies Questionnaire
The first Personal Tendencies Questionnaire reproduction. At the same time, the ,question-
(PTQ), the one used in the pilot Study (chapter. naire does not include many other SVIB items
III) is *scribed On page 108. The revised presented in tables 25, .26, 27, 28, and 63. In-
questionnaire which was used in the Second vestigators - who are planning this kind of
and Third studies (chapters IV and V) is research might well consider the following:
reproduced in full on the following pages. It is Administering the entire 400-item SVIB.
further dacribed on pages 107-109, and in Dropping the SVIB lams from the PTQ and
tables 6, 10-13, and 17. In appendix 4, the per- adding items on treatment ideology (appendix 8)
sonality scales of the PTQ are presented and adding to certain scales such as Empathic
separately. The A-B predictors and clusters in Interest and Tolerance for Ambivalence.
the PTQ are presented on pages 36-45. The
questiahlfaire contains niany, items from the Fortran proirams for scoring the PTQ may
Strong Vocational Interest Blank (Form M) be secured from the author.
With the permission of the publisher, but this A second revision of the PTQ is described in
permission does not extend to further appendix 8.

116

I3 kJ
 APPENDIX 117

Personal Tendencies_Questionnaire
It will help us most if you will record your initial
reaction by checking or circling the"'T,".the "?" or 4010111!P
The "F."

When I am a part of a team orzroup that is working or

playing together, i prefer not toet so involved and
caught up in the activity that I lose my feeling of
Separateness.
I can ge cept Just criticism without getting sore. 2
The 14ader of a group-Is likely to be most effective
if he acts on the assumption that everyone in the group
Ilk is glad to have him as a leader. F 3
I prefer that my hours of eating and sleeping be regular,
not changing from day to day. T ?
What youth. needs most is strict discipline, rugged
determinationi and the will to work and fight for family
and country. 5
What I have hoped for in life generally \s coming to we. 6
It's a good thing for a teacher to leave basic problems
unresolved, so that students have a figure out for
Othemselvet whether there is a.clear answer. 7
When I chooSe to do something, the fact that it may not
be allowed is relatively unimportant.
Human nature being what it is, there will always be war
or other forms of serious social conflict. ? F
Knowingthat something .will be very hard to understand
makes it more interesting to me. F 10
I have often thOughl at some length the various ways
In which I would be affected if my nearest relatives
were suddenly to die. ?. F 11

I enjoy putting my own affairs aside to do someone a'

fa'cor. 12
I feel very' insecure about.reaching any decision of
Importance entirely on my own. 13..
I am inclined fo keep quiet when out ina'social group: T. ? 14

Some of my friends think that my ideas are imi5ractIcal, if

not a bit F 15
My accomplishments in.life are .about as high as my expect
tations. ? F 16

.
Certain items are adapted. and reprinted from the Strong Vocational' interest Blank for Men
(Revised), Form M, by Edward K. Strong, Jr., with the permission of the publishers, Stanford .

ynlversIty Press. Copyright © 1938 (renewed 1965), 1945 (renewed 1973) by the Board of
Trustees of the Leland.Stanford Junior University.
 4.
EXPLOEINO THE PSYCHO-SOCIAL THERAPY

People fall.Very.naturally.into distinct classes, such as

the strong and the weak. ? 17'
I would. prefer being president of a society to being a
committee chairman or a member. 1,8
. I believe ybu should ignore other people's faults and make
an effort to get along with almost everybody.' . ?' 19
The man who.can approach an tinpleasant.task with enthusi-
asm, is the man who gets ahead. 20
While listening to a friend recount an experiende, I very,
seldom emp'hathize to the point .of feeling what he must
have felt in the situation he is describing. T 21
I have mechanical IngenUity (inventiveness). T 22
Insofar as the study of philosophy makes one doubt-his
basic beliefs, it should be encouraged.- 23 .

,if one works. hard enough, he is likely to make a good life

for himself. 24
I enjoy letting my thoughts wander aimlessly, and find my-
self thinking about all sorts of unusual and unrelated things.. T ? . 25,
,
I thinkthat my future will fulfill.my aspirations. )? 26
Those religions are to be most respected which impose no
uniform beliefs on their members. T ? 27
The real fault is to have faults and not try to amend them. T ? 28
Once a husband and wife, have contemplated divorce, t6,9;
can probably never be truly happy with one another again. '29
, I would rather spend an hour solving several easy, math prod -4..
!ems than solving one hard one. F 30
I like the sense of privately feeling my way into some sym-
pathetic parilcipation in every kind of human emotional
experience I hear about, regardless of how tragic or unsavory
Some of them may be. F 31
Even the,strongest love for a person is entirely compatible
with the presence pf a variety of negative feelings toWard
the same person. 32
If people would talk less and work more, everybody would be
better off. 33
When I go out in the evening, like it best if all the
activities are planned. 34
In trying to understand what another person is like, I find
It best to put little emphasis on the, feelings I get when I
am with that person. 35
I am inclined to keep in the backtround on social occasions. 36
Most of our social problems would be solved if we coyld
somehow get rid of the immoral, crooked, and feeble-minded
people. 37
 APPENDIX 3 119
14

It. Is better to let the past be forgotten; to be remorseful

is a waste of time. 38
The best theory is the one that has the best practical appli-
cations.' 39
I prefer deciding what is morally right myself, rather than
accepting the moral code of society or institutions. 40
. In listening to a lecture, I often am amused by thoughts of
double meanings or possible puns which the lecturer prob-
ably doesn't intend.
I prefer having many women friends rather than only, a few
women frlends.. F 42
I am very sensitive to the emotional attitudes people some-
times want to convey but are unwilling to state openly. 43
I refrain fr,pm giving advice an help becaLise I don't want.
People to be dependent on me. 44
I mustadmit that, I would find it hard to have for a close
friend .a person whose manner or appearance made him
_somewhat repulsive, no matter how brilliant or kind he
_might be. 45
Our society would have fewer problerni if people had less
leistire time. T ? F 46
Arguments.about-irreconcilable differences:---such as contra-
dictory religioUs beliefs or basic valuesare more interest-
ing than arguments about points on which one person is
able to perquade the others to agree with him. T ? F 47
I can correct others without giving offense. T F 48
It is not possible to know when one is doing good or evil. T ? F 49
I_ would rather open a heavy door myself to having it open
automatically. T F 50,
P.

.1I prefer not to associate with the kind of people whO some-
T 51
, timesget silly and childish. .

A beautiful sunset would bel still mre beautiful if it lasted

longer, and were not a tragic reminder of how transitory
everything good is. F - 52
The thing I would particularly hate about military service is
the requirement of obeying orders of my immediate superEtor. T ? F 53
It Is very interesting to hear about, or to observe, an in-
tensely emotional experience of gomeone I know. 54
When at a boring lecture, while I may get restless I seldom
get the urge to set off a firecracker, jump up and down,
scream "Fire" at the top of my lungs, etc. 55'
Sometimes wheh I think about how much other people have,
or how much they have accomplished, I feel ashamed. 56
At the end Of a textbook chapter, a good summary is more
valuable than a set of thought-provoking questions. 57-
.120 ? ? EXPLORING THE PSYCHO-SOCIAL THERAPIES

It would 'not be very-interesting to try.to feel one's way

Into what the experienqe of a blind person is like; it-would
be better not to think abo,ut it unless yqu had to. 58
Occasional sharp words are no threat td any genuine friend -
.ship.' 59
believe in individual reildrOn rather than institute hal-
ized religion. 60.
There are times when I feel as if. I were a small child again,
or when I wish I could be.one for a shorftime. T 61
I am able to meet emergencies quickly and effectively. 62
When a person has .a problem or worry, it is best for him to
think about it rather than turn his attentionto More cheer-
ful things. T F 63
Fun-house mirrors that completely distort your body-shape
provide an interesting experience. - 64
I .get annoyed when people l'hardly know greet me as if they
were old friends. T 65
If given a chance, I would.do something of great benefit to
the world. 66
Obedience and respect for authority are the most important
virtuesrchildren should learn. 67
I.Stimulate the ambition of my associates. ? 68
The unfinished and the imperfect often have greater appeal
forme than the completed and polished. 69
If I had to choose between carrying out a program of a supe-
rior whom I respect and working for myself, I would, choose
the latter. 70
I rarely let my imagination wander to the point where I think
thirigs which are so unusuakrsexual, violent, or irreverent
that they could be discussed only with close friendsif
anyone. 71
A person will get along better with his close relatives if he
recognizes that he resents them as well as loves them. 72
I conform to conventions even when I don't lik'e them. 73
The worst crime a person could commit is to attack publicly
the people who believe the same thing he does. 74
I don't understand how men in some European countries can
be so derhonstrative to one another. .75

1 1

I like to have many social engagements.° T

Sometimes I feel I learn almost as mich from watching
people's faces and their hands while they talk, as I do from
listening to what the9 say.
I am always on time with my work. 3

L3J
 APPENDIX 3 121

I ,Would: rather have a few Intense friendships than a great

'?
many friendly but casual relationships.
I p( refer, to take an elevator ratheiothan to climb a flight
.4
of-stairs.
--When I am *one or performing a task requiring,no mental
'workil enjoy doing things like rhyming words, dis-
covering 'puns, 'and Making up. little songs. 6
Rather than develop or operate a new machine, I would
prefer to interest the public in the matiine through. public
addresses. T
am unable to work efficiently when I am in a subservient
position.
Quite apart, from considerations of morality or prudence,
the confused sensations of extreme drunkenness would be a
very objectionable experience to go through. F

I seldom become enthusiastic over new ideas or experiences

the wajr, some people do; rather,. I tend to take these things
9
in stride. ,

Sometimes I ruMinate about things I have done wrong.

No one can be sure of conquering his diffidulties; willpower
F 12
is not enough. .

T 13
I find it easy to get along with people younger than myself. ?

Science should have a_ s much to say about moral values as

T 14
religion does.
I geta kick ottt of trying to solve a puzzle even when I
15
fail
I get little pleasure or fun out of playing with words and
languageas by talking nonsense, baby-talk or in a foreign
16
accentand seldom do that sort of thing.
It would be herd to have complete respedt for someone as
leader of your group if you have any relzfundamentar
17
disagreement with him.
18
An insult to our honor should always be punished.
The challenge of novelty and excitement in a completely
new situation is worth the price of disrupting one's old
19.
established ways.
Kindness and generosity are the most important qualities for
T 20
a wife to have.
21
I tend to regard.little children as principally a nuisance.
Avery perton should have complete faith in some super-
natural power whose decisions he obeys without question. . T F' 22
,Clouds that are frankly clouds and coverthe whole.skir are
preferable to the little floating ones that leave you never
knowing whether the next moment will be bright,ot dull. 23

Io
122 EXPLORING THE PSYCHO-SOCIAL THERAPIES

It would be difficult to maintain a warm:friendship with a

person who has some.markedly disgusting characteristic. T
I prefer a job that requires making .difficult decision/
rather than Joilowing Instructions carefully.
When playing with very young children, -1 find it easy to get
into their world and experience things as they do.
I follow up subordinate4 effectively.
Unquestioning loyalty is the first requirement of good
citizenship. T
A slow-acting anesthetic, which would provide a period of .

relaxed, hazy, mixed-up sensations would be preferable to an

anesthetic which would.up you under instantaneously.
A person should not probe too deeply into his own and other
people's feelings, but: take things as they are.
I like having a few intimate friends rather than many
acquaintances
Wild and impractical ideas are often the ones that are: most
worth. pursuing..
I. feel that since you only live once, you should enjoy
yOursetf.
The happy person tends always to be poised, courteous,
outgoing and emotionally controlled.
I prefer reading a book to going to.a*movie.
I can detect In myself no strong antisocial irhpulses of the.
sort which, under certain circumstances, might lead to
crime.:
The expert ski.jumper should enjoy the sport all the More if
it remains a source-of tensionand even alarm.
I resent anybody whose position IS superior to mine and who
can exert a certain authority over me.
A distaste for hard work usually,reflects a weakhess of
character.
I rather like floating in water, for the pleasant sense it
gives you of your own,identity as against the formless mass
of surroupding water.
It Would be exciting to arrive in a city for the firSt time
and find it enshrbuded In a heavy fOg.
The man who truly loves a woman must regard hdr as the
best in the world in every important respect.
People should have more leisure time to speiid in relaxation.
If young. peop e get rtfbelilous ideas, then as they grow up
they,sught to g t over them and settle down.
The igt ;Yqin i some places of Pelebriring a funeral with a
tt ken" ast is disgusting to anyone' who respects. the dig-
4144f h man sorrow.
 123
APPENDIX 3

While watching a play ot,listening to music I am often com-

pletely carried away, becoming totally absorbed and feeling
T 46
as if I were a part of the play or the music.
The present is all too full of unhappiness. It is only
47
the future that.Counts..*
It wouldn't makeriwvery unhappy if I we. A ptevented from
.
48
'aking.
M numerous social contacts.
To be Wandering around a familiar city and suddenly realize
that you are lost, would be an npieasant experience. 49

I prefer team games to games i whiCh one individual corn-

50
petes against another.
.

51
: I do not like to mix socially with people.
The rhere'you get to know and like a person, the more you
? 52
are aware of his-weaknesses and failings.
Nobody ever learned anything really important except
F I 53
.through suffering. ,

In watching football, a close game in which your side

finally loses is, more enjoyable than a game :in which your
side is.way 'ahead almost from the start 54
would rather cut down weeds with a scythe than use a
55
mowing machine.
My enjoyment of music and art takes the fprm of quiet pleas.
Ure; andiseldorn have the intense .emotidnal reactions to
them that some-people do. T 56

.It is only natural that a peyson should have.a much better

acquaintance with ideas' he believes in than with. ideas he
9. 57
opposes.
58..
'llerld to set very high standards for my own performance.
If there were a'harmlesa drug which, would temporarily make
one's sense of smell as sensitive as that.of a dog, it would
59
great fun to try taking some...
I could cut-my mooringsquit.MY home, my family, and my.
I.
60
friendswithout suffering*great regrets.-
I do natend to develop attachments to inanimate things
or objects-(such as autos, houses, tools, etc.) to the.
point where it almost seems as if they had a per6onality or
61
were alive: .
.
.
We may confidently expect that:mankind will somedayAttain
. .

ustable. social order In whict(marital infidelity will be

r..
nknown.
get annoyed. When a st%ngertalks lb me on the bus., train,
63
Y Airplane.
Opticarillusions and other experiences that put you in con-
. .

flictabout what is real,a9d-what isn't are on the whop

64
quite enjoyable.
124 EXPLORING THE PSYCHO-SOCIAL THERAPIES

Even if you were an expert in all the reloWitl languages, it

would be disquieting to be in a multilingual groUp where.you
would, never know what language to use next. . ? 6
sometimes. have daydreams in which I become a "heroic"
typeof figure; either all-.Powerful, all-knowing, and.suc-
cessful or someone. who has sunk..to the lowest depths of
deprivity; weakness, .and suffering. - 66
When it comes .to a difference of opinion in religion,.we
must be carefUi.nOt to compromise with those who believe
differently from-the'way we-do. 67
Hard work offers little guarantee of success. 68
Regardlesi of what caused It, dizziness would just `in it-
Self be a very disturbing experience. 69.
What tOls country needs most, more than laws and political.
programs, is &few courageous, tireless, devoted leaders_ in
whom.the people can put their faith. 0
Playing with words, jas-in punning, ought to be avolded,-_
since it interferes with the normal use. of words for corn.'
municating ideas clearly. T ?. F' 71-
l.am alalkative ind Vidual. T ? F. 72
No thoughts are b sically evil.. 73
It is.easy for me t takerorders and do what I am told. T ? F 74
Life would'be more meaningful if we had more leisure time. T ? F 75 '

There are times when I get a notion of making a mess o9 1111111111

things and being destructive, such as having the urge to
throw mud on a freshly painted wall. T ?
Perfect balance is the essence of all good composition.
No sane, normal, decent person could ever think of hurting
.a close friend or relative.
I would especially like to be a psychotherapist.
I enjoy mimicking or caricaturing people-and their manner-
isms, or enjoy watching others do it if I am not good at it
myself.
I am inclined to be shy in the presence of the opposite sex.
Regardless of whether it is necessary or not, it is better
to have a clear schedule or plan for each day's activities.
I would feel like a fool if, like I put a lot
of energy into entertaining peopl I hardly know.
Sometimes) have a vague feeling of anxiety as if I had done q
wrong and would be found out.
Good pay, with small Opportunities to learn over the next five
years, is preferable to small pay, with large opportunities to
learn aver the next five years. T F 10
 - 125
, APPENDIX 3 .

The tyPe of humor whiCh is based on the fantastic, bizarre,

or Impossible has little appeal for me. 11,

I would rather do house-to-hoine canvassing than gardening. 12

I would have resentedit Hilly parents tried to make me con-
form to a-certain pattern'of behavior.
Life is moat enjoyable when it is filled witkuncertainty
about what is-corning next.
I -sometimes imagine what it would be like to do the impos-
sible such as being a member of the opposite sex, living in
another era, flying, etc. 15

The world has treated me at least as well as it has-treated

most. others I know. 16

A man who does not believe in some great cause has -not
17-
really lived.
I feel remorse when I think of some of the things I have
F 18
dope.
I would rather do Work involving m'any details than work
19.
involving few details.
-
.0f all the different philosophies which exist in this world,
there is probably only one which is correct. 20
I am not prone to think things thiatotake me laugh or chuck-
.F 21
le to myself.
I become resistant when others attempt to influence me. 22
23
.1 prefer conceited people to jealous people.
I am able to smooth out tangles and disagreements between
people.
I remember rather accurately a good many fairy stories and
** nursery tales.
I prefer playing safe-to taking .a chance. J.

It's only when a person devotes himself to an ideal or cause

that life becomes meaningful. T
Rather than nights spept at home, I like nights away from
? F
homd
Occupations; Indicate whether you would like that kind of-Work. (*regard Salary, social stand
ing, and future advancement. Cbbsider only whether you would like to do what
is involved in the. occupation. -

'Draw a circle or check: L if you would like that kind of work.

if you are indifferent to that kind of work.
D if you would dislike that kind of work:
Work rapidly. Your first impressions are desirable.

. Auctioneer
Author of a novel:
Building ContractOr
Carpenter
 EXPLORING THE PSYCHO-SOCIAL THERAPIES
-

Cashier in a bank D 33
Electrical Engineer D 34
Employment Manager D '35
Foreign Correspondent L D. 36
-Floorwalker L. D 37.
interpreter L. 38
Jeweler L D .39
Labor Arbitrator L 40
Machinist L D 41
Marine Engineer L. 42
Medhanical Engineer D 43
Office Clerk. L D 44
Photoengraver D 4,5
Poet L 46
PriVate Secretary L D 47
Retailer L D 48
Ship Officer L I D 49
Specialty Salesman p 50
Surgeon L D 51
Toolmaker L D 52
Wholesaler L 53
Sotioortik ects: Atiriculture L p 54
Algebra L. .D 55
Rookkeeping L i D 56
Geometry L I D 57
Literature L I D .58
Manual Training L I D .59
Mechanical Drawing L 1 ~= D 60
Amusements: Solving meeitienical puzzles L 61
Performing sleight-of-hand tricks 62
Drilling in 'a military company L 63
Convention L l D 64...
Musical Comedy 65
Detective stories .L I D 66
Social Problerri Movies L D 67
Making a radio set' D 68
Activities: Repairing a clock L D 69
Adjusting a carburetor L 70
Repairing electrical wiring L D 71
Cabinetmaking L' I D 72
Being called a nickname .1 D 73
Looking at shop windows 74
Expressing judgments publicly D .75.
regardless of criticism
4 .1 131
Looking at a collection of antique furniture L. . 1.
Peculiarities of .Absentminded people D
people: People who haVe made fortunes in business L D 3
Foreigners 4. D 4
People who talk very slowly D 5

143
 APPENDIX 3 127

-People who chew gum 6

7
AthietiC men

If you would like to know your scores on this inventory check he're 8

About your parents when you were 10 to 15 years of age:

(Please be as specific as possible.)

Your father's oCcuPlition:

Your mother's occupation: 1.1 -12

Father's Mother's
Education (Check appropriate boxes) EduCation
No formal education !.
Eli Primary school attendance
1
El Secondary school attendance
ci Secondary school diploma 14

113 College or university attendance El 15

College or university degree 6
Specialized education. Please specify
what kind:

When you were growing up, did you ever know a person who had a. seriouj mental
disorder? I No I

(If yes) How well you you know this person?

Friend, relatlive 1 Close friend, relative 16
ICasual ac9aintancW1
Did you have occasion to see this person at times when he was suffering from
his disorder?
No Yes, occasionally I Yes, often

About your college education (undergraduate):

Number of college yeari, or parts of years completed 18

Major field of study 19

Minor field of study 20

Your vocational objective in college 21.

Your age in years at last birthday Sex: 1 Male I Female 22--

23

4c
 e

Appendix 4
Peitsonality Scales in the PTQ
Tables 31:to 50 present the items In each .of 20 'personality scales con-
tained in the PTacThe Internal, consistency of these scales is shown in
table 6.. "- ... .

The A-B predlOtors and clusters in the PTQ are Shown in tables 10 to 13, .

and table 17. The A-B predictors are not internally Consistent, as discussed
o'h pages. 10-19. The 'internal consistency of the .A-B clusters.ispresented
on page 109.

Table 31. The Need for Closure

The five Items are from Irvin L.Child.0985), "Tolerance of Complexity." Of his 20-items, 18 are included in the question-
naire. Qf these, 5 are from the "F Scale " (see "Antidemocratic", table 41) and are scored there. Six other items were,
found not to correlate well with the total scale and weredropped. ,

r with'
Item Total Score
A

1-7 It's a gad thing for a teacher to leave basid problems

unresolved, so that students have, to figure out for
themselves whether there is a clear answer.. (True) .66
1-23 insofar as the study of philosOphy makes one' doubt his' basic
beliefs, It should be encouraged.-(True) --.60
2-32 . Wild and Impractical Ideas are often the ones that are most
worth pursuing. (True) .56
1-27 Those religions are to be 'most respected which impose no
uniform beliefs on their members..(True)

1-47 Arguments about irreconcilable differences such as contra-

dictory religious beliefs or basic valuesare more interest-
ing than arguments about points on which one person is able
to persuade the others to agree with him. (True) .52.
. Thls sCale has much,the samemeaning as Child's. Tolerance of Complexity. But the content is somewhat more Intellec-
tual. Accordingly, it was retitled the Need for Closure (With signs of .the items reversed). The items excluded are as
follows: 1-3 The leaderof a'group is likely to be most effective if he acts on the assumptidothat everyone in the group is
glad to have him as:afitader. 1.57 At the end: of a textbook chapter, a good summary is more valuable than a set of
thought-provoking questions. 2-28-Unquestioning loyalty is the first requirement of good citizenship. 2-42.The man writ,
truly loves a woman must regard:her asthe'best in the world In every Important remierct. 2.52 The more you get to know
and like a person, the more you are tiWare.O1 his weaknesses and fail) . 2-82 We may cortUdently expect.that mankind
will someday attain a stable social. order In Which :Marital infide will be unknoWn...

128 tJ
Vr
 .

APPENDIX 4 124'

-Table 32. Regression in the Service of the Ego

This unpublished scale was developed, by .David Singer: Some of its characteristics in a college population are
presented by. Child (1065). Of 30 Reins, 27 were InclUded in the PTO.Of these, 3 were found not to correlate with the total
score even wheri they were part of the score, and these 3 were dropped. TWo other Items were removed because they
were badly needed in the. Empathic Interest SCale (see the next table):

with
Item Total Score

2-66 I sometimes have (*dreams in which I become.a "heroic" type

of figure; either all-powerful, all-knowing and successful
or someone who has sunk to the.lowest depths of dePravity,
weakness and suffering. (True) .58

1-25 I enjoy letting my thoughts wander aimlessly; and find myself

thinking about all sorts of unusual and unrelated things. (True) .53

3-1 There are tines when I get a notion of making a mess of things
and being destructive, such as having the urge to throw mud
On a freshly painted wall. (TrtLe)

1-61 'There are times when I feel as a 1 were a small child-again,

or when I wish I could be one for a short time. (True)
. .

3-11,1 The type of humor which,is based on the fantastic, bizarre, or

impassible has little appeallor me: (False) _.50
`3-15 I sometimes imagine what it would be like to do the impossible
such as being a member of the opposite sex, living in another'
gra, flying,"etc. (True) ,, .49k
,
, . ,
2-6 When I am alone or per4orming a task recNiring n_o mental work,
I enjoy doing things like rhVming words, discovering puns,'
and making Up little songs. (True)'

1-71 I rarely let my imagination wander to the where I think`

things whiclii are so unusual, sexual, violeht, or irreverent
that they would be disCusied only with close friendsif. -
anyone. (False)
,

2.36, 1 cari.detect'in mysetif no strong. antisocial impulatia of the

sort-whicli,.Under certalin circumstances; mightleacitO
. crime.
.(False). .44
:2,56.. ,.my enjoyinent crf,musiclind.art takes.the form of quiet pleasure,
and I.Seidom have .%the,Intseemotional reactions to them.
'thatsame .people.do. (FaIS ) .44.

Continued on the, next. page:

180 EXPLORING THE PSYCHO-SOCIAL THERAPIES

Table 32 (concluded)
r with
Item Total Score

in listening to a lecture, I often.am amused by thoughts of double

meanings or possible puns which the lecturer probably doesn't
intend. (True) .40

I enjoy mimicking or caricaturing people and their mannerisms,

or enjoy watching other "t if I an' not good at it myself.
(Tiue) .40

2 -16 I get little pleasure fun out of playing with words and
languageas b lking nonsense, baby-talk or in a foreign
accent and seldom do that sort of thing. (False)

2-46 While watching a play or listening to music I am often completely

carried away, becoming totally absorbed and feeling as if I were'
a part of the play or the music. (True) .38

3-21 I am not prone to think things that make me laugh or,chuckle

to myself. (False) .38
1-51 I prefer not to associate with the kind of people who sometimes
get silly or childish. (False) .34
2-61 I do-not tend to develop attachments to inanimate things or
objects (such as autos, houses, tools, etC.) to the point where
it almost seems as if they.had a personality or were alive.
(False) .33
1-55 When at a boring lecture, while I may get restless I seldom get
the urge to set off a firecracker, jump up-and down, scream -
"Fire" at the to of my lungs, etc. (False) .31
1-11 I have often thought at some length the various ways I would be
affected if my nearest relatives were suddenly, to die. (True) .31

2-10 I seldom become enthusiastic over new ideas or experiences the

way some people do; rather I tend to take these things in stride.
(False) .30.
3-25 I remember rather accurately a good many fairy stories and
nursery tales. (True) .26

3-40 I rather like floating in water, ffor the pleaiant sense it gives
you-of your own identity as against the formless mass of sur-
rounding water. (True) .20

items dropped were 1-21 and 1-35 (see bottom of table 33) and 1-1 When I am part of a team or group that is working or
playing together I prefer not to get so involved and caught up in the activity that .1 lose my feeling of separateness.
 APPENDIX 4 181

Table 33. Empathic Interest- j

The empathic interest Items are from a variety of sources. Four of the original nine items are from Regression in the Ser-
vice of the Ego (table 32). Of these, two 'wew found to have low correlations with Empathic Interest (and also with
t Aegreasion) and these two were dropped (see below). Two other IteMs-marked with an asterisk(!) correlate with both
Males but are Inciiicled In Empathic Interest becalm it is a weaker scale, psychometrically, than Regression.
r with
Item 'Total Score'

'1-31* I like the sense of privately reeling my way into some .60
Sympathetic participation in every kind of human emotional
experience I hear about, regardless of how tragic or unsavory
some of them may be. (True)

1-43 I am very sensitive to the emotional attitudes people some:, :" .49
times want to convey but are unwilling to state openly. (True)
n
.2-2 Sometimes I feel I 'earn almost as much from watching people's .41
faces and their hands while they talk, as I do from listening
to what they say. (True)

1-54 it is very interesting to hear about, or,to observe, an .37

intensely emotional experience of someone I know. (True)

1-58 It would not be very interesting to try to fe.el one's way .37
into what the experience of a blind person is like; it would
be better not to think about it unless you had to. (False)

2-30 A person should not probe too deeply into his own and other .37
people's feelings, but take thingsias they ate. (Fal )

2-26* When prying with very young children, I find it eas o get .36
into their world and experience things as they do. (True)

The items excluded are as follows: J.21 While listening to a friend recount an experience, I very seldom emphathize to
the point of feeling what he must have felt in the situation he is describing. 1-35 In trying to understand, what another
person is like, [find it best to put tittle emphasis on the feelings Irgef 'when I am with that person.

3.

1
 i
132 EXPLORING THE PSYCItO-SOCIAL THERAPIES

Table 34. The Work Ethic

Mire Is and Garrett (x971) developed 19 Items measuring the Protestant Ethic. Some of the Items concern "spendthrifts"
and are not included here although this aspect of the-Protestant Ethic would be of considerable interest(see page 162).
Seven items relating to work and leisure were Included In the PTO. Of these, fwo were dropped because of low correla-
tions with the scale.

r with
Item Total Score

2-43 People should have more leisure time to spend in relaxation. (False) .63
2-75 Life would be more-Meaningful if we had more leisure time. (False) .60
1-46 Our society would have fewer problems if people had less
leisure time. (True) . 59

1-20 The man who can approach an unpleasant task with enthusiasm
is the man who gets ahead. (True) .53

1-24 If one works hard enough, he is likely to make a good life for
himself. (True) . 52

The items excluded are as follows: 2.39 A,distaste for hard work usually reflects a weakness of character, and 2.66 Herd
work offers little guarantee of success.

Table 35. The Need for Order

Of th,e five items, two are new and three are frOrn Child's Tolerance for Ambiguity. (1965)

r with
Item Total Score

1-4 I prefer that my hours of eating and sleeping be regular, .65

not changing from day to day. (True) .

2-19* The challenge of novelty and excitement in a completely =.57

new situation is worth the price of disrupting one's old
established ways. (False)

1-34 When I go outin theevening, I like it best if all the . 53.

activities are planned. (True)

3-7* Regardless of whethe; it is necessary or not; it is better .48

tip have a clear schedule or plan for"tach day's activities. (True)

3 -14' Life is most enjoyable when it is filled th uncertainty .48

about what is coming next. (False)

From Child's Tblerance for Ambiguity. Three other items from this scale were included in the PTQ,.but the two triads
ncorrelated, and the second triad has no-internal consistency for our sample of.133 mental health professionals.
3e latter items are: 2-23 Clbuds that are frankly clotids and cover the whole sky are preferable to the little floating
. 0.10s that leave you never knowing whether the next moment will be bright or dull, 2-41 It would be exciting to arrive in a
city for the first time. and find it enshrouded in a head fog, and 2-.65 Even if you were an expert in all the .relevant
-ges, it would be disquieting to be in a multilingual group where you would never know what language to use next.

1
 APPENDIX 4 133

Table 36. Tolerance of the Unrealistic Experience

Of Child's 12 items (1985), 9 were included,in the PTO. Of these 2 were dropped from this scale because of very low cor-
relations with the scale.
r with
Item .Total Score

2-69 Regardless of what caused it, dizziness would just in itself .61
ge a very disturbing experience. (False)

2-9 Quite apart from consideration's of morality or- prudence, .59 ,

the confused sensations of extreme drunkenn'ess would be a
very objectionable experience to go through. (False)

2-64 Optical illusions and other experiences that put you in con- . 56
flict about what is real and what isn't are on the whole quite
enjoyable. (True)

2-59 If there were a harmless drug whiCh would temporarily make .56
one's 'Sense of smell as sensitive as that of a dog, it would
be great fun to try taking some. (True)
4

2-29 A slow-acting anesthetic, which would provide a period of . 47

relaxed, hazy, mixed-up sensations would be- preferable to an
anesthetic which would put you under instantaneously. (True)

1-64 Fun-house mirrors that colopletely distort your body-shape .46

provide an interesting experience. (True)

2-49 To be wandering around a familiar city and suddenly realize that .43
you are lost, would be an unpleasant experience,l(False)

The Items excluded are as fellows: 1-58 It would not be very interesting to try to feel one's way into what tl(e experience
of a blind person Is like; it would be better, not to think about it unless you had to, and 2-71 Playing with words, as In pun-
ning, ought to be avoidecysince it interferes with the normal use of words. for communicating ideas clearly. This last
item was found to correlate highly with the F Scale and is therefore included in the Antidemocratic scale, table 41.

ti
184 EXPLORING THE PSYCHQ-SOCIAL THERAPIES

Table 37. .Tolerance. for Ambivalence

Of Chli id's 10 items (1965), all were includecHlowever, 3-of.the 10 are.found to have.low correlations with the scale in our
sample of 133 Mental health professionals and they have been dropped from this scale.
r with
Item Total Score

2.24 It would be difficult to maintain a warm friendship with a .54

person who has'sdrne markedly disgusting characteristic. (False)

2-37 The expert ski jumper should enjoy the sport all the more if .51
it remains a source of tension and even alarm. (True).

2-17 It would be hard to have complete respect for someone as .50

leader of your group if you have any really fundamental
disagreement with him. (False)

1-29 Once a husband and wife have contem ted divorce; they can .49
probably never be truly happy with one nother again. (False)

1-32 Even the strongest love for a person is entirely compatible .41
with the presence of a variety of negative feelings toward
the same person. (True)

2-54 In watching football, a close game in which your side finally

loses is more enjoyable than a game in which your side is way,
ahead almost from the start. (True)

1-72 A person will get along better with his close relatives if he .30
recognizes that tie resents them as well as loves them. (True)

The items excluded`are as follows: 1-52 A beautiful sunset would be still more beautiful if it lasted longer, and were not a
tragic reminder of how transitory everything good is, and 2-45 The custom in some places of celebrating a funeral with a
drunken feast is disgusting to anyone who respects the dignity of human sorrow. This last Item was found to correlate.
highly with the F Scale and is therefore included in the Antidemocratic scale, table 41.

Table 38. Preference for Decision Making

Of 11 items, Child etal. (1969), 4 are included. _

rwith
Item Total Score

1-60 I believe in individual religion rather than institutionalized = .61

religion. (True)

1-40 I prefer deciding what_is morally right myself, rather than .61
accepting the moral code of society or institutions. (True)

-2-25 I prefer a job that requires making difficult decisions .56

rather than following instructions carefully.. (True)

1-13 I feel very insecure about reaching any decision of .51

importance entirely on my own. (False)

5
 APPENDIX 4 135

Table 39. Personal Standards

-Thls item, written thr the PTO, was thought to be correlated with Remorse, tabla,40. It does not so correlate, so it stands
alone in analysis.
Item
r 4

2-58 I tend to set, very high standards, for my own performance.

Table 40. Remorse

For' his "Superego" scale, Child (1965) developed 20 itema. Of these, 9 were included in the PTQ. Of the 9, 2 were
dropped because of loW correlations with the scale for our sample of 133 mental health professionals. It is doubtful that
the remaining .7 are representativeof Child's Superego scale, and a new name was Invented.
r with
Item Total.ScOre
lir .64
3-18 I feet remorse when I think' of some of the things I have done. (True)

3-9 Sometimes I have a vague feeling of anxiety as if I had done .56'

wrong and would be found' out. (True)

1-38 It is better to let the past be forgotten; tote remorseful .50

is a waste of time. (False)

2-11 Sometimes I ruminate about things I have done wrong. (True) , .44

1-8 When I choose to do something, the fact that it may not .42
be allowed is relatively unimportant. (False)

1 -49' It is not possible to knOw when one is doing good or evil. (False) .34
1-g8 The real fault is to have faults and not try to amend them. (True) .32

The items excluded are as follows: 2-33, I feel that since you only live once, you should.enjoyyo-urself, and 2-73, N
thoughts are basically evil.
136 EXPLORING THE PSYCHO - SOCIAL THERAPIES

Table 41. Antidemoc Scale

(not a manifest. scale)
Most of the Items in this scale are lrom the F Scale which was denielobed by Adorng, Frenkel-Brunswik, Levinson, and
Sanford (1950) as an indireet measure of ethnocentrism. The F Scale has been lound to be strongly correlated. with at-
titudeatoWard the serious mental disorders, especially "custodialiSm" vs. uhumaniam.'-' (Greenblatt et al. 1957, pp. 26,
206, 224, 363, and 374j There are a large number of variants of this scale which are rlo4 necessarily comparable. From the
original 42 -Item scale, 14 items were Included in the PTO of which 5 were also in Tolerance of Complexity (table 31) and 4
were also in independence of Judgement (table 42).
Mental health professionals score very low on this scale. On the 14 items, the average mental health pilbfessional
responds &II an antidemocratic way to only 2, even though these items are the less exteme ones of the original 42. This
extreme skewness poses measurement problems. For this reason it seemed desirable to Include as many items as
possible. The overlapping items were all assigned to this scale rather than to the other two scales. Additionally It was
found that 3 other items correlated highly with the 14 and were not well correlated with their scales. One of the 14 Items
was dropped because it did not correlate well with the others.

r with.
Item Total Score

2-45* The custom .in some places of celebrating a funeral with a .67
drunken feast is disgusting to anyone who respects the dignity
of human sorrow. (True)

3-3 No sane, normal, decent person could ever think of hurtind .59
a close friend or relative. (True)

1-33 If people would talk less and work more, everybody would be 8,
better off. (True)

2-44 If young people get rebellious ideas, then as they grow up

they ought to get over them and settle down. (True)

2-71* Playing with words, as in-Punning, ought to be avoided, .54

since it interferes with the normal use of words for communi-
cating ideas clearly. (True) .-

1-37 Most of social problems would be' solved if we could .53

sornehbik.get rid of the immoral, crooked, and feeble-minded
people

1-74* The worst crime a person could commit is to attack publicly

the people who believe the. same thing he does. (True)

Jot

continued on next page

 APPENDIX 4 137

Table 41.(concluded)
r with
Item Total Score

1-9 Human nature being what it is, there will always be war .43. .

or other forms of serious social conflict. (True)

1-67 Obedience and respect for authority are the most important .42
virtues children should learn. (True)

1-5 What youth needs most is strict discipline,i rugged - .41

determination, and the will to work and fight-for family and
country. (True)

2-70 What this country needs most, more than laws and political .39
programs, is a few couragepus, tireless, tlevoted leaders in
4
whom the people can put their faith. (True)
,
1-17 People fall vdry naturally into distinct classes, such as .39.
the strong and the weak. (True)...

2-18 An insult to our honor should always be punished. (True) .32

2-22 Every person should have complete faith in some supernatural .32
power whose decisions he obeys without question. (True)

2-12 No one can be sure of conquering. his difficulties; will- .24

power is not enough. (False)

2-53 Nobody ever learned anything really important except through .23
suffering. (True)

Item .1-63, When a person has a problem or worry, it is beet for him to think about it rather than turn his attention to more
cheerful things, was dropped because it did not correlate well with the other items.
Item 2=45 is from Tolerance of Ambivalence (table 37) 2.71 from Tolerance of the Unrealistic Experience (table 36) and
1.74 from Dogmatism (table 48).
138 EXPLORING, THE PSYCHO-SOCIAL THLRAPIES

Table 42. Individualism

' Ine t a manifest scale).
Barron (1953) developed a scale of Independence of Judgement by validating items against not "yielding" in asocial
pressure experiment (Asch, 1955). Thesittems, then, have behavioral validity not necessarily manifest validity. In his
item pool Barron included the items of the F Scale and a number of these items predicted "-not yielding,'' all but one of
them (1-9) being scored in a reverse direction from that in the F Scale, table 41.
Of Barron's 22 items, 19 were included in the PTQ. Of these, four are from the F Scale and are included in "Antidemo--
cratic," table 41. Six other items did not correlate well in our sarnpte- and: were dropped. Individuatsm is considerably
more manifest' in this scale than in Barron's.

r with
Item Total Score

1-45 I must admit that I woulci.find it hard to have for a close

,.
.46
friend a person whose manner or appearance made him some
what repulsive, no matter how brilliant Or kind he might be. (False)

1-39 The best theory is the one that has the best practical appli- --.45
cations. (False)

1-69 The unfinished and the imperfect often have greater appeal for :43
me than the completed and polished. (True)

3-2 Perfect balance is the essence of all good composition. (False) .42
14.5 , Some of my friends think.that my ideas are impractical, if . .41
not a bit wild. (True)

1-19 I believe you should ignore other pkple's faults'and make

an effort to get along with almost everybody: (False)

2-20 Kindness and generosity are the most important qualities for 39
a wife to have. (False)

2-30 A person shoul not probe too deeply into his own and other .39
people's feelifr, but take thingsQas they are. (False) .

2-50 I prefer team games to games in which one individual.. competes .39
against another. (False)
The items that were excluded are as follows: 1-75, I don't understand how men in some European countries can be so
demonstrative to one another, 2.4, I would rather have a few intense-friendships than a great many friendly but casual
relationships, 2-14, Science should have as much to say about-moral values as religion does (See table 45), 2-34, The hap-
py person tends always to be poised, courteouS, outgoing, and emottonally,controlled, 2-60, I could cut my moorings
quit my home; my family, and my friends:-without suffering great regrets, 2-74, It, is easy for me to take orders' and, do
what I am told (see table 43). .
 APPENDIX 4 139

Table 43. Deference Anxiety

Judgement was added
Of nine Items (Child et al. 1969), six were included in the PTO. One Wen', from Independence of ::
because of Its high correlation with the six, Items. . .

- . r with

Item Total Score

2774* It is easy for me to take orders and do what I am told. (False) .65
The thing I would particularly hate about military service is .56
1-53
the ee-qUirement of obeying orders of my immediate sliperior. (True)

1-73 1.conform to conventions even when I don't

4
like them. (False) .50 4

I would have resented it if my parents tried to make me con- .49

3-13
form to,a certain pattern of behavior. (True)

I become resistant when others attempt to influence me-. (true) .49

-3-22
I r.esent anybody whose position is superior to mine and who .47
2-38
can exert a certain authority over me\ (True)

I am unable to work efficiently when'l am in a subservient' .42

-2"8
position. (True)

Item .2-74 is from Independence of Judgement (see table 42).

Table 44. Nurturance Anxiety

Of 10 liems (Child et al. 1969), 6 were included, but only 3 were kept.
r with
Item Total Score

I get annoyed when people I hardly know greet. me as if they, .69

1-65 _

were old friends. (True) . .

.,. 1-12- . I enjoy putting my own affairs aside to do someone a favor. (False) .69
I get annoyed when a stranger talks to me on the bus, train .54
2-63 i.
.
0
$,

or airplane. (True)
don't want pe ple to be
The items that were excluded are as follows: 1-44, l'refrain from giving advice and help because I
feel like a foi if, like=
dependent on me, 2-21, I tend to regard little children as principally 'a nuisance, and 3-8, I would .
some persons, I put a lot of energy into entertaining people I. hardly know. . .

Table 45. Science.

This item is in the Independence of Judgement scale (see table 42). For oul- sample, it does not correlate wellwith the
Dent and Kwiatkowska, 1970). It was
scale. On the other hand, in item analyses it frequently discriminates (e.g.,
therefore continued in the analyses as a single item for exploratory purposes.
Item

2-14 Science should hav as much to say about moral values as.
religion does.
 1.40 EXP RIN THE PSYCHO-SOCIAL THERAPIES

`Table 46. Preference for Intellectual Challenge

Of 19 Items (Child at al., 1969), the 3 Items best correlated with the total score were included in the PTO.

r with
Item Total Score
2-15 I get a kick oyt of trying to solve a puzzle.even when I fail. (True) .75

1-30 I would rather spend an hour solving several easy mat prob- .70
lems than solving one t!ard one (False)
1-10 Knowing that something will be very hard to understand .66
makes it more interesting to me. (True)
I.

Table 47. Extroversion..

Child et al. (1969) deyeloped 20 items which they called Affiliation or Sociability. Of these, 7 were Included in the PTO.

r with
. Item Total Score
1-36 am inclined to keep in the background on social occasions. (False) .78
% -
. 1-14 am iitlined to keep quiet when out in a social group. (False)
2-72 am a talkative individual. (True) , .67°

2-1 like to have 1-nany social engagerlients. (True) ... .65

% $

2-51 do not like to mix socially with people. (False) --.54

3-6 I am inclined to be shy inlihe preence of the, opposite sex. (False) .54
2-48 It wouldn't make me very unhappy if I were prevented from .49
making numerous social contacts.-(False)
 APPENDIX 4

Table 48. Feivor.

itsTelation to the F Scale (Kerlinger
Rokeach (1960) has developed several forms of a "Dogmatism ".'scale, and explored
thesewas more highly correlated with the F
and Rokeach, 1966). Onlyelght lima were included in the PTO _and one of also did not correlate well and
Scale than with Dogmatism. Therefore, it.was transferred (see table 41). Three other items
'were ditipped. The remaining four Items all have a donee of commitMent, or fervor...,
A

r with
item Total, SCore

. .78
3-27 It's only when-a persOn devotes himself to an ideal or cause
that life becomes Meaningful. (True)
.74
.3-17 A man who does not believe in some great cause has not really
lived. (True)

1-66 If given a chance, I would do something of great ben -:air o I .59

the world. (True)
.50
2-57 It Is only natural that a person\should have a much better
acquaintance with ideas he believe in than with ideas he
opposes. (True)
When it comes to el:difference of 6pinion in religion,. we must be
The items that were. excluded are as follows: 2-67, waj4ive do, 2-47, The present is all too full of
careful not to comproinise with those who believe differently from the which exist in this world, there
unhappiness. It Is .onlythefuture that counts, and 3-20, Of all the different philosophies
. Is probably. only one Which is cfirrect. .

Table 49. Life Satisfaction

hese items were written for the PTO:
r with
Item Total Score.

1-16 My accomplishments in life are about as high,as my expecta-

tions. -(True)
.72
1-6 What I have hoped, for in life generally is coming to me. (True) .

.! .
.
.70.
1-26 I think that my future will fulfill my aspiration,s. (True) ..
> - .,
-7.59
1 -56 Sometimes when I think about how much other people have, or
how much, they. have accomplished, I feel ashamed. (False)
0
316 The. world has,treated me..atleast as well as it has treated',
most others I know. (True)
 142 EXPLORING Ti'PSYCHO- SOCIAL THERAPIES

Table 50. Work. Activity

Child at al. (1969) developed .12 items representing Preference for Work. Three of these work activity items were in-
cluded in the PTO to provide a contraet with thit Work Ethic (table 34) in our efforts to understand the meaning of rejec-
tion of manual and mechanibal activities among "A" therapists.

r with
item Total Score
2 -5 prefer to. take an elevator rather than to climb a flight .78.
of stairs..(False)
2-55 I. would rather cut down weeds with a scythe than use a movVing .70
machine. (True)

1-50 I would rather,open a. heavy door myself to having it open

automatically..(True)

:v
 7
Appendix 5 5

Supplemental Analyses of the

Phipps Clinic Data
This appendix presents some analyses not tioti of SVIB items. In the original instrument
essential to the content of chapter IV, but there are some blocks of 10 which the respon-
which provide some .eXplanation for the aent .is:: asked to rank. In some cases an-item
methods that were. used. can 4e extracted in a meaningful way without
Atithe time the'PTQ was constructed, I did reproducing the entire 10 items. In general,
not haVe data on which, Phipps residents. had - such items are not inclUded if it would'recluife
prescribed drugs. tried to approximate the including the entire block of items.
no-drug group by selecting for item analyses One such item is among the original 23 of
the 47 therapists who completed their resi- Whitehorn and Betz: Interest the public, in a
dencies before 1959. TableS 51-54" present the new machine through public addresses: When
item analyses for these 47 therapisti.' These . they reported their findings, they did not-
tables explain how SVIB items were accepted present all the. detail concerning the 23 items
for inclusion in the PTQ. and many investigators have taken the "A-B
It seemed appropriate to accept about 80 Scale" from .the Whitehorn-Betz reports
items into the PTQ. The .04 level of without looking at the original SVIB. Nowe this
significance yielded about this number after item is one of 10, all concerned with the "new
allowance for other relevant considerations: (1) Machine," and this Is about the, least
Nearly all the original 23 items were included mechanical of them all. In'other words, accep-
*regardless if whether they were significant in tance of thil item in the original SVIB means
the new item analyses. (2) A few items were in-. rejection of the new machine, When Ai -item
serted to break up a long run of A-therapist stands alone, it is likely to reflect acceptance of
dislikes. (3) Some items significant at .04 were the new machine. (One group of psychomefri-
excluded because they are dated, very skewed % cians found it carried the wrong, sign in their
in their -responses or. redundant. data so they inverted it!)
There is one other consideration in the selec-

+4.

162
 EXPLORING THE PSYCHO-SOCIAL THERAPIES

Table 51. SVIB items COrrelated With Therapists' Percent Improvement

in Patients Who Are Schizophrenic, Depressive, or Neurotic
(Doctors who completed residency before 1959)
N =48 N =42
Item Number Schizophrenic Depressive Neurotic
SVIB PTO Item Gamma p Gamma p Gamma p

..- _
19. 3-32 Carpenter** -.49 .001 -.51 .002 -.47 .02
56 3-41.Machinist -.37 ,.02 ( -.37 .07) (.-34 .07).
60 3-43 Mechanical Engineer** (-.29 .06) (-.29 .06) (-.33 .08)
94 3-52 Toolmaker** .--.45 .003 (-.29 .07) (-.34 .06)
188 3-71 Repairing Elec. wiring -.36 . .03 -.38 .02 -.42 .05
310 John Wanamaker, Merchant -.37 .03 -.40 .03. -.64 .001

38 3 -37 Floorwalker -.54 -.03' -.62 .02

71 .3-46 Poet .34 .03 .30. .05
.155 Excursion ( .36 .08) .47 '.02

87 3-49 Ship Officer*, (-.28 .08) -.38 .05

130 Physiology ( -.64 .06) .68 .02
132 * Shop work (-..29 .06)
t
-.39 .04
332 Immediate vs. 5 years(A) (-.26 .09) -.40 .05
359 3.23 Jealous vs. conceited(A) -.37 .03 ' ,. -.54 .01
, .
.
90 3-50 Specialty Salesman** (77.33 .07). (-.34 .09)
.105 3-56 Bookkeeping --.38 .04 -.51 .01
119 .3 -58Literature -.49 .04 .67 .007
129. * Psychology .54 .08) .59 .04
148 3.61 Solving mechanical puziles (-.26 .10) -.43 .03
149 3-62 Performing sleight-of-hand tricks -.46 .003 -.37- .05
190 Operating machinery , .-- (-.28 .09) .(-.35 .09)
279 4-6 People who chew gum .51 .0Z ( .34 .15)
-280 4-7 Athletic men -.41 .03 ..50 .0f
316 '* Chairman, Educational Committee .34 .05 ( .40 .06)
340 3-10 Small pay, large oppor..vs. Good pay(B) .51 .02 *.57 ;. .02
I
I
In parentheses are relations between the .05 and the .10 level.
*Not Included in the PTQ; see footnote to table 53.
**Original 23 A-B items.
(A) The A's prefer the second choice.
(B) The B's prefer the second choice.
One of the doctors did not have enough depressive patients to give a reliable improvement rate; five of theth did not
have enough neurotic patients.
 APPENDIX 6 145

Table 62. SVIB !toms Correlated With Therapists' Percent Improvement

in Patients Who Are Schizophrenic
(Doctors who completed residency before 1959)
N =47

Item Number Gamnia p

item
13 Auto Repairman -.31 .04
33- 3-35 Employment Manager
33 -.34 .03
44 3;38 Interpreter .32 .04;
46 3-39 Jeweler -.36 .03.
.33 .03.
158 3-64 Conventions
Full-dress affairs .32 ..04
159. :

.41 .01
170 Snakes . .
- .40 .04
184 3-67 Social problem movies
189 3.-72 Cabinetmaking** -.35 .03
195 Arguments . -.34 .04
.35 .03
221 3-75 -Expressing fudgments klublicly regardless of criticism .40 .02
297 ' Opportunity to understand just how one's superior...
.

.48 .02
311 1-18 President of a society or club** -.30 .05
314 Member of a society or club .04
Chauffeur vs. chef (B) .32
323 ,

.40 :008
334 3-26 Taking a chance vs, playing safe (B) -.32 :.03
341 3-19 Work involving few details vs. many. details (A) -A7 .004
352 3-28 Nights spent at home vs. away (A) -.37 .04
355 2- Few intimate friends vs. many (A)
Accept Just criticism without getting sore** .50-
367. 1-
Able to meet emergencies quickly and effectively .58 .006
376 1-62 .33 .03
383 1-68 Stimulate the ambition of my associates .04
387 * Am approachable (43-4-0)*.** ,

Between the .05 and the .10 level: .45 .10

9 3-30 Author of a novel # .39 .06
40 3-36 Foreign Correspondent -# .26 .10
93 3-51 Surgeon # -.44 --07
136 Zoology .31 .06
138 Fishing
.28 .08
139 Hunting .42 .08
140 Tennis .25 .10
147 Observing birds .34 .08
162 Animal zoos (opposite in sign to 136 above)
Handling horses .30 .06
191
Adjusting difficulties of others (43-4-0)*** .58 .09
209 .36 .06
232 Looking at a collection of rare laces .06
Freedom to work at one's own methods of doing the work .35
299 -.29 .07
364 Usually get_ other people to do what I %I:sot...done
Not Included In PTO; see footnote table 53. OrIgInal 23 A-B items.
Figures In parentheses indicate distributions skewed toward "T."
I Included to break-up a list of Items all disliked by "A" doctors.
(A) The A's prefer the second choice. (B) The $'s prefer the second choice.
146 EXPLORING THE PSYCHO-SOCIAL THERAPIES

Table 53. SVIB Items Correlated With Therapists' Percent Improvement

in Patients Who Are -Depressives
"stators who completed residency before 1959}
N = 46

jSVjem N pmber
IB PTQ Item Gamma p
17 3-31 -Building Contractor** -.39 .01
42 * Hotel Keeper or Manager
48 3-40 Labor Arbitrator
-.34 .04
.33 .04
64 3-44 iffice Clerk -.74 .004
74 3-47 rivate Secretary -.40 .03
80 .3-48 Retailer -.39 02
99 3-53 Wholesaler -.43 .01-
102 3.54 Agriculture -.35 .04
153 Amusement parks --.31 .05
166 3-65 Musical Comedy -.56 .04
205 3-73 Being called a nickname -.36 .03
21.5 Writing reports -.35 .03
233 4-1
245 4-3
, Looking at a' collection of antique furniture -.33 .04
People who have made fortunes in business. .008
260 * Side-show -,freaks -.41 .02
326 3-12 HOuse-to-house canvassing vs. gardening .(B) .51 .02
338 Work in a large corporation with little chance of
becoming president vs. work for small business (B) .32 .04
3.73 2.3 Am always on time with my work -.45 .01
Between the .05 and .10'level:
25 Civil Service Employbe -.36 .07
37 Farmer -.28 .07

Not included in the PTQ. The.criteria for inclusion were generally as follows; (1) The significance levet .04 (two-tall)
generated about 80 items; this was felt to be an upper limit in a questionnaire of about 225 Items, including these 80 A-B
Items. (2) Included also are a few items (9, 40, 93) needed to break up a sequence all of which tend to be disliked by A
doctors. (3) A number of original A-B items are included in the PTQ even though reanalysis has shown they are not
significant predictors. They are Included.to facilitate comparisons with other studies. (4) Some items significant at .04
are not Included because they are "out-of-style" (155, 159, 260, 310), burled In long contingency series (296.297), vary
skewed (129, 130, 278, 387), or redundant (13, 103032).
Original 23 A-B Items.
(B) The B's prefer the second choice.
 APPENDIN5 1.47

Table 54, SVIB Items Correlated With Therapists' Percent Improvement

in Patients Who Are Neurotics
(Doctors who completed residency before. 1959) .
N=42

Item Number
SVIB 'PTO Item Gamma

8 ,3-29 Auctioneer -.40 .04

11 * Auto Salesman -.38 :05
21 3-33 Cashier in a Bank -.53 .02
101 3-55 Algebra -.78 .008
103 * Arithmetic -.63 .02

115 3-57 Geometry -.61 .01

173 3-66 Detective stories -.47 .02
185 3-68 Making a radio set.**. -Al
-.52
.03
.005
186 3-69 Repairing a clock
237 4-2 Absentminded people .44 .02

255 4-4 Foreigners .-.47 .03

269 4-5 People who talk very slowly .52 .02
278 Men who use perfume .49 .05
296 Opportunity to ask questions and consult about .diff. .38 .05
336 1-70 Work for yourself vs. carry out program (B) .51 .02
353 2-35 Reading a book vs. going to movies (B) .47 .01
386 3-24 Smooth out tangles and disagreements between people -.73 .01

Between the .05 and .10 level:

7 Athletic Director -.35 .07
15 Bank Teller -.44 .08
22 Certified Public Accountant -.42 .06
51 Lawyer, Criminal , -.35 .07
67 Pharmacist. -.31 .10

70 Playground Director -.35 .06

79 Reporter-sporting page -.33 .09
83 Scientific Research Worker .37 .07
116 History .38-- ,.06
171 Sporting pages -.39 .06.

212 Doing research work .38

239 Quick-tempered people .40 .09
240 Optimists -.32 .10
262 People with protuding jaws .60 .06
264 Blind people .34 .09

267 People who always agree with you .39 .07

268 People who talk very loudly .47 .09
275 Bolshevists .41 .06
304 Henry Ford, Manufacturer -.34 .08
303 Thomas A. Edison, Inventor .41 .09

348 Present a report in writing vs. verbally (B) .39 .06

350 Playing baseball vs. watching (A) -.41 .06

Not included in PTO; see table 53.

',Original 23 AB items.
(A) The A's prefer the second choice.
(B) The B's prefer the second choice.

'.166
148 EXPLORING THE PSYCHO- SOCIAL THERAPIES
rr.

. When the drug-prescribing data became "robbed" some of the variance in therapist per-
available to me, it was clear that Phipps
. formance that we seek to explain.
residents 'had started to prescribe drugs One other set of analyses may be of interest
almost immediately after the drugs had been to some readers. It will be recalled that the,
introduced in 1%55; 12' of the 47 therapists had "functional reversal" refers to the fact that the
prescribed drugs for at least some of their "A-B Scale" reversed when applied to
schizophrenic patient4; Item analyses were therapists treating outpatients in the Veterans
repeated for the other 35 who had prescribed Administration. This has widely been inter-
no drugs, and these analyses are shown in preted that the A's are more effective with
tables 10-13 in chapter. IV. It is these analyses schizophrenics, while the B's. are more effec-
that are used for determining how the PTQ is tive with neurotics. Other explanations are
scored. considered in chapter VI. Because of this
Clustering items resulted in the variables widespread interpretation, careful attention
presented in tables 14, 15, and 18, chapter. IV. was given to any SVIB item that might
The means and standard deviations of these possibly have similar or.dissimilar correlations
variables are presented in table 55. with the percent-improved of schizophrenics
In table 55, it can be seen that the percent of and the percent-improved neurotics. Using .10
schizophrenic patients improved, is 60.8 for. the as the significance level, no items were found
35 therapists who did not prescribe drugs. For that carry, the same sign with both criteria.
the therapists who prescribed drugs, the per- Two were found with opposite signs:
cent improved is 80.5 for their patients who got
drugs, and 76.0 for those that did not. The Item 113, Geography, has a gammi of + .54 at
dramatic increase in percent improved is part .04 with the percent-improved of neurotics, and
a gamma of .41 at .09 with the percent-im-
of a time trend that went on during the period proved of schizophrenics.
that Whitehorn and Betz collected data. This Item 209. Adjusting difficulties of others, has a
time trend is present in the period before the gamma of + .49 at .10 with the percent -im-
drugs were introduced. proved of schizophrenics, and a gamma of .72
The time trend poses the following ques- at .10 with the. percent-improved of neurotics.
tions:' Were the patients less disturbed as time Actually, of the 35 therapists, 30 say they like
went on (e.g., coming to treatment earlier)? Or this item, the other 5 being indifferent to it.
These 5 therapists are low on the schizophrenic
were the therapists getting better at treating criterion and high on the neurotic one.
schizophrenics? If the former were the case,
there should be, no correlation between
therapists' A-B scores and the time they began Unfortunately, neither of these items was in-
their residencies. If on the other hand, there is cluded in the PTQ. The geography item is the
a correlation between the therapists' A-B only thing in the PTQ that reflects an-
scores and the time they .began their residen- thropological concerns, the kind of concerns
cies, and this correlation is at least as large as that are widespread in psychoanalytic
the correlation between the percent-of- writings. The "adjusting" item makes some
patients-improved and time, then it is difficult sense in that social adjustment is a serious
to explain the trend in terms other than an "im- problem for schizophrenics but not for
provement" in the therapists. neurotics.
The latter is the case. Among the 35 It is doubtful that these two sets of relative-
therapists who prescribed no drugs, TOTL S is ly weak relationships can be taken as support
correlated .44 with the year they began their for the popular' interpretation of the functional
residency, while the percent-improved cor- reversal of the "A-B Scale." Much more
relates only .31 with that year. For this reason, substantial reversals are presented in chapter
the time trend was not partialled out in the VI for some personality disorders, as compared
various analyses. Such partialling would have with both schizophrenics and neurotics.
 1.49
APPENDIX 5

Table 55.-PhIpps Clinic Therapists Means and Standard peviations

Standard
Variable Mean Deviation

35 Whoa Did Not Prescribe Drugs for Schizophrenics

W-B 23 12.11 4.41
W-B 22 . 22.91 7.32.
% IMPRV S fl 60.80 23.16
% 1MPRV D * 78.97 15.61
% 1MPRV N ** 83.97 18.08

TOTL S .21.57 6.59

TOTL D 24.34 8.03
Ton N 14.74 5.§6
SD 5.49 3.32
DN 4.09 1.76

S 16.09 4.30
D 14.77 4.38
N .. , 10.66 4.45
S OTHR 8.63 .:.r . 3.12
D OTHR 6.63 2.64

N OTHR 10.03 2.62

SXPRSA - 3.97 1.64
SXPRSS . 5.49 1.62
SQUEST 2.80 1.73
DBUSNS 15.66 5.14

I3CIVIC 3.29 . 1.60

NSOLVE 6.57 3.82
NQEST1 4.49 1.27
NQEST2 . 1.69 1.25
.

12 Who Prescribed Drugs for. Half or More

% IMPRV S (Drugs) 80.50. 11.22
ABDRUG 7.17' 2.82
% 1MPRV S (No drugs) *** 76.00 30.23

' N= 34 N = 32 N=

IC
--Appendix 6
PTQ Analyses of the Phipps Clinic 'r.
'Thekapigts,
At.the time the second study was planned, we have is the 35 doctors who did not prescribe
the former Phipps residents' were to be asked drugs. Of this 35, 3 are dead and 2 more are no
to respond to the PTQ for two reasons: (1) to longer in clinical work. Of the remaining 30, 17
determine the long-term stability of the A-B responded to the PTQ. These 17 were analyzed
measures' -by comparing their earlier responses for whatever could be learned from so small a
to their. present ones; (2) if stability were high, group.
to determine whether any of the presently With respect to long-term stability, it is
measured personality characteristics relate to already known that profiles derived from the-
their original success with schizophrenics. SVIB have considerable stability over as much
These objectives were seriously challenged as 40 years (Wnitsky, 1973). The present test-
when it became apparent that' it would be retest spans an average of about .25 yeais. In
necessary to subdivide the sample for various figure 4, the correlation coefficients on .the
Purposes (e.g., prescribed drugs vs. did not, base of the triangle represent the long-term
hospital vs. nonhospital, etc.) The largest cell. stability of the three major A-B predictors.

Figure 4 ..
PHIPPS RESIDENTS' CORRELATIONS AT TW.O TIME E. PERIODS

35 17 (who responded to the PTO),

Criterion Criterion
subjects subjects
at time period 1 at time period '1

C C(percent of patients improved)

.70. .07

.75 ..61

.73 .41
at time period 2

.62 .79 .82

TOTL S TOTL D TOTL N

150
 151

The TOTL S predictor has a long-term stability words, the failure of the TOTL S predictor is
of .62. Thvcoefficients for TOTL D and TOTL primarily a failure of the 17 therapists to ade-
N are .791ind .82 respectively. quately represent the original criterion for
Thelact that the TOTL S predictor has the schizophrenics. There. is no readily available
lOwest stability. suggests the following ques- explanation why the nonrespendents include
tion:Since this predictor contains .a number of so many 'who are either high' or low on the
items from the original "A-B Scale," and since criterion.
these: therapists may have known about the Finally, aPe can ask whether the man est
scale, 'could they be distorting their answers personality scales in the PTQ are related to the
consciously or unconsciously_ as a result of their success these therapists had 20 years ago. For
knowledge? If we divide the individual .items the 17 therapists' percent of schizophrenics im-
within the TOTL S.predictor into those- that proved there are no significant correlates.. For
were part of the 23-item scale' and those that . 10 in nonhospital settings there are the follow-
are not, we find no difference in the, stability of ing correlates with percent of schizophr,enics
the two groups of items. If anything, the improved: Deference Anxiety, .66; and Nur-
original items are slightly more stable. In turance Anxiety, .63. These do not confirm the
other words, for this group of therapists wh "correlates of correlates" in table 16.
prescribed no drugs we can find no evide of For the 417 therapists' percent of depressives
distortion which could be traced to kn wledge improved, the current correlates are: Work
-'of the "A -B. Scale" and its component items. Ethic, .48; Remorse, -- .52; and Fervor, .57..
Looking now at the coefficients on the left of For the percent of neurotics improved, the cur-
figure 4, for the 35 Original ,therapists, the rent correlates are: Need for Order,', .50;
three A-B predictors reproduce the original Remorse, ' .68; and Nurturarice Anxiety, .51.
criteria, 'Percent of patients improved; at .70, Of these six coefficients, only two confirm the
.75, and .73 for TOTL S, TOTL D, and TOTL N, "correlates of correlates" in table 16 (Work
respesctively. For the 17 who responded to the- Ethic and Nurturance Anxiety). Failure to con-
PTQ these coefficients are .50, .89, and .73 for firm could mean that the 46 therapists 'in table
their original responses to the SVIB, showing 16 are' not representative of what -the 35
considerable loss of power for the TOTL S therapists were like' as residents, or it could
predictor when we move from the 35 tifthe 17.' mean that the 17 therapists are not represent-
If we compute their current responses to these ative, or it could mean that the 17 haVe
items, the TOTL S predictor has no power at changed enough. over the 20-year period to
all (.07).. TOTL D and TOTL N do much better change the correlations. We know, of course,
at .61 ,and .41 respectively. that the 17 .are not representative for percent
The complete failure of 'the 'TOTL S predic- of schizophrenic patient& improved. -
tor is due in part to the somewhat lower test- In any event, the relationships above are not
'retest coefficient of .62. But the major loss in, unreasonable in themselves. (It had been ex-
power is due to attenuation of the criterion (see pected that Remorse would predict lack of suc-
page 61). The 18 therapists not included are cess with neurotics, but it does not show ,in
much more at the extremes of the criterion, table 16.) The general conclusion of this appen-
i.e., very high or very low in percent of dix is that theespondents are too few and not
srchizophrenic patients improved.. In other representative of the original group.

For the 18 nonrespondents, r is .92, significantly dif-

ferent from the .50 at the .05 level (two-tail).

i
Appendix 7
AdditiOnAl Analyses - Second arid
Third `Studies
This appendix presents some additional Mothers' occupation and education were also of
tables and some additional analySes. Specifical- no consequence.
ly, the intercorrelation of the PTQ- manifest Once when I asked Barbara Betz what con-
personality scales is presented in tables 56, 59, tributed to a person's becoming an "A"
and 62 for the 46. nonhospital psychiatrists, the, therapist, she said that she had found that "A"
43 hospital psychiatrists, and the 51 Loysville therapists had had personal experience With
helpers. Iii general these tables show that persons who suffered severe mental disorders.
while the two groups of psychiatrists are not Accordingly I asked respondents whether,
greatly different from each other in the pat- when they were growing up, they had had a
terns of intercorrelation of these basic per- friend or relative with a severe disorder, and,
sonality variables, the pattern of intercorrela- if so, how qften they had seen this person suf-
tion for the Loysville helpers is decidedly dif- fering. One scale for their responses is: no such _

ferent, probably reflecting the needs of this in- friend or relative, had such but did not see, had
stitution to deal with quite different clientele. such and saw occaaionally, had such and often
Table 57 presents the intercorrelation of the saw suffering from the disorder. This scale has
A-B predictors and clusters for the 46 a significant positive correlation with .D (.33)
nonhospital psychiatrists. This can be com- and with NQESTI (.32) and is certainly worthy
pared with similar intercorrelations for the 35 of further investigation. At Loysville, the scale
-no-drug Phipps residents in table 14, chapter is negative with improvement in situationals
IV. ( .45).
Table .58 presents the correlations of A-B Age of Therapist. It is not uncommon to
predictors and clusters with the PTQ manifest observe that the 'older generation is less
personality scales for the 43 hospital tolerant. Whether this is generational or
psychiatrists. Comparison of this table, with a _maturational is unclear. Among the honhos-
similar one for the 46 nonhospital psychiatrists pital psychiatrists, :age is negatively related to
(table 16) shows how different are the cor- SD, TOTL D, DBUSNS, and NSOLVE. Among
relates of the A-B variables. As discussed psychiatrists as a group, age isipiegatively
above, the A-B predictors and clusters are only related to. Regression in the Service of the Ego,
useful for the no:drug situation, except Tolerance of Ambivalence, Deference Anxiety,
ABDRUG (see tables 17-19). Tables 60 and 61 and Nurturance Anxiety. However, the inter-
present means and standard deviations for correlation matrices of older and younger '
various subgroups of the sample. psychiatrists are not different, provided
Background Characteristics. Psycho-
. hospital and nonhospital groups are kept
therapists are all Class I in the Hollingshead separate. It is the hospital - nonhospital 'break
scale of Socio-Economic Status. But we can ask which is critical for the correlates of the A-B
about the social class of their origin their predictors and clusters.
fathers' social class. This variable produced not Females. Considerable attention was given
one significant correlation, nor do its coni- to how the femaleein the second study 'might
ponents, fathers' occupation and education. differ from the males. Unfortunately, there are

152
 Table 56,-Intercorrelation of the PTQ Personality Variables,
Nonhospital Psychiatrist and Residents

Penionlan r's N 48

Netd la
Closuri. REORSN MPTHCI waketi NORDER UNREAL ARNE DECISN POND REISORS ANTIDM INDIVCL DEFANX NRTANX SCIENC .INTLEC

32 Regression in the

Service of the Ego

33 Empathic Interest 7556 .379

SI nIficalltly different from zero;

34 Work Ethic .210. 7.188 .089

at .05, r ,2g1

35 Need for Order .272 -.177 -.367 ;225

36 Tolerance of the at .01, r .376

Unreal'Experlencei 7.418. 1.387 .258

37 Tolerance of

Ambivalence -.266 .075 ,115L.081 ).105.,1 .034 .

38 Preference for

Decision Making -.529 .175 .323 .018 7345 ,229 .326

39 High Personal

Standards -.205 .044 .013 -.216 .311 7150 .024 292

40 Remorse .168 .021 .011 .336 .102 -.151 a .126 -.114

41 Antidemocratic .283 -.367 -.323 '.267 .208 -.527. -.088 -.388 -,..083 -.091

42 Individualism -,354. , .569 , .254 7081 -.232 ,324 .170 .063 -.143. .137 -.377

'43 Deference Anxiety .L.349 ,521 .236 -.222 -.143 .194 .040 .073 -.161 -.004 -.210 .452

44 Nurttkrance Anxiety -.254 .103 .056 -.154 .031 .201 .254 -.059 .082 - .165..221 .061

45 Science and

Moral Values .001 -.013 .057 .191 -.214 .301 -.288 .163 .075 .105 .095 -.006 -.168

48 Preference for Intel.

lectual Challenge ;02 .355 -,169.7.059 .043 7054 .073 -.128 :.038 .127 222 25 265.

47 Extroversion M15'. .285 ..153 7074 -.319 7151 -070' 176 .095 -.090 - .141...149 .220 .076 .027 .114
I

-.213 M47. 1!1 .272 .188 .207 .037 .081 L.0e5 -.050 7,016 .093 .010 -;048 7,047
50 Work Activity

The correlation between XTRAVR and WRACT is .299.

r
1 173
 0' Table.57.-Intircorrelition of. the Al Predictors, Drug.FrOe Relitionships,
NOnhospital PsychlatriSts,and Residents

Pearsonlan i's N = 48

.1,011 re

for Schizophrenics f)Neurefics

.VARIABLE W.B 2 t:1011 5 S SXPRSA ISXPRSS SQUEST SD 'T01,1, Q D DBUSNS ION TOIL N N NSOLVE, NOEST1

f011. S .827

Variables are described on pages 3545.

S .529 .864
co

Significantly different from zero: 0

SXPRSA .165 .535 .735
. .4
at .05,1 = .291

SXPRSS .370 .599 .747 .455

SQUEST, .610 .591 A72 .005 .115

SD .901 .874 .516 104 .301 .554

TOIL D .737' .637 .298 .',158 .164 269. .80

D .288 .175 .04\ 164 .019 7,025 .262 ..759

DBUSNS .458 .284 049 051, .021 .004 .438 .831 .838

DN :245. .125 -.125 , -.082 -.227, .335 .822 .412 .652

TOTL N .501 .348 .082 .168 .070 .011 .511 .813 .382 .560 .577

.503 .350, .130 .222 .114 .082 .472 .533 .339 .488 .366 .960

.365 .6.... .879

NSOLVE .522 .347 .083 . .070 .140 .091 .512 .486 .202 .437

NQEST1 211 , /05 122 .191 .126 .100 .135 AS .588 A61 .169 .397 A49 151

II 1

. NQEST2 .071 .0531; -.053 -.080 -.106 -058 .142 .119 -.118 -.081 A60 .201 .040 .023 -195

175

I
 Toble 58. = Personality Correlates of the A.13,Predictors and Clusters, Drugfree Relationships,
Hostiltal Psychiatrists ant Residents

(Pursonlin is N 43)

for Depressives
Paw*/ Nilo
App. 4 Tabb NEW for Schizophrenics , for Neurotics

*I 22 TOIL 1 6 URSA IXPRO 11011EIT SD TOIL D D DINAND ON TOIL N N NIOLV! NOR NW

31 Need for Closure -.039 .073 .227 .281 ,.129 -,025 -,125 -.012 .105 .027 -.019 .095 .114 .029 .180 -.047

32 Regression In the

Service of the Ego -.244 -.095 .075 .142 .114 -.228 -.257 .263 -114 -.194 -.225 -089 .031 -.116 -M03, -.232

33 Empathic interest ..211 -175 .024 .093 .024' -.147 -.346 -.195 .022 -.076 -.082 -.081 -049 -.143 .106 7103

34 Work'EthIc 071.. -054 -M42 ; 009 7188 .1 - .051,w,014 008 ,080 .032 313 321 333 .164 -.177

35 Need.for Order +.011 7.059 072 1.228 -.139 .112 -.037 -027 030 106 ,,104 038 .097 022

36 Tolerance of the

Unreal Experience 310 3 -.016 -.062 158 -.147 -.417 -.392 ir,169 7.350. -.273 -.309 -.273 -.222 -.321 -.003
37 Tolerance of

' Ambivalence -.120 -082 -M23 7014 .036 -.087 -.123 .080 .241, .04 .03,4 :016 .000 -.003. .049 :029
z
38 preferenceior

Decision Making

39 High Personal
.028 , .174.
-.333 .413 .298-.171 -.063 ,7,082 ,005 -.127 -195 .076 .143 053 .313 -.125 x.

Standards .179 ,201 .219 .207, .185. .132 ..116 .153 ,121' .146 .079 -.009 -M36 -.057 -.M06 '267

40 Remorse -.086 .016 062 -.002 .055 -.158 -.042 .127 -090 -095 -.202 -.125 '-.072 067 -295 09

41 Antidemocratic 230 114 .037 .0 7.001 ,260F .165 , .147 .113 .173 M24 .170 .170 258 .107 -.241

42 Individualism -.211 -,142 -M79 ,026 -.021 -.287 .168 -016 ;072

43 beferenCe Anxiety .023 .094 .018. .1p -.136 -.023

44 NOrturance Anxiety .168 .038 -,157 ,205 -A40 .020

45 Science and

Moral Values .046. .198 .123 -.020 .0 .166

46 Preference for Intel.

lectual Challenge -.180 .001 .162 .175 .081 .021

47 Extroversion .342 .348 .326 .224 .183 .241 .262 .227 .070 .044

50 Work Activity -.086 -.153 -213 -.222 -150 -.260 7.035 131 277

176'Single underlining meens.signif Icantly different from zero at .05; two.tall, Double tinderlin

Variables In the heading are described on pages 35.45,

 Table 59,!-intercorrelation of Ile' PTO Personality Variables,
Hospital Psychiatrists and Residents,

Pursonlen r's N 43

MN My Nod la

ANINI...111
41).41416Ru* Clan EARN .MPTHCI VIRKETH ROARER UNREAL MINI DION RiNDRD RENON ANTIDM INDINX DEFANX NNTANX SCIENC INTLEC

32 Regression in the.
Service of the Ego -.366

33 £mpafhic Interest -.382 .438

Significantly Different from zero:

34 Work Ethic -.107 .021 -.011, .

.at .05, r 21
35 Need for 'Order .161 -.151 -.317 -.077
36 Tolerance orthe at ,01, r .389
Real Experience -,127 .487 .258 .-437 L,364
37 Tolerance of
mblvalnce -.467 .276 .431 ,103 -.450 ,381
,

38 ference for
. Dec *Making -.455 '.322 .414 .087 -.334 .221 .388

39 High pars al

Standards ,,,... -.158 .186 -,237 -.180 -.007 -.045 . .021 .176

40 Remorse 024 .283 .232 .032 .008 .146 ,090 .022 -,055

41 Antidemocratic ,272 -.302 -g 329 .069 -A06 -.317 -.034 .028 .099

42 Individualism -.166 A13 A77 .211., .019 290. ,325 .183 -.027 .322 -A55
A

43 Deferende Anxiety 7A00 .131 .110 .036' -226 -.023 262 .283 ,070 .001 -.052 .051

44 Nurturance Anxiety ..141 -.011 7.123 .092 ..128 -.269 -.148 -.024 .112 -.177 .090 -.003 .319

45 Science and
Moral Values -.049 -04 -,076 -.055 -.025 -.152 -,099 .0 .049 -.278. 190 -.174 .235 .190

46 Preference for Intel.

lectual Challenge 7,262 .313 .079 -.111 .016 .434 .113 .184 .241, .164 -.098 :146 :174 -.203 .069.

46 Extroversion -.017 .076 .058 -.172 -.148 .352 .316 .168 -.262 -.052 -.071 -.151 -.015 .051 -.121

50 Work Activity -.020 112 ,082 :043 .119 .061 .055 7,218 220 .055- .009 106 -.011 .010 -.066 ,124
,

The correlation of XTRAVR with ViRKACTis -.069

 Table 60.-PTO Yarlables Psychiatrists and Residents

Nonhosphal Psychiatrists Hospltel Psyc

Verlible Experienced Residents All Experienced ReiWints

,

i o
x..: 0 i4 i a a a

Need for Closure . 5.18 2.85 4 ;57 241 4,98 '271 . 4.81 1.94 3.82 2.70 430 , 21
136 24,86. E14 2189 8,47 24,05 8,86 27,18 1,10 2E65 8,06
Regression . 2347

10.70 2.05 1.1.05 2.13 11.36, 2.48 11,21 2.29

Empathic Interest, 10.63 215 10.86 1.88

4.88. 2,00 4.64 3.37, 4.80 2.46 5.00 259 ' 3,77 1.41 437 214
Work Ethic .

1.50 4.43 2.90 4;30 2.01 4,05 .2.01 3.73 262 188 232
Need for Order 4.25

7,25 3.24 '8,29 1.73 7.57 2.89 7.14 165' 7,09 2,91 7.12 3.25
Tol Unrealistic
1105 1.99 11,27 200 10.671 1,07
Tol Ambivalence 9,50 258 104 2,03 178' 2,44

6,25 1.74 710 1,52 6.48 1.70 8.14 1.68 6.86 1.55 ' E51 1.64
Pref. Decision Making

1.84 0.45 1.79 0.58 1.83 0.49 . 1.62 0.74 1.82 , 0.39, t72 0.59
High Standards
4 8,43 2.68 8.87 2.83 10.62 238 10.50 2.50 10,56 2.41
Remorse . 9.06 2.91

- 6.43 4.42 '715 t82 E62 ,4.43 .5. 4.17 6.00 429
rIlldemocratic, 7.75' 5.00

.3.67 13.30 3.37 12.67 3.48 12.nta 3.82 12.72 3.51,

Individualism 13.16 3.28 13.64

2.97 6.57 2.93 ,. 5.41 3.02 5,48 2,69 6,73 3.40 1.12 3.10
Deference Anicy 4.91

1.72 1.28 107 1.64 2,19 1,51 1.95 1,47 2.09 1.63 -2.02 134 ,
Nurturante Anxiety /

Science 41,25 0.92 0.64 0,84 1,07 0.93 1.00 0.84 1,00 0,82 1.00- 182 ;
4.50 1,67 310 218 4,20 117 4.05 115. t 18 1.68 412 1,69. ,t1,
Pref. Intellectual
r
8,88 3.71 114 3.69. 8.96 3.71 ' 8.33 2.99 917 3.05 9.07 3,P.
Extroversion

Fervor 4.97 2.88 4.14 2.11 4.72 267 i 5.43 2.75 4.45 2,34 1,193 217 p
'7.71. 2.45, 8.64 1.97 8.19 2.24 A
Life Satisfaction 7.22 2.62 7,93 1.86 7:43 242
1,86 1.93 1,91 1.63 1.88 1.76
Activity 2.09. 1.94 2.71 1.86 2.28 .1.92
0
21,11 ' 8,30 21,05. 1 21.37 7.56
B - 18,66 7,45 23.43 7A7 20.1( 7,70 6.96

2214 E55 2141 ' E70 21.57 5,85' 21:14 6.36 -2135 E05
TijTL S 1166 6.73
14.25 15.76' 3.92 15.18 3.72 15.47 3.79
,, 15,06 3,64 15.36 15,15 339

4.15 1,48 4,24 1.51 4,36 1,50 4.30 1,49

SXPRSA 4.22 1.52 4.00 1.41

317 1.62 4,36 1.22 4.09 1.50 4.33 1.59 414 1.83. 4.23 1.70
SXPRSS
3.24 1.37 2.84 1.73 2.93. 1.58
SQUEST 2.53 1.59 286 1.70 2.63 1.61

6,79 3.09 5.26 3,92 5.81 116 5.95 3.55 5.88 133
SD 4.59 4,10 .

172 2317 6.59 2138 6.95, 25.32 , E38 24.37 E20

TOIL D 2247 7.23 26,43. ,
1.4.63' 341 14.00' 4.05 15.41 2.91 1412 .3.54
D 14,13 .3.55 15,79 2.83

'-- 14.14 5.06 15.77 4.12 14.98 462

DBUSNS 14.13 t61 18.86 .2.54 14.96 4.26

3.75 1,55 3.86 0.95 3.78 1.38 3,57 166 315 1.62 3.77 1.63
DN.
14,62" 4,30 15.68 5.00 15.16 4.64
TOT L,N,, . 13.97 4.39 16.07 3.95 14.61 4.33

13.43 1.74 14.41 4.84 13,93 4.31.

12.63 3.80 14.64 3,86 ,13.24 3.89

E54 E67 1.43 E18 153 593 145

NSOLVE 456 110 7.79 4,.02 3.68

4,72 1,22 4,79 1,12 4.74 1,18 448 1,17 482 1.22 4.65 1.19
NOEST1

1,81 1.09 11.36, 1.01 1.67 1,08 1.76 , 0.77 1.95 0,84 ,. 1.86' 0
1i0EST2
130 711 1.33 1.80 , 1.65 ,7,95 209 7A1 206 7,67 : 217 )
ABDRUG E13
21. 22 . .43 OA
Number of oases 32 14 .- 46
Of
14

Coded variables-in the Stub are described on pages 35.45 and 49.

i so
 Table 61.-PTO Varlables-Selected Groups

Normal Volunteers
Psychology Interns
Male Male

Variable Wyss Ille,Stall

,*Pre.Med Psycho'. Mors Fjmales Females Males

X. 0 07 0.. o X 'a

Need. for closure 453 2,47. 5.78 1.86 3.22' 1.48 4,38 2.06 .4.00 2.24 3.32 -155

Regresslon . 22.86 9.13 25.67 3.57 28.67 9.7, 1 27.38 6.23 28.35' 1.35 31.32 5.37

Empathic Interest I 1051 2.61 10.00 1.66 10.56 2.60 ' 11.19 2.89 , 12 :00' 1.62 11.73 1.81

Work Ethic 649 2.54 4,89 1.17 .4.00 2.06 5.10 2.05 . 3,65 . 1.69 3.45 1.63

Need for Order 429 2.25 3.33 2.94 3.44 2.60 .2,10 1.79 1.71 1,49 3.59 t92

Tol Unrealistic 657 324 7.78 2.73 9.67 3,50 9.38 3,20 . 10.00 3.71 9.68 1,76

Tot Ambivalence 9.37 240 8.22 1,79 9.22 2.39 8,80 , 2,43 ,10,59 1.66 10.23 1.48

Pref, Decision Making 6.22 1.88 6.33 .212 6.78 . 0.97 6.38' 1,60 6.71 1.16 6.82 1.26

High Standards 1.69 0.65 1.56 0.73 1.56, 0.73 1..71 0.64 1,94 0.24 1,73 0.63

Remorse 9.61 2.71 8.33 2.45 9.44 2.01 10.05 2.18 ,10,35 2.45 9.68 2.57

Antidemocratic 11,55 6.73 , 9,89 3.14 7.89 5.90 7.43 4.21 , 355 2.64 3.77 2.39

Individualism' 11.80 3.36 10.33 3.00 11.89 3.30 12.81 ,;2.89 14.71 2.80 12.68 2.48

Deference Anxiety 5.75 3.27 4.78 119 , 618 2.86 6.67 3.23 7.06 2.9 6.82 2.50

Nurturance Anxiety 1,47 1.57 1,44 1.13 1.22 1.48 1.19 1.33 2.47 1.94 232 , 1.29.

Science 152 0.87 1,22 0.67'' ,1.44 0,73 0,85 0.80 .124 0.75 1,27 0.83

Pref, Intellectual 3,98 1,82 '' 4,11 2,15 4.44 1,33 352 2.09 -3.59 2.27 3.86 '1.61

Extroversion 8,65 3,10 ' 10.56 3,78 1.78 HT 8.76 3.59 V.06 3.51 9,45 4.67

Fear 2,80 6.00 1 194 5,78 2,64 4.57 2,56 311 '1,61 4,27 1,96

Life Satisfaction . 1,97, 756 L 1,88 8.00 1.80 6.52 258 8,35 1..13 ' 726 1.99.,
Work Activity , 2.10 ' 150 aoo aor aoo oho 181 ,1,69 1,94 1.4 111" 193
ri
...4 Vii 22 1959 . 6.02 18.89 4,48 .. 18,56 3.32. 2011 7.36 24,71 5,46' 2295 7.60

18,45 4.58 19.11 2.26 21.67 3.94 21,67 5.31 24.24 5.53 22.18 3.78
TOTL S

S 15,25 343 14.00 2,24 17.11 2.85 15.67 267 16,24 3.96 15.59 2.34

SXPRSA A.73 1.34 3.78 0.97 4.11 1.54 3.62 1.60 4.29. t53 4.32 1,39

4,65 .:1.55 4,22 .1.20 5A4 1.67 4.71 1.42 4.65 1.32 4.86 1.36
SXPRSS

259 1.37 1.89 155 257 1,32 262 1,32, 356 178 223 1,23
SQUEST i

3.20 2.71 5.11 2.57 4.56 2,01 6.00 3.92 8.00 . 2.96 6.59 3,65
SD 1
,

TOTL D 20,73 5.77 6.38 .. 21.78 4.09 22.90 6.48 . 29.71 4.71 27,23 6.08,

D 1441 3,38 15.11 3.55' 14.7t 2.44 1311 3,00 17,11, 245 16,77 351

D8USNS 13,65 4.04 15.56 4,88. 13,41 117 13,62 3.61 .1714 3.23 1750 4.29 '

DN 11Z, 156 4.4 114 2,44 159 3,19 1,40 459 106 3.86 152

TOIL, N J 15.02 4.70 1122 4.47 1450 550 1167 4,61 15,41 423 14,45 315

N 14,16 4,44 1123 409 13.67 .3.97 1267 420 413,65 420 13.23 3.39.

NSOLVE 752 359 4,33 3.04 5,78 163 6.57 2.93 724 .3.56 555 322
,

NQEST1 3,67 152 ,3,33 166 189 1,27 343 154 4.18 1,19 4.50 \1,06

NQEST2 2.06 0.90 2.67 0.71 1.89 1.17 1.38 0.92 1.53 1.01 1.86 .0.99
. ,
AEORUG 7.14 2,41 6,00 224 659 3.10 6.71 2.00 7.53 215 7.50 2.20

Number of Cases 51 9 21 17 22 1.8,

182 . Coded variables in the stub are described on pages 355 and 9.,
 Table 621,1ritert?' orrelatlon of the PTO Personality Scald Loysville

Penonlen is N 51

homily VON Nod la

NORDER UNREAL ANDRA DECISN UNDRD RENORD -ANIMAL, 'NPR DEFANX:NRIANE SCIENC INTLEC
Apo.4TobloNwotor Clown REORSN MPTHCI WRKETH

Regression In the

Service of the Ego -.194

33 Empathic Interest -.100 .429

"Significantly different from zero:

. 34 Work Ethic' 349 -.530 -271.

at .05, 276
.35 Need for Order' ,394 -,402 -.241 26

36 Tolerance of the
at
Unreal ,Experlence .510 .379 "L382 -.381

37 tolerance of

AmblvalenCe -.182 .167 .209 'L.109 -.087 .118`'

38 Preference for

Declilon aking -,271 .331. .116 . -.320 -.404 146 ,208

39 :High Pars al

Stan s -.149 .162. ,037 -363 .092 .320 .103 139

40. Remorse' X154 340 ,164 -105 170 .169' 051 .029 293

41 AMideMociatic .025 '7307' 7.517 477 145 =364' 7.469 109 299

42 Individualism -391 293 .328 -244 -160 ,179 319 ..342 .354 .068 -,601
0
I 6

43 Deference Anxiety -.289 .183 --,149 -376 -.257 205 -.156 292 -329 .149 330 -,061

44 Nurturance Anxiety -.160 .267 .048 -.280 -.154 .297 .208 -.036 -.167 .063 .282

45 Science and

viorap/aRieS -.138 .074 .138 -233 -.158 .125 142 384 -..168 .175 315 .194 -.155

,46 Preference for Intel..

lectualChallenge 389 .243.. .302 354 -.170 MO .424 200 ,029 -334 .194 .343 -.254 -.095 .269

47 Extroversion . -.018 . -.046 -.1237 286 356 123 233 214 -.1k .212 .138 .249 -357 ,.166
 160 EXPLORING THE PSYCHO-SOCIAL THERAPIES .

..
not enough female psychiatrists for separate However, in ercorrelations suggest a
study. However, we can compare female reasonable me sure of agreement. With
psychologists-in-training with males, and we respect to diagnosis, it w s found that two of
can compare female nornial volunteers with the judges and- a,number of elpers, particular-
males. With respect to means, the females are',. ly those helpers no longer- at Loysville, had
much lower on the need for order, suggesting quite fferent (and idiosyncratic) fr s of
that they may be well adapted to working with refere e for the six diagnostic ps. e ,
,
schizophrenics. They may also be higher on in= consensus diagnoses do ir6t include es ar- ,
dividualism, possibly an asset in-dealing with ticular judges and\helpers.
conformists and depressives.
With respect to patterns of intereorrela-
tions, there are only a few differences between
..
helpers, a number of background alyses t
Finally, with respect to the Loysville
were conducted. It was found that the elper's
females and males. (There are no females in the scores on the PTQ personality 'variables were
Loysville sample). There are decided dif- not related to (1) the number of youths he chose
ferences for three. A-B clusterS: SD, SQUEST, to claim,' and (2) the severity of the youths he
and N8OLVE (tables 10, 11, and 13); Not only claimed regardless of whether severity was
are the correlates different, the females are measured by judges or helpers. There were
higher on these scores reflecting their rejec- some significant, relilions with the age of
tion of the traditional maledominanee of these youth claimed. Specifically, older youths were
interests. Clearly, these three clusters 'do not claimed by those higher in Nurturance Anxie- -
measure for women what they measure lOr ty (.32), lower in the Work Ethic (-- .43), higher
men. The other A-B clusters !nay have meaning on Preference for Decision Making and Regres-
for female therapists, but what is really needed sion in the Service of the Ego (.32 and .28).
is %study where success criteria ,for women are Since little else relates to the age of youths,
available for complete analyses. these,,findings probably do not affect our con-
Loysville. There were some further analyses clusions. 'Youths who came to Loysville later
Of the. LOysville sample which maybe ,of in- were treated. by helpers who were higher in
terest to some readers. Interrater agreement Nurturanee -Anxiety (.55) and Preference for
on such things as' severity and improvement is becision Making (.33) and lower in An-
not easily computed since no two youths were tidemocratic ( .28). ., ' )
rated by the same group of judges and helpers.

ao
Appendix 8
Somatic 'vs. Psycho-Socjal Orietitations
Drug- treatment is. ConSidered. to be impel.tant SXPRSA and required for schizophrenics not
for psychotherapy because drugs change on drugs may not be so important if drugs are
the patient's behavior. In this way drugs in- prescribed. Since this difference was found for
'fluenee the therapeutic felationShip. The the same group of therapists, it. can be pie-
therapisit's treatni4nt orientation 'toward sumed that .the difference: is not in the
drugs is also important in the sense that his therapist's personality but rather in the pa.
treatment preferences reflect his .personality. tient side of the relationship, i.e.; a change in
Therapists who prescribe &tugs may be. dif- . the .patient when drugs are prescribed. (We
ferent-from those who do not. Therefore, when have information about drugs only for. the
we compare drug treatment with no-drug schizophrenics at Phippi. All other diagnostic
treatment there. may be. changeS from both groups are drug free in our data.)
sides of the therapeutic ielationship. The other A-B predictors for schizophrenics
The situation is further complicated in the are also probably invalid when drugs are
present researCh in. that when. we pass from pirescribed; probably because they are subject
drug. treatment to no-drug treatment, the / to semantic reversals. Specifically, the W423,
meaning of some -AVIA ..predicfors may change. and the cluster, SD, which is highly correlated
These semantic reversals iithe meaning of the with W-B 23, do not predict for schizophrenics
"A -B Scale" were found-when we passed fibm that are receiving drugs,,ABDRUG, whichis
high-empathic-interest to low, and -when. we comparable predictor for schizopfirenics
:passed-from- hospital to nonhospital (although receiving drugs, produces similar correlates to
we coUldsfind rip relation between empathic.in-. those for W-B 23 and SD, suggesting that,
terest and hospital/nonhospital). If it is though the predictors don't generalize because
reasonable to. assume that either empathic in- of semantic reversals,.. the personalities they
4trest or hospital/nonhospital .is - associated represent are somewhat constant regmdleaa of
with drug; tre'atment, then this reversal adds a whether -drugs are prOcribed; Specifkally; the
`third compliCation to the significance :of drug 'correlates suggest ti* eMpathic inteiest .and
treatment when we use certain A-13 predictors. tolerance are needed by,,schiophrenics.
In the present research it is difficult to The question addressed in this appenikmis
separate- the ttree effects: the)change in pa- as follows: Are such personality characteristics
tient beha!viorlthe ideologies ofiheraPists, and also related to drug treatment behaviors and
. the semantic reversals: fortnnately, some 4-B ideologies? As noted, we foundlno difference in
predictors are not. subject to the semantic -. the PTQ personality scales between hospital,
reversal. Specifically, and SXPRSS and. nonhospital psychiatrists. But hospital/
have fairly.. constant meanings
meanings acrosa various nonhospital is clea-rly only a tentative approach
groups of therapists; They are really manifest to measuring treatment ideologies. It is useful
'scales. SXPRSA has predictive poivei for the lickause it can be used to deserike therapy
tgeatMent qf schizophrenics for two ?eparate studies, as well as therapists, bid it is 'a crude
groups of therapists if no drugs are prescribed: `measure .because drugs are used in many
The correlations fall short of significanee if the nonhosPital settin ' 3.
patients-are on drugs... This appendi* t ies to push aIittle further
Thus, we consider it possible that the active inte\the complexitie f treatment grid trapv:
jeadeship represtnted.. by merit ideologies in .t lation to persona5ty,
 162 EXPLORING THE PSYCHO-S0e1ALTHERAPIES

Specifically, it presents (1) the SVIB correlates 383. Check Yes to "Stimulate the ambition
of drug-prescribing among 17 residents at the of my associates." 4c;
Phipps Clinic, and (2) the personality correlates .396. Check "Practically" never. ,,borrow"
of treatment ideologies among 25 therapists at , rather than "Borrow frequently (for
a short-term inpatient 'facility.. personal use)," (Note that this also is
II not in cluster 2 below.)
a

SVIB" Correlates of DiugPrescribing Although some of these items do not appear to

have much in common, several of 'them do
_ Antipsychaltic drugs were intfoduced at t e
reflect a preference for free active execution;'
Phipps Clinic almost immediately after they tints* 299, 321, and 328. This group of .items
became available in 1955. Dr.' Stephens has would conform to some of the stereotypes of a
made available/ the proportion of each busy, driving, no-nonsense. physician.
therapist's scligophrenics for whom drugs The second 41uster consists of six items. The
were prescribed. By correlating this, propor- theripists who prescribe more drugs are'more
,.. tion with' the therapist's SVIB responses we.
likely to 6.
I can characterize the therapists who prescribed 103. Be indifferent toward or dislilte "Arith-;
drugs for most of their schizophrenic patients. metic."
# We must, bear in mind that when a new type,of 145. Dislike "Poker."
treatment is introduced, whether a therapist 146. Dislike "Bridge?'
uses the treatment is likely to be influenced 280. Be indifferent toward rather than
not just be his treatment philosophy,.but also 4 'liking "Athletic`Men." (Npte that this
by his general stance toward new treatments. item .dois net belong with 'Athletic
Moreover, for those therapists whose residen- Director' in the fourth cluster helow.)
cies spanned the introduction period, the pro- 360. If they must choose betweed"Jealous
portion may 'not- be as high as for those who people" and "Spendthrifts," ihey prefer
cameidter. Accordingly, we will analyse the 17 the former.
therapists who finished their residencies in 399. They check "Never make wagers"
1959 or later. The proportion of schizophrenic, rather than "Frequently make wagers."
patients receiving drilgs varied from 11 per- As in 'cluster. 1, some of'the items don't, fit in..
cent to 100 percent4ith a mean of 55 percent. but the dislike for poker, wagers, and spend-
The ace analysis is presented in table,63.. thrifts Is yery, suggestive of some aspects of
When the items are intercorrelated., four the , Protestant Ethic. This is of interest
clusters emerge. The largest of these clusters becauge. Hollingshead and Redlich (1957)
consists of seven: items. The therapists: who, reported that psychologically, oriented
prescribed drugs for. most ,.of their schizo- psychiatrists are overwhelmingly froni Jewish-
phrenic/patients are more likely to: ckgrounds while somatically oriented ones
-. 26. Like "Clergyman." (Note that this is not are primarily from Protestant backgrounds.
in cluster 2 below where one might think Cluster '3 is a so,cial 'bluster. The five items
it belongs.) ' .. . are:
299. Rate higher -than some other,aspects of
work "Freedom in working out o- 's 25g. Dislike "Gruff men."
.own methods' of doing the work." . .:300. Rate highei than some other aspects of
302. Dislike "Enrico-Tarns°, singer?' ( ote work '4e(o-workers-48%itenial, tom-
that this does not belong with disliking petent an adequate in number."
"actor" below.1 *lot 354. Prefer "Blonging to many societies"
321. Would rather be a 1Street-car Motor- rather than only a few.
man?, than a.."Stredt-car .conductor." 363. Check "Yes" to "Win friends easily."
328. Would rather "Execute plans" than. 3641 Check "Yes", to "Usually get, other'
"Develop plans." .
...--.. . . 4.t people to do what 'I want done.
.

4e,
 , APPENDIX 8 . Nib 163:
, .

Table 63. SVIB Items Correlated With. Doctors' Prescribing Drugi '
for Their Schizophrenic Patients t
(17 doctors who ended their realcSancles In 1959 or later) . .

item Number
SVIB PTO Item Gamma
ti
1 Actor (not movie) .62' _:05
7' Athieti4 Director 4.50 .05 ."
26 * dlergyman . 11 .01
Playground Director
103 Arithmetic r ,04 4

130. Physiology .1.00 .03

P ker -.52 .04.
146 ridge .01
165 -.68 :02
,254 Gruff. men -.66 ,.02
261"' People with gold teeth._ .57 .05
268 People whotalk very, loudly -.73 .04
.280 4-7 1- Athletic men , -.78 .01 6
284 Determine, the cost ofthe machine ,04
285 -Superviie the manufa.clure.of tits machine . 58 . .

298 / Certainty that one's wotk;f0)11 be judged fairly --.60 .02

,299 Freedom in working out n rpthods ol doing work , .64 :03
30 Co-workers-congenial, compe1ent, and adequate in number .60

.: 302' Ehrico.Cacuso, singer --.56 , .03

.'321' =street -car motorman vs: Conductor (lOW drug) .64 r, .02
328 cpevelop plans vs. execute Plans (high drug) A- .03
354 Belonging .to many societies vs. few (lowdrug)\ .64 .02
(360 Jealous people vs. spendthrifts (low dryg) .64', ,02.
363 Win friends easily . 65 .04
314 .Usually get other:people-to do what lwant done :52 .05
383 Siimulate tf1 ambition of my.associates 1".

396 BorroW frequently Vs. never (high drug)4 .0;

399. (high drug) -.59 .04 4
BetVieen the ,10 aod the .05 leVel (two=t ii)
.

82 School teacher . ----.49 1_0,-

. 83 Scientific Research VVorker .55 :09: ,,

- 140 Tennis r -.57 .1b'

Ortjanizing .a play._ . -.44 :10 ,,,- s

231 Climbing along the edge cif.a prec A8 .08.

*Religious people .43 .10, .,
250-
V9 4,6,%. People who chew gum . 64- ..06
289-, TeatOf others to use the machine :06,
308 William H. Taft, Jurist :48 .09
312 Scipretary4f*a Sticiety or club' ,-; . 54 - -.07
358:-. Tail.men vs. shott merhigh drug) L---.55 - .06 1

(high drug) means that the second choice Is preferred by do (who prescribed dru s for most.of their schizophrenic
patients,
` .

. I...
 EXPLORING THE PSYCHO-SOCIAL THERAPIES
6
Many 'Of these items are suggestive of rather behavior that treatment ideology. Relating \
"casual social viationships.:
The fOurtlA
'
has four ite s' and con-
treatment behavior and treatment ideology to
personality variables \was one of the reasons
-

sists of disliking Actor;

sista. tt
Athi Director, for developing a Simon revision of thp PTQ;
Playground, Director; and Vaudeville. Do these In this revision (Fir C), the following kinds .

items reflect a' dislike of strutting for an au- of items were dropp d: (1) .A-B. items not
dience? .
1 .
'validated in the 35-doctors atuilysis (tahlei
it should be made clear that these( clusters- 10-13), (2) personality scale items which failed
are\not so .much aspects of a common d mension to correlatewell with thescales they were Sup-
as they are separate -dimensions. Sin e all of posed to be part of, and (3) items which most..:
the items are derived from a common c iterion, respondents answered in the: smile way., 'In
they tend to be correlated, but the corielatiOn. place of these dropped items, most of t4e item's
among the clusters is small and not st4stical- in table 63 were fitted' in Iii addition,.one extra
ly significant. In terms of the theory outlined page was added consisting of ,quesiioiii drawn
on pages 11-12, drug-prescribing is.- a ipheno- from Strauss et al. (1964); these are". further
type, a resultant of several genotypic folies. It described below. ,
is not a single dimension from the viewpciint of Form C was administered to 25;therapistain
personality correlates. a shore-term inpatient facility: 7 These
In. summary, the' doctor who presctibes
drugs for most of his schizophrenic patie4s is experienced psychiatrists
tt
therapists, 14 ,men and 11 ,wom i ; riniged. from
paraprofes-
similar to the stereotype of the active, drting, sionals. All had' had experience \working. with
oral, sociable but impersonal, physic4an, patients for whom drugs wera Prescribed.
uite in contrast to the relaxed, informal, per- Some of the therapists were ex-addicts. The
onally involved stereotype a the psychiatrist. purpose of collecting these data wa to find out
hile there is nothing here to stiggest that the: whether any of the personality scaleS in the
s matically oriented therapist, endorses the PTQ ccirrelted With the drug-prescribing
w irk ethic, he does reject SVIB items relating clusters shown above, or With the Str uss et al..
to poker and wagers, which is also an impor- treatment ideologies. .
t t aspect of the Protestant Ethic. First, with respect to the drug- prescribing:
the roughly 30 SVIB items in:table 63, on- clusters, it should be noted that the ctusters
ly 3 were included in the PTQ and these 3 are were again found to be unrelated to each \ titer
not epresentative. Therefore, the PTQ's col- except air .a significant, negative re1atio be
lecte in the second and third studies (chapters tween the second (dislike poker; etc.) and .t
IV a d V), are of no use in pursuing these (social) cluster ( .42, p < .05). The th .

analy s further. However, just before this' . cluster tends to be negative with, the of
mono aph was finished, a second revision of three which tend to be positive with ac
the PT (Form C) was developed to' include other. ThtiS again it appears that these clusters
many o the items in table 63 plus some addi- represent several different personality
tional qu stions on treatment ideology. These -, characteristicS.
are .consi ered further in the next section. The "social' cluster does not correlate
significantly with any of the- PI'Q personality
scales. It tends to be negative with-Nurturance
= Treatmen Ideologies Anxiety (-7 .37, p < .08) and positive with Life
Satisfaction (.37, p < .08).
The last se tion was concerned with the cor- The first cluster, active execution, is .

relates of rug-prescribing," the actual negatively associate 4 with EmpathiC Interest

behavior. Tre tment ideblogy is only one ( .49, p < .02). Thd second cluster, which in-
determinant of treatment behavior. Not only , eludes poker; bridge, and spendthrifts, is
cas...ireatment havior be influenced by the negatively associated with the Preference for
treatment settin , but as Lionell's study in- Intellectual Challenge ( .51, p < .01). Its
dicates, there are ther aspects of personality relation to the Work Ethic is negative, :a min-
which have a stron er influence on treatment significant .29.. The' fourth cluster, which in-

1 E)
 E'sr
AP i) NDI.- 8
165
°
Psychotherapy is by -far the most humane form
.41 ek-;, actor, 'and p4playgreui14.: direCtor,1 is of treatment for hose, hospitalized for mental
associated with Regression in the disorders. (True) `'S
.,..Seryiee*the:Ego:(;-- p,-,C.:011):: In a majority of c ses, ityingto treat psychotic
These findings stIggest the:Possibikity that schizophrenics wi intensive individual psycho- _
.,.,,drtUrprescriiing.,;behivior, may roate4 to therapy it, a waste- i if the4therapist's time. (False)
lie74100:peOotilo characteristics thit may 'The three were no sig -ficantly correlated -
yrelate r to :,,effectiveness of the Tisycho-soeial with each other, -but the-S matic tended to 'be-
therapies. 'TRis Nnoh4sizei More the negative With the of er tw
daniers. of trying tO 'Study' the psycho- social The Se,matic orien ation was positively cor-
':therapieS*ithdut ',Careful measurement alsO of related with the eed. -for . Closure (.54,
the.- ditigi! therapies- which are Often ad- p < .01) and 'negat ely' with Individualism
.ministered concomitant yr. 4;
p < .05). T ,Sdcielogic one was
With respect :4( 'ideologies; they following's:, negatively correlated with the Need for Order
-were tidaptedlroni'Stratps et al. (1964); ( .40, p < 4054 Old i ositively'corre ated with
Somatte treatment if
Deference Anxiety (.5 , p0 < .0.1).T Psycho-
By and large, when patients are treated with *logical one w-as positt ely correlated ith Em-
drugs 'and shoCk therapies, imprOVement tends, pathic Interest (.40, <- .05) 'and negatively
to be only temporary. (False)- correlated with the- i eed fo,r Orde (.44,
Unless 'organic 'bases of mental disorders are p .04) and Nurt ranee hnxiet ( .47,
better ,upderstOod treatment can not be ex- p < .03)
pected to'improN% mateHally. (rtge) findings sug: est that there is an in-
These,
'Drugs and shock therapy continue to be more teraction, between th psychosocial therapies
effective than other forms of treatment for
persons' hospitalized because of mental dis-, and 'drug treatments. In chapter VII we have
otders. (True)- A already noted- the interaction between
diagnosis and treattn, nt, and the possible in-
Sociological treatment orientatioA: 'teraction between tre tments and the defini,
The .ioCial environment of the hospital is an tion ofthe outcome of eaments.) The interac-
`important factor .in whether or not the patient tions are not adequat ly handled by random
will improve. (True) . assignment ,te,
In fir, while a particular
The chief limitation of using the social environ- may ay e effe tive even though the
ment in the hospital as therapy is that it modi- therapist does, not bel4 vie in it, few would deny'
fies 'only the external, Outward aspects of.the 'that the.therapist's be ief in therapy is usually
patient's behavior. (False) .
,
-11 potent, force in treat i ent.
The role of Social factors in the development of The issuer present d in this appendix are
mentar disoidert is: frequently overestimated.
(False) the most diffiltult that re encountered in this
research. I recognized very early that the in-
v.;
,. dependent variables ere somehow intertwin-
Psychological treatment orientation: ed with each other an :with control'vOriables.
, As causes of mental distorder, psychological
factors tend to be less important than biological,
The hospitallnonhospit I distinctidn is a useful,
Or social environmental factors. (False) . preliminary Solution.
 .In these 'stud' s, the analyses, h been : ' analysis assumes 'a com p nsatory model.
- - based on fairly 'small Cells-46. of 1, , 43 of Variables are additive or subtractive; they: can -'
that, 'and so forth too small or reliable not be,additiVe here ancyithbiractive there. In 1
multivariate analyses. Taken all egether, .fact they cannotrien be additive in one part of [
several- 'hundred PTQ's wer collect d. Why the sample, .ancOrreleVant in 'another. There is/.
not throw them all together, nd ecind ct a fac- no way for th6, factor analysis of the total
tor 4snalysis? 'Why not use the Major riterion _-:. , group to show the:Auite different loadin
. variables as dependent variables in i multiple SQUEST,.SD, D13t1SNS, and NSOLVE fo the"
regressien? Theie are tbe issues considered in 'two snbgroulps.. The analyiis. for the to '1
this appendix.,,,While definitive a_ ns ers will group tends to. dump' -these variables, the
. , not be for coming, one general con lusion is 4 variables of prime, interest, into minor factors.
e' analytses are no: bstitute .This. is not to say that the' factor analysis is
for some good hard thinking 'about t e data. not useful. ''Among' the manifest personality
Iri older, that the contrast' is lear; the variableS from the PTQ. there is a reasonably
analyses; reported corisist , largely forming common .structuie at least so :far as 'the !first
subcells on certain objective bases, s ch as sex, two factors are concerned. even though the
treatment set professional st tus, etc., cells are. small. MOreOvbr, the factor analysis
and t en paring the interc rrelation Seems to place the manifest scale, SXPRSA, in
ma the covariance matri s among the second' factor which hai loadings on
the cells. Although the currently Preference for Decision Making and .High Per-
a ilable :statistical testS'Sre not id al for this sonal. Standards. BLit then, if meanings are
urpOse, it was possible to establish hat, so far manifest, who needs to analyze? It is par-
as A-B predictors are concerned, n are dif- ticularly. with respect to the. A-B clusters that .

ferent from women; and hospital p ychiatrists we Seek information, and it is here that the fac-.
are different from nonhospital, one for analysis fails. us. = . .

Table 64 presents factor analyse 'for the 89 Mu\ltiple regression likewise assumes an ad-.
psychiatrists. The analyses Vore c dtiCted on ditive,:compensitory model. unless we build in
the-A-B clUsters, above the dotted .1 ne, and the some interaction terms in advance. Since our
PTQ personality scales below the line. Work task is \to find where such terias are needed,
Activity was excluded becausei it is so skewed they cannot be anticipated in the modeLipie:
that the Pearsonian r's are not reliable. Most of failure tai find any common variables repre.,
the A-B Predictors. are not included- because senting sucdess with various diagnostic geronrisi-'
they contain overlapping items. -The clusters, makes it difficult to pool our data.
however, contain no overlapping. tems. For the Analysis of variance, Of course, can, test for/.
group as a whole, one additional variable was interactions, but the model usually take/ the
added, namely, hospitalinenhospitai which independent variables. as attributes, thus
defines the subgroups, hoping that it would throwing away a great deal of infOrthition
hOlp.sort out. the difference. It did. not. when these variables are in fact ordered
While 89 is still too small, it is large enough_, classes as they are in our .data, Moreover,
to illnitrate the danger's of relying on factor . analysis of variance tends to throW inforMatiOn
analysis exclusively for data reduction. Factor away if the independent variables are not or-

166
 of Principal Components--A.B.C1 Ofers and P TO'PertOnality Scales.
Psychiatristkand Residenta,
, r
43 Hospital
Peychiattists

2 3 4 1 2 3 4

-07 -21 -39 08 -41. 09 -70i 12

SXPRSA -72' 24 .

-.65 -09 17 -01 ;-49 -60 , -11 19 -4i. -44 33

,SXPRSS .,; . 1

-51 -2 -21 -70 04 -28 -18 -.45 '07 i -49

SQUEST -19 -03 ,

-32 -46 -23 07 -727 42 ,! -62

SD , 136-, -26 -68 -11. 36

-06 -01 ;19; 18 pi: 60/ .720

DBUSNS -09' -17 61

-30 -42 04 7.:-4 57 t ..-41

-71 -08 46 ,

NSOLVE " 19 ...7717 . 24 :

4
1 1

po 726 -22 -58 . 25 -21 62 18 -05. 42: 66 1,1 22

NOEST1
:

!07 713.f lin-35OP

mi-19asp .0 amiAmi
.03,r 69 10
NOEST2 12 -16 09 ......
36 -09
oil_ms , b
al MMMMMMM Nos
MO 11011.1114111111111111101 Mil
i
NCLOSR7
MOP no Rolm 011 MO Mi MI NW= MIN
A60 -53
, - - : 6q -25 03. -03 59, ' +42
....
21
;
.-
,
1

REGRS'N 60 -13 -06 -.,56 1 66 , 27 724 -24 .69 701 , 11 19

..5 -42 .02 04 73 16 =10 -12 62 $2 , 031 09'

ro

; MPTH91
,-07' 24, , -44 -26 :7;08 -02: 44
W.RkETH -10 09 64 10 07

-28 46 00 -06 -41. 26 14 '-09 7.42 .26 2i,, 34

NORDER
13 ;760 07 '65 1 -26 -06
UNREAL 72 -05 .7,03 18 46

-04 0i) 10
AMBIVL .07 -32 -19 -06 .70 -19' 15 44,'s 7'-30

65 r49' -07 -24. , -67 -04 05

DECISN :, '32, -74' 10 11 ' '

-06 22 -46 -13 -17.

STNDRD -63. -00 -09 -08 ' ,H44. r. ii ,

-37 -26 11 14 -07 35

REMORS 11,,, 26 40* ,-12. ',12 33

-11 -55 -17 00 23

ANTIDM -74. 04 06 -15 , -4p.. ' 13

INDVDL .66 (' 10 2 -44 , 156 28 -05 16,, 13 23 17

DEFANX 47 1 00 -14 -55 39 00 / 37 if 41, -12 43 07

-04 63 -.40 /` - 01 701 58 12

NRTANX 27, 00 -45 04 15 . .

'f '
'-.08 -14 21 x-10 0 J-01 . 06 20 -02
SCIENC 21 10 51 18

-01 -42 02 20 -12 -15 49

INTLkc -05 '', .04 27 -74 (27

-65 'I I 00. -53 06 -12

KTRAVR 00 '., -35 704 -44 22 16 11 ,

-42 ---=-11 -35.. -57 32 \ -28 -26 18

43 - -10
.

FERVOR -27 ',-26

-64 09 ';, -59 06 -02
SATSFN 04 "-54 -03 -06 23 00
-08
NONIHSP - 4
17 i,,, -08 701

% Of te
10.9 8.7 7.3
Variane 14. 12.9 10.5 8.9
I

Decimal points omitted.

1.93
1.94
 EXPLORING L THERAPIES

thogonal. floPefully other multivariate ant 'models. In the present stuff les,' the
will be developed that more adequately ;fit the ks came not from multivariate models :.
requirements of naturalistic observation and T y came when I asked the substant've ques-.
natural experiments. --
na: How might these subjects dif er from
In the meantime, clinical researchers need to psyehotherapists ?. Hew might these t erapists
respect the costly character of their data; and / differ from each other?
not be overly impressed with the prestige of

w.

O
Appendix 10
Extracted from:
Commonviealth of PennsylVania,
Department of. Justice,
. ; Juvenile Court Judges' Commission,
xi Caseload Classification and
Supervision Approaches
(mimeo), pages 17 to 23.
Although words, or ."labels" are used to juvenile justice system, and then adapted the
describe an individual or his behavior, this 'Pro- cross-classification of offender typologies
cess only has meaning in terms of .assisting us develgped by Marguerite .Q. Warren. This ap-
inOnderstanding his behavior and identifying proach allowed the Task Force to consider
possible supervision approaches. It should be many classification systems and then adopt a
noted that delinquent behaviors areusually in- classification method built on considerable
dications of underlying problemscor situations. research (Warren, 1968).
The supervising officer must be able to see the As in the caseload classification recommen-
juvenile as a human._ being capableof many dations, seldom will a juvenile fit completely in
behavio.and- he should be aware that terms one supervision category. However, inmost in-
used to describe some of the child's behavior stances, a delinquent will usually fit into one
are only conceptualizations intended to Offer a category more than another.
better understanding of the causes of the In using this system, it must be remembered
behavior. that it is based on the understanding that
There is another caution which must be con- supervision must be provided either because of
sidered in classifying behavior. When a the child's needs or the nature of the offense.
Juvenile's fi'ellavior is clissified there is a For example, the classification of Situational
tendency to view this classification as a final Offender, in thiS system; is used to describe a
decision, and thus to perceive the child as such, client who at least requires informal adjust-
no matter what new behaviors he may demon- ment. Many children who adjusted at in-
strate. This is a misuse of classification since take could be seen as Situational Offender's,
behavior can change at any time, and the but some mitigating circumstance requires ad-
superirising officer .is responsible for working justment at the intake level.
with a juvenile as he is at the present time, not
as he -was when he was originally -classified.
Thus; classification is an ongoing process The Situational Offender
which takes into account the growth and
develoriment of the probationer and any The juveniles described by this term are
changes in his behavior that:rnay occur. .usually found to be normal children who don't
[The lothibit] provides S basis for the appear to have any strong need for supervi -.
classification of juvenile offenders' behavior, sion for example, the youngster who is social-
based on the data used to determine ap- ly well: adjusted, but who made a mistake, or
propriate disposition, as developed in the first was led on by the group. Involvement in delin-
section of the paper. The Task Force studied quent activity., is usually accidental or the
. various classification systems that had been result of a specific situation which was over-
_

developed by other professionals in th whelming. 169

 ,

Juvenile Off e di 13 haiiq

SUBCULTURAL. MANIPULATOR' ,
L

SITUATIONAL 'IDENTIFIER NEUROTIC ' ANTISOCIAL, '. 'CONFORMIST / ASOCIAL

qature of Delinquent act an No Substantiated 1

Nuubstantlated No SuOtantlated ', , Client was 0 Client .probabl, /
Dffense accident or caused research research research f011ower inVolvid
,
'cornmited.de-
a by situation
With,ithers unguent act

alone

Prier. : s!, Probably ''....\-'1 Behaviors all y

Property ands Properly offense. , irwOhred with, IMpuisiVe
Delinqu none accepted by person offense's involving personal ..,1 others and hOstile acts
Behavior peer group' gains Identified as 'which May have
. , , .
follower . been violent
' . , ,
Threat to
Leaderrho co Id: ',,,POSSibl'e threat , , Exploslie be,
self andior No Maybe threat .Possible threat be threat to thidugh pei, ,havior, makes
community, .
to others because of anxiety others '.' :, groUp pressUre ititn a tbreat .
, 1

Family Probably ' Parents are Parents are chit ' Parents inconsls.. Nieak parental`' Rejecting
Structure acceptable : suspiciouS and like and client is , tent in providing ', diSciplineier' 'parents with
distrustful expected to be love and reject ','indifference 'possible
adult, tion physical'
.cruelty
,

-1 . .
Attitude . Realistic and Believes, delinquent ':Denies'self imalle Justifies behavior Admittreapons Denies 4apoti, ,' .

Towards regretful behavior Is of delinquent , , as warranted. or bilitilar ', ' 1

,
sibilltY
Current acceptable reasonable' behaliOr:,' projects blame
Situation 1
, / ,
1
, i ,, onto world

Emotional Probably sound No obvious

.
.
Anxious and; Emotional ' : :Behavior defer. No impulse
Adjustment problems confused ,
lsolatien ,
mined by need: , control'.,
and Control
far Self infantile

approval demands

, (

School Probably good Identified' as, Unusual behaviors ' , : ', History of using History of History of
Adjustment acting out and noted In records. !,., ,.,:', classmates and attention seeking acting out , ,o,

associatIO with
.i
, / instigator ; : behavior- performs, uncontrollable, ,

"bad" youngsters well,when ,' rage

Observed
. ,
Peer Group Probably intense peer Probablyia loner -may , Uses other's and ,,;No close relation. A lohr peers
Relationships , acceptable group have SeleFted friends, .' seeks them as 7 ships, but .. see Nth as
relationships ° ,, iSaCkiri.! , attempts are made' strange
, )
 APPENDIX 10

Supervision these cases should be

in i social perceptiveness and ability to pre let via'
directed toward lielping the client 'Solve the group treatment and increasing oppor
specific problem which legi to his violating the for legitimate accomplishments via training in
law. the example mentioned above, the pro- work, athletics, etc. The second supervision
bation officer would; attempt to help thfi+ method allows the offender .to :work throUgit
juvenile understand'his need to make decisions his childhood trauma In a relationship which
independent of peer group. influence. will revive or develop his capactty. to depend,
on' and be concerned about others. 'This ap- _

preach usually involves sustathed long-term

Tice gulicultur al-Identifier dividuil assistance 'requiring aUthOrititive
external. controls and offers limited Probabih .

Clients described by, this term *ill usually. of success.,This approachls Usually considered
be fund to have developed normally, but will impraitical in probation: settings.
have, internalized the value system of a minorirl e ';
ty group.. A' common example of this kind of
juvenile is a child from a middle;class home "
Conformist
who joins' a cOmmune.
Two Levels of supervision' re recommended. This client is best charactetized as having an
The first is, directed toward stopping the delin; unustlally strong limed for approval, .whtlb.
quent behavior by' demonstrating to the client presenting himself as. problem free. Sui3ei
through discipline, that delinquent behavior is sion recommendatio'ns include the !us# 's
";!pot an effective method in satisfying his or her clear, consistent, external structure in;
needs. This method, also emphasizes teaching.. concern for the offender can be 4xpres$e
the individual how to meet his status needs in a control ofilis behavior, use of group woi.,
way that is' acceptable in The larger culture. crease social perceptiveness, use.4Dpeer group .,
The second level of sUperviSion recommended as a pressure toward nondelinqueddy, and the ,
is focused on, changing the content of the teaching of skills in order to 'jell) change self-
-client's value system; and,,therefore, his self- definition in the direction of ideiluacy and in-
control. This approach requires working dependence.
through a relationship with a strong identity
model who represents the values of the' larger
culture. This approach will help the delinquent
broaden his self-definition. Asocial. Type
This client is usually chatfeteriied
Antisocial-Manipulator little' control over his behaVier:..411& needs
mediate gratification., and . self - centered.:
Clients.,dekribed by this' term are usually. +
Supervision recommeridstiomeinclude;s clear;
characterized , as not having incorporated ap- and concrete structure- = Of.' lozp pressure,
propriate social norms. They appear to have no warmth, and acceptance fronvaii.eXtremely
remorse for their delinqUent behavior. Super- tient person. SuperVisicirt. shpui:13e'Wriv and
vision recommendations move in 'kw° separate supportive in the directionO. cOnfOlnity0 and:
directions. One method is; to edcourage the° attempts to reduce the fear 0fabandnnMent'
manipulator to develop hits manipulative skills and rejection should beJziadethsrOngh:t.eSghing
in a socially, acceptable direction, to increase rather than counseling: f.
 Index of Namesr, trti ,

Adler, Alfred 7 Endicott, jean 19, '51;- 46',11.11:

Adorno,'T.,W. 22, 27, 96, 136 Erikson,, Robert V. 55,:
- Alexander', JalSes F. 93, 96 '4° Ewalt; Jack
Aristotle 11, 103 Eysinck, _Hans 13;-13n, -97
Armstrong, Stephen H. 14n, 97
Asch, Solomon E. 96, 138 Fairweatbei, 'Getifie
AstruP, Christian 50, 100 figyss ---
Auld, Frank 100 -106
Fiske, Di W. 87j.97
Barker/E. N. 4, 100' Forsythe, 'Alan B. 91,100, 104
Barron, Frank 27, 96, 108, 109, 138 Fox, Harviy 18;97;,-
Fox,: N. 99i
Barton; Cole 93, 96 Frenkel-aruniwik,'Else 22, 27, 96, 136
Bednar, Richard L. 3,-96, 111 Fretid, $igninna,2, 4,`7, 10, 74, 75, 94,- 97
Benedek, Therese 4, 7.5, 96 Friedman, William-L.78,-96, 111
Bergin, Allen E. 8, 14, 14n, 81, 96, 97,-98, 104 Fromm-Reichmann, Frieda 8, 10',. 74, 97
Berzins, Juris 1. 3, 9, 18, 78, 96, 9.7, 106, 111, 113, Fume, George X. 28, 85-
115
Betz, Barbara J. 2, 8, 9, 10, 14-20, 19, 24, 29, 32,
. 34, 35, 41, 49, 51,.56, 61, 76, 78, 80, 87, 90, Galilee, Galilei 83
92, 94, 95, 96, 100-102, 111, 112, 114, 143, Garfield, Sol 8, 81;97, 98
148,- 152 , Garrett, James B, 27, 99, 108, 113, 132'
Beutler, Larry E. 9, 96 Gebhardi M. E. 9; 13n, 97, 'i
Bloom, Victor-100, 106 'Gendlin, Eugene T. 8, 99, 114 '
Bowden, Charles L. 19, 51, 96, 111 Gerard, Roy 55, 81, 97
Bucher, Rue 4; 30, 100, 164, 165 Gibson, Robert W. 76, 96
Buckey, Harold M. 88, 96 Gilbert, Doris C. 22, 80, 97
Burnham, Donald L. 76, 96 Gladstone, Arthur 1..76, 96
Golding, StephenL. 18, 99
Caine, T. M. 87, 96 Gonies, Beverly 14n, 97
Callahan, Daniel. M. 19, 34, 99, 112 Goodman, L. 97, 105
Campbell, David P. 97, 112 Greenblatt, Milton 97, 100,43.6 ,

Campbell, Donald T.-13, 97 Grinspoon, tester 88, 97, 99,411.

Cancro, R. 99 Grissell; James 100, 106
Canon, Walter 15 -
Grossman, David 13n, 87, 97-
Carson, Robert C. 19, 97 Gunderson,-Jahn G. 88, 98
Carver, Harry C. 107 -
Gurman, Alan. S. 96, 98, 99.
Chartier, George M. 19; 78, 97, 111
Cheney, Warren D. 9, 97 Harmatz, Jerold S. 99,',111
Child, Irvin L.t1, 27, 97/107, 108, 128, 129, 132- Hasenfus;Nancy 79, 98
135, 139, 140, 142 Hays, William L. 98, 105 -,
Chodoff, Paul 80, 97 Henry, William E. 30, 98 ;;
Colby, Kenneth. Mark 10, 75, 88, 97 Hewitt, L.E. 55, .98 . .

Collins, John L. 8, 14, 98 Hogan, Terrence P. 18, 99

Cooperman, M. 97, 134, 139, 140, 142 Holland, John. L. 9, 13n,,97
Copernicus, Nicolaus 83, 84, 103 Hollingslread, August B. 4, 46, 98, 162
.Cristel, 4Jlan H. 5, 87, 88r100 Holt, Robert-R-. 8, 98
Horwitz, L: 81, 98
DejUlio, Steven S. 8, 98 Hunt, Howard 87, 97
,
Dent, James K. 3, 21, 97,, 139
Dixon, Bernard 100 Ingram, Gilbert L. 56, 62, 98
Draper, Franklin M. 19, 31,.51, 47
Dublin, James E. 18, 97, 113, 115 Jacobsen, Carlyle F. 15
Dymond, Rosalind F. 99 Jackson, Don D. 8, 98
Jenkins,, R. L. 55, 98
Ehrlich, Danuta 4, 30, 100, 164, 165 Johansson, C. B. 97, 112 .
Elton, Charles F. 18, 97, 113, 115 Jung, Carl G. 4, 7, 10, 98

172
 INDEX OF NAMES
,
, .

Cynthia 79, ea Razin, Andrew M. 18, 19, 96, 98, 99,

Zarin, Bertram:8, 100 Reahl; J. E. 9, 13n, 97
_rift, Daniel 16, 98 Redlich, F. C. '4, 46, 98, 162
Kemp, David E. 17.;,98, 108, 111, i 3 Reich, Wilhelm 7, 12, 78, 99
-Kendall, M.' G. 98, 105 Ricks, David F. 11, 75; 99
Kepler, Yohannes 84 Riemann, Fritz 4, 9, 74, 99,.
Kerlinger, Fred'98, 114, 141 Rierdan, Jill 4%
kiesler,/Dortald J. 7; 8, 86; 98, 99; 114 Reiser, Morton 87, 97
Kish; Leslie 98; 106 . Rittenhouse, Joan 87, 88, 89, 99
Rieman, Gerald -L. 12, 98 off, Merrill 99
Kluginan, Sainikel F. 98," 113 Rogers, Carl 4, 8; 99, 114
Knupfet; GeneVieve 8, 98 Rokeach, Milton 27, 98, 99, 114, 141
:Knutson,. Jeanne N. 98 Ross, Wesley F. 78, 96, 111
KObayeslit, Shigefumi 55, 98
-/Kraus, Anthony IL 87, 88, 98 Sabshin, Melvin 4, 0, 100,164-465
Krieger, George 8, 98 , Sanders, Richard 9, 13n, 97
Kruskal, W. 97, 105 Sanford, R. Nevitt 22, 47, 96, 136
Kullback, S. 98, 106 Schatzman L, 4, 30, 100, 164-165
Kwratkciwaka, Hanna Y. 21, 97,139 Schiavo, R.-Steven 93, 96
.Schorer, Galvin 100, 106
-Lambert, Michael J. 8, 14, 14u, 98 Schubert, Daniel S:P. 79, 99
LeBow; Michael D."18, 99 Schulz, Clarence G. 88, 98
Lefkowitz, Joel 18, 47 Segal, Bernard 99, 114
Levinson, Daniel J. 4, 22, 27, 80, 96, 97, 99; 136 Seidman, Edward 18, 99, 113
Lewin, Kurt 11, 98 Semmelweiss, Ignaz 10
Lionells, Marylou J.' 4, 5-6, 13, $5, 98; 104, :164 Severy, Lawrence J. 3, 96
' Lorr, Ma'urice 19, 34, 99, 112 Shader, Richard I. 88, 97, 99, 111
Lowinger, Paul L. 100, 1064 Shaffer, John W: 34, 100
Luborsky, Lester'8, 13, 87, 97, 98 Shapiro, L. 9.9
Luborsky, Lise 13, 98 Sharaf, M. R. 4, 99
Shinohara, Mutsuharu 55, 98
Maier, Norman R.F. 86 Shore, Milton F. 88, 98
Manosevitz, Martin 98, 113 , Simon, Ralph 9, 13n, 97
Martindale, Colin 79, 98 'Sims,John H. 30, 98,
Massimo,-Joseph L. 88, 98
.
Singer, Barton 13, 98 .

May, Philip R.A.,14n, 19, 51, 91,49, 100, 104 Singer, David 27, 108, 129
McGuigan, Sharon 99, 113 Sjoberg, Bernard M. 88, 96
McNair, Douglas M. 19, 78-79, 99, 112 Sloane, R. Bruce-5, 87, 89, 100
Menninger, Karl A. 4, 8, 10, 99 Smail, D. J. 87, 96
Meyer, Adolf 14. Smitlw-Edward W. L. 79, 100
Mire's, Herbert L. 27, 99, 107, 113, 132 Somers, Robert H. 100, 105,
Mizushima, Keiichi 55, 98, Spear, F. G. 88, 100'
Mobley, Max J. 96, 111 Spitzer, Robert L. 1Q, 51, 96., 111
Morgan, J. P. 42-43, 80 Spray, S.' Lee 30, 98.'
Mosher, Loren R. 76,,99 Stansfield, Ronald G. 15, 100
Muench, George A. 88, 96 Staples, Fred R. 5', $7, 89, 100
.
Newcomb, Theodore M. 11, 74, 99
Steinberg, A. 100, 113 .

Stephens; Joseph H. 17, 26, 30, 34, 35, 49, 50,

, 98, 100;108, 111
Nunnally, Jiim C. 99, 102-103, 109 -Stone, B. 9, 13n," 97
StrausatAnseIm 4, 30. 100, 164-165
Orne, Martin 87, 97 -,..Strong ;\ Edkvard K, -Jr. 117
StrAW, Hails 4, 5, 74, 100
Palmer, Ted B. '55, 81, 99 Suinn, 'Richard M. 8, 96, 104 .

Parloff, Morris B. 3, 87, 97, 100 Sullivan, HarrY'Stack 4, 6,.10, 84, 100
Parsonsx.Bruce V. 93, 96 Sundberg, Norman D. 13, 100
Parsons, H. M. 86, 99 . Suncliand, D. M. 4, 100
Parsons, Lowell B. 55, 99
Peyman, D.A.R. 88, 99 Thomas, Alexander'99
-Prinzhorn, Hans 4, 7-8, 13, 99 Tourney, Garfierd 100, 106
Truax, Charles B. 8, 99, 114
(bay, 14erbert C. 55, 99 Tuma, A. ITussain 14n, 87, 91, 97, 99, 100, 104
 174 INDEX OF NAMES
T--,Continued Whipple, Katherine 5, 87, 89, 100
Tyler, Leona E. 18, 100 Whitehorn, John C. 2, 8, 10, 14-20, 24, 29, 34-35,
42, 49, 51, 56, 61, 75, 76, 78, 87, 90, 92, 94,
Uhlen.huth, E. H, 47, 112 95, 100, 101, 111, 112, 114, 143, 148
Williams, John R. 64, 81
Vandenbos, Gary R. 8, 100 Williams, Richard H 97, 100, 136
Vinitaky, Michael 100,'150 WoloWitz, H. M., 97, 134, 139, 140,A42

,
Wagner, Maxie Earle 79, 99 Wright, Wilbur and Orville 20
Warren, Marguerite Q. 55, 100; 169
Waskow, Irene 8, 100 Yale, Coralee 91, 00; 104
Weingarten, Eric 9, 13n, 97 Yorkston; Neil J. 5, 87, 89,100
Weiss, Lillie 78, 97
Wexler, Milton 1$, 100 Zlotowijtz, Howard I. 34, 100

ti

*1

;s%_%:'
 Sp" Index
ABDRUG, x, 48-50, 76, 161 defined, 136 -
A-B clusters, see also clusters and drug-free predictors and clusters, 47
correlates of, 45, 47, see also DBUSNS; SD; in factor analysis, 167
/WEST; SXPRSA; SXPRSS and Loysville helpers,.160
defined, x, 40 - and MP variables, 153, 156, 159
drug associated, 46-49, 162-164 means and standard deviations of, 157, 158
drug-free, 42-45 and success with depressives, 79n, 80
means and standard deviations of, 149157, 158 and succe s with subcultural identifiers,
and socioeconomic status and other background 63, 64
. characteristics, 151, 152, 160 . antisocial, see asocials; conformists; delinquencx;
A-Bpredictors manipulators, subcultural identifiers
correlates of, see "A-B Scale;" ABDRUG; D; aptitudes, therapeutic, 7-8, 16-17, 21
-PIN; N; S; TOTL D; TOTL S; WB 22 art, therapist's liking for, 73, 77
defined, x, 18 asocials (delinquent)
s drug aisociated, 49-50 and AB variables; 65
drug-free, 35-45 defined, 58, 170-171
and drug therapy, 32, 50, 161 improvement of, 59 r
Meanwand standard, deviationsof,'149, 157, 158 inadequate' numbers for item analyses, 59, 63
stability, long term, of, 150-151 and MP variables, 64
"A-B Scale," x, 2, 26, 35, 41, see also WB' 22 severity of, 59
correlates,'21-25, 42, 47, 111-115 assignment of clients to therapists, 56, 87, see also
criterion-based test, 102 matching clients and therapists; random
defined, 17-18 assignment
i calopment and use, 17-20 atheoretical observation, 20
ortion of answers to, 18n, 151, see also per- attitude to wilid mental disorders, 22-24, 80, 114,
sonality
aonality measurement 136 'I'
authoritarianism, 22, 79n, 80, 114, see also Anti-
functional reversal of, 32, 78-79, 148
homogenizing, see billow multidimenisionality of democratic
meaning of, 31-33 avoi nee of others, 9
means and standard deviations of, 149
multidimensionality of, 18, 21, 103 baekgro d variables, 60, 87, 148, 152, 160
multisemantic, see semantic reversal behavior harts for patients, 16, '92
semantic reversal of, 19, 22, 32, 92, 95 behavior f the therapist, see process, therapeutic
validity when drugs are prescribed, 32, 50, 161, bias, 87 89, 1103, 106-107
- see also semantic reversal mini zing, 87, 89
active execution, 162, 164 biology, 4-15
active personal involvement with schizophrenics, metho ologic 'limitations, 83, 85ff, 91
see participative involvement "blanket blank question," 84-85
adaptability of the therapist, see plasticity; Bonferom s t, 60n
"Super A" business i terests and pursuits, 43-44, 79
adjustment problems see also DBUSNS
of the client, 9, 148
of the therapist,?, 8, 81, 94, 108
aesthetics, 21, see also art, therapists' liking for categorical studies, 1
affiliation, see Extroversion - causationsi 86, 165, 169-171
age; of client, 60, 73, 160 chaining correlations, 45-46, 51, 73, 76, 79, 151
of therapist, 31, 152 change agent, 73
alchemy, 13 chemotherapy, see drug therapy
alienation, 75 client, defined, 78
alpha, see Coefficient Alpha client-centered therapy, 4, 8, 13n
American ,Psychiatric Association, 10 clinical interpretation, see interpretation, clinical
analogue therapy7-11-24, 29, see also pseudo- closure see Need for Closure
, therapy cluster, cOinitioirof, 18, 35, 41-42
anthropology, 148 z of items, 43-44, 49, 162-164
Antidemocratic, see also authOritarianism clustered samples, 106-107
and ABDRUG, 62 Coefficient Alpha, 27, 105, 109
CoeffieientAlpha of,' 27. common sense, 74

175
176 S EJECT. INDEX

C-Continued and improvement in delinquents, 65

comparative studies, see also differential hypoth- means and standard deviations, 149, 157, 158
esis . and MP variables, 47
of diagnoses, 1, 95 DBUSNS
of therapies, 13, 87, 95 . and AB variables, 40, 44, 154
complexit14 see Need for Closure Coefficient Alpha, 109
compulsive, 9, 15 defined, 36, 38 .

conceptualizations, 11-12, 83-85 actor analysis, 167

conflict among researchers, clinicians and a i d improvement in delinquents, 65
istrators, 88-89 interpreted, 79-80
conformists (delinquent) me s and standard deviations, 149, 157, 158
and AB variables, 65 an, i. P variables, 47
defined, 58, 170-171 m U.: -mantic, 47, .155
and helper's individualism, 63 DCI IC ,t a'

improvement of, 69 an' AB variables:NO, 44

inadequate numbers for item analyses, 69, 63 de fined, 8, 43 - -

and MP variables; 64' in erprete , 79-80

severity of, 59 m ans and tandard deviations, 149
conservatism, personal, in the therapist, 42 decision make g, see Preference for Decision
construct validity, 18, 102-103 Making
contact with serious mental disorders when grow- DeferenCe Anxie y
ing up, 127, 152 and AB variables, 47 52
contamination of the criterion, see criterion vari- Coefficient Alpha, 27
ables defined, 139, 79n
contraindications in therapy, 10 factor analysis, 167
control group ft
and improveinen't delinquents; 64
d improvement in schizophrenics', 151
deterioration in, 88
and effectiveness of psychotherapy, 14, 14n "means and standard deviations, 157, 158
ethics of, 86-87 and MP variables, 153; 156, 159
improvement in, 15 and treq.tment orientation, 165
limitations of, 85, 86-88, 89-90 d linquency, juvenile, 1, 54-72, 76-78, 79, 94, 160,'
normal volunteers, 29 see also asocials; conformists; manipulators;
control variables, see age; background variables; neurotics; personality disorders; situa-
diagnosis; -sex ',, tionals; subcultural identifiers r

corrections, see delinquency emocraey, 114 .

correlates of correlates, see chaining ependency in patients, 79, 80

correlation coefficients, 105-106_ ependent variable, see criterion variables; out-
partialling, 148 come measures
costs of data collection depressives, see also D; DBUSNS; DCIVIC; DN;
A-B measures cheaper, 16 - D OTHR
exploiting existing records, 16, 20, 92, 105 drug-free predictors and dilaters, 35-45
and minimizing error, 106 improvement in, 34, 149' ,

personality studies vs.process studies,6, lb,114 interpreted, 79-80

countertransference (specifically' referred to), neurotic and psychotic, 34; 92
4,75 pluralistic or unitary, 12,-34, 92
criterion subjects, 34-35, 46, 51, 160-161, see also reversal with personality disorders, 77
validation group deterioration
criterion variables in control.groups, see control group
attenuation of, 60n, 60-62, 161 in therapy,pee worsening
contamination,of, 58n, 92 diagnosis, 12
definition of, 16, 49, 58-61 confounded in combination, 91
intercorrelation of, for various diagnostic groups confusion with evaluation, 3
8, 10, 40, 41, 50,.61 of delinquents, 55, 57-58, 169-171
means of, 59, 149 expertness in, 91, 160
Phipps improvement data, 16, 34, 35, 49, 92 hierarchical nature of, 12, 55 rel
progress in defining; 3 ignored, 8, 19
transformation of,. 105 importance of, 3, 12,.82, 84, 90, 92, 94
custodialism,12, 136 pejoratite dangers of, 3, 58, 169
Quay; 61-62
D (predictor) reliability of, 12, 81, 111, 160
and AB variables, 40, 154 differential hypothesis, 1, 7, 8-10, 94, 11.4
dlied, 35, 42 and "A-B Scale,': 19, 78-79
 SUBJECT INDEX 177

tinued experience ofthe the apist, 9, 61-62, 81

arid differentiation in patholoby, 9, 80 experimental nianipu tion, 86-90
empirical support,. 41, 45, 60-61, 80-82 administrativp a :.cts, 86, 89'
and experience in therapists, 9, 81 exploration 8 f 102-105, see also research
and generalization, 5, 12, 79 strat
.

illustrated, 3, 81 Extrove
dimensionality, see 'Coefficient Alpha; internal and ariabl s, 47, 52
consistency; multidimensionality; multiple Co = fi ent Alp a, 27
traits of therapists d d, 140
distortion of test responses, see personality, d rug presc "bing, 162-164
measurement of / actor analy is, 167
DN (predictor) d improvem nt in delinquents, 64
and AB-variables, 40, 154 eans and sta dard deviations, 157-158
defined, x, 36, 42 and MP 'varia es, 153, 156, 15.9
and improvenient in delinquents, 65 and personal i volvement with schizophrenics, 75
means and standard deviations, 149, 457, 158, and SQUEST, 47, 102
and-MP variables, 47 extrovert, 7
dogmatism, 27, 114, 137, ,I41, see also Fervor
D OTHR, x, 40-42, 1487149 . factor. analysis, 6, 92, 109; 166-162.
drug therapy feelings, others', '74
and AB research, '20 femininity,' 113
controlling for. 35, 148 Fervor, see also dogmatism
importance of, 3, 13, 48-51, 85 Coefficient Alphas 27
interaction with psycho-social therapy, 2, 34, correlates, 45
46, 48-51, 75-76, 91, c148, 149 defined, 141
and somatic orientation, 28, 29-31, 161-165 . in factor Analysis, 167
and improvement in depression, 151
ECT, 41 . means and standard ,deviations, 157; 158
education of the therapist, 73, see also training field independence, 114
effectiveness, 3,'14, 86ff Followup Study of Former NIMI-1 Trainees, 28,
of psychotherapy, 13-14 30, 31
efficaciousness, 14 Freudian school, 4, 74
imbedded figures test, 114 F Scale, see a/so.Antidemocratic
Empathic ihterest and the "A-B Scale," 22, 23, 114
and AB variables, 47, 52 and attitudes toward serious mental disorders,
and active execution, 164 22, 80, 114 a
Coefficient Alpha, 27, 92, 109 and custodialism, 22.
defined, 21-24, 131 defined, 136
explanatory power, 21-24, 78 and ethnocentrism, 22, 63, 114
in factor analysis, 167 and Fervor, 141 .

and improvement in delinquents, 64 and Individualism, 138

and 'improvement in psychotics, 79 and mental health professionals, 24, 107'
And improvement in schizophrenics, 76 and Need for Closure,128
Means and standard deviations, 157; 158 and rejection of business interests, 80
and MP variables, 153, 156, 159 and rejection of manual activities, 22, 24..
arid treatment orientation, 30-31, 165 and subcultural identifiers, 63 .
empathy, 8, 114 functional psychoses, 12, 79'
episodic improvement, 56n functional reversal of A-13 predictors and scales
ethical issues in research, 86-87 and improvement rates for schizophrenics' and
ethnocentrism, 22, 63, 78, 80, 136 neurotics, 4,1.
evaluation and item overlap for schizophrenics and neu-
ambiguity of, 3 rotics, 42, 148
interaction with independent and control vari- and personality disorders, 77-79
ables, 93, see also differential hypothesis; and semantic reversal, 78
functional reversal and severity of disorder, 79
iriethodologic issues in, 3, 84-90, 92-93 and sex of client, 78
null findings, ,1, 13-14; 84 and V.A. outpatients, 32, 78, 79
in: program evaluation, 8, 8142; 87
role of the therapistin, 10-11, 16, 85, 89 generalist approach, 1, 7-10
expectations, 86-87 acceptance widespread, 13
. of patients, 87-88
evidence summarized, 76-77, 80781, 94
of 88-89 and experience of therapists, 9, 6041, 81
. 178 . SUBJECT INDEX
generalists approach /coat/ '.
..
- -
,
defined; 138
and increase in variance of outcome, 14 in factor analysis; 167
and personal adjustment of the therapiat, 8, 81 and field independence, 114
and plasticity, 12, :78, 82 aid improvement in. delinquents, 64
generality of findings a.
means and standard deviations, 157-1518.I
assumed widely,-8, 10, 13 and MP variables, 153, 156, 159
and comparative approach, 1 and somatic orientation, 165
and diagnosis, 7, 12, 46, 74, 80, see also 4erier- individualism, rilkged, 43, 79, 79n
4ilist approach information statistic, 106
in research reported here, 79, see afro female inpapent/outpatipnt, 9-10, 19, 79, see also pay-
therapists - chopathology, differentiation in; severity t
and selection tests, 18 . insulin shock, .41 st,

and therapeutic modalities, 2, 5, 81 :3`

interaction, statistical, 9, 78, 92-93, 165,i66
and therapists' education, 73 interaction, therapeutic, see process, therapeutic
genotype, 11, 109, 164 internal consistency of scales, 109-1).0, see also
geography, 39, 92, 148 Coefficient Alpha ..
and criterionnbased tests,.102-103
habit patterns, 9 of drug-free clusters, 43-44,'199
habits, bad,.14 inappropriately appliod to" inultidimensional
"Hawthorne Effect," 86 predictors, 18, 103, 112 I'
homogeneity of scales, see internal consistency of and test-retest reliability, IA'
scales interpretation, clinical
homosexualitSr in schizophrenics, 113 of correlates of correlates, 46
hospital/nonhospital affiliation of the therapist of increase in variance in the criterion; 14n.
and "A-B' Scale," 32-33 of the pilot investigatiqn, 22-24
correlated with interest in' psychopharmacelogy, of the second and thirdatudies, 73-82
30 interrater agreement, see reliability measures
and drug-free preditors and clusters. 45, 47,
introvert, 7
involvement with schizophrenics, 14e participative
and drug-prescribing predictor, 51-52 involvement
mean and standard deviations, 157 item analysis, 19, 61-62;i 102, 103, 104, 105.
response rates, 30 for the A-B predictor,323 I

semantic reversal, 32, 78, 166 for "other" A-B items, 25

for Quay Correctional Preference Survey, 61 -.62,
somatic vs. psychosocial orientation, 31, 161, 165 71-72 l'!
hyperglycemia, 15 for rejection of manual activities, 24
hysteric, 9, 74 .
of SVIB for the "A-B.Scale," 17
of SVIB for drug-associated.Predictor,'18
ideologies, treatment, 30, 62, 161-165, see also of SVIB for drug-free prediCtors, 36-89
psycho-social treatment orientation; social. of SVIB for drug prescribing, 162-164
treatment orientation; somatic orientation , ofSVIB for success with 'delinquents, 66-70
ideology and behavior, 6, 30, 79n, 114, 164 items, selection of, for PTR
improvement, episodic56n pilot investigation, 21
improvement criteria, see also MPRV revised questionnaire, 26, 35, 41-42, 107-108,
at Loysville, 57, 58-60 109-110, 143, 146n '

it Phipps, .16 second revision, 164

and severity, 56n item selection
e trend at Phipps, 148 for comparability with other research, 107-108,
indkiendence, field, -114 146n
independence. of judgement, 136, 139, see also for internal consistency, 199 -
Individualism for modernity, 41, 146n
independence in sampling, 106 ' for neutrality; 108-109
independence of variables, 35, 40, 41, 43, 60-61, for rapport, 107, 108
see also criterion variables, intercorrela- for sequence, 146n
tion of for specific 'objective's, 21, 26, 61, 107-108
independent variable in evaluation studies, need for specific populatiqns, 107
to define, 1, 3, 6, 13, 14, 91 for. variance, 108, 146n i

Indiana study, 9 items used out of context, 103 143

A Individualism
and A-B variables, 41, 52 Johns Hopkins University, 14, 15-16, see also
behavior and ideology, 79n,. 80n Phipps Clinic t'
Coefficient Alpha, 27 juvenile delinquents, see delinquency e
 SUBJECT IN DE CL*". 179
*,
Kemp extension of the "A-B Scale," '108,113 mil' U setting, herapetaR influence in, 54, 56
knowledge, advanced-and' 1 ss advanied fields of, mill u therapy, 2, . 14n, 15, 81-82, 9d.
6, 83,.102 J. .1 minn,imizing biases 89, 90',' see also optimizing
nesota Multiph sic Personality Inventory, x,
'Minnesota
labor-managenient stri 44 8, 81, 113,
leadership,-7, 75, 76 m1onotonic 'assumptio , 105-106
leisur. , see Work Et 'Moralism, 113, see a o Remorse; Work Ethic
lengt of stay, 60 /ntioi tivation'of responde t, see respondents' inoti-*'
-1-
1,- vation
"leri h of the test 107 ' .

*Life Sitisfaction 1413.V see also criterion- ariables;

.
improvement
'
Coefficient Al a, 27
--i,
.criteria; Item analy s
correlates of, 45 and AB variables, 40, 65
defined, 141 ! defined, xi .

in factor a lysis, 167 " ' : for Loysville data, 58-61

means an standard deviations, 1 7, 15$ rheane and standard deviati s, 59, 149
and social, treatment orientation, 164 Ond MP variables, 64
linear assumption, 105-106, 166-16 \ T4r Phipps-data, -35, 41, 148
literature, theilipist's liking "for, 7$-74, 75, 77, multidimensionality,' 18-19, '164
see alsp NQEST1 nirultiple traits of therapists, see o "theratist
Loysville,' PennsOyania, Youth Deelopment .:, variable" ,
Center, 2,i28-29, 54-55, 63, 76-77, 94, 160,. as aspects of therapy, 21, 22, 24
: as inferred from independence of sters,. 43,
)nanifest personality seale, 19, 45-46, 95 ' 92, 162 .

manipulation,' einerimental, see exper as inferred from'selection tests, 18

multisemantic predictors and Clusters
'maniptilatioir "Az.B Scale," 18-19, 22, 32
manipulators (delinquent) ,DBUSNS, 43
/ and AB variables, 65 'and females, 21, 34, 160
defined, 57, 170-171 and hospital therapists vs. neinhospit I ones,
improvement of, 5,9 32, 91
and MP variables, 64 and preMed students vs. psychology. ajors,
and Quay diagnosis, 65 32, 91
Quay SurveY iteni analysis, 72. SD, 42, 43, 79 see also semantic reversal \
reaults interpreted, 76-78 and somatic. orientation vs. psycho-sOcia ,, 29,
severity of, 59
SVI.item analysis, 67-68 iLAtivaria,te
34) linear Models, 102, 166 -168
.... \
\.
manual arid mechanical interests, rejection of music,, therapists' liking for, 77 I
and'"A-B Scale," 17, 22, 112.
,andfernininity, 113
and. Scale, 22 N (predictor)
and functional psychoses, 79 and AB variables, 40, .154
implications for personality theory, 92 cliniCal interpretation; 73
meaning of, 19, 22, 112-115 see also SD defined.'xi, 42, 43 t
PTQ item analysis of, 24 and improvement in ddlincitients, 6 ,;

redundancy of items, 21 means. and standard deViations, 149, 157, 158

and SD; 42, 43, 79 and MP variables, 47 ;

semantic reversal, 22, 43 National Institutes

.
i3f health Clinical Center, 21,
and Work Activity, 79n, 142 .28 . , V ,

and Work Ethic, 79n, 113, 132 natuialistie.methods,'83, 89, 90, 92, 94, 168
matching clients and therapists, 9, 81 need for 'approval; 171'.
complimentarity and similarity, 112-113, 114 Need for Closure \ ',

delinquency studies, 55-56, 62 and AB, variables', 47, 52

McNair reversal, see functional reversal Coefficient Alplia,`27 .

meaning, scientific search for, 102-103, 107 defined, 128 '

meaning of communication' in therapy, 6, 92-93, in factor analysis, 167
114-115 general therapeutic trait, 81
measurement, see criterion variables; pe,sonality and 'improvement in delinquents, 64,.81
measurement affecting therapy, 90, 93 - and love and marriage,. 107 ..

mental disorders, 1, 77, see allo specific disorders means and standard deviations, 157-158
contact with, when growing up, 127, 152 and MP variables, 153, 156, 159,
.
mental health programs, design of, 1, 3, 81-82, and somatic orientation, 165 44 '

94, see also evaluation . Need for Order `- .,

methodologies, 83ff, see also research strategies and AB variables, 47, 52

.

o ,
 180 SUBJEdT INDEX
Need foi Order (cont.) Coefficient Alpha, 109'
Coefficient Alpha, 27 defined, xi, 39, 44-45 Nk

defined, 132 . in factor analysis, 167 b 1

in factor analysis, 167 and improvement in delinquents, 63, 65,56
and improvement in delinquents, 64 means and standard deviations, 149, 157, 158
and improvement in neurotics, 151 and MP variables, 47, 155
means and standard deviations, 157, 158 and sex of therapist, 160
and MP variables, 153, 156, 159 null findings .
and psychological orientation, 165 common in therapy studies`, 1, 29, 83
and sex of therapist, 160 explifiation` for, 13, 84
and sociologic orientation, 165- in hospitals with 'iA-B Scale,",19, 29
need-fear dilemma, 76 inferences from are risky, 75,A03-104
neurotics. null hypothesis, 103.
and AB predictors and clusters, 35-45; 63, 65 ,f nurturance, 5
and "A.43 Scale," 19, 40, 78-79, 112, 148 Nurturance Anxiety
.. clinical interpretation; 73-75, 79 and AB variables, 47, 52
delinquent, defined, 54-58, 170-171 and age- of therapist, 152
and the functional reversal, 78-79 Coefficient Alpha, 27
improvement rates at Loysville, 59-61 defined, 139
improvement rates at Phipps Clinic, 34, 41, 40, in factor analysis, 167
149 and improvement in delinquents, 64
and milieu, therapy, 81 mans and standard deviations, 157-158
and moralism, 114 and MP variables, 153, 156, 159 .
and MP iiariables, 47 and psychological orientation, 166
and Neeii for Order, 151 reversals for schizophrenics and personality-
and Nurturance Anxiety, 151 disorders, 77. ,
and personality disorders, 76-77, 81) and social orientation, 164
and Quay item analyse, 71 and Success with psychotics, 79, 151
and QUay variables, 62, 64 and time at Loysville, 160' 1

and Remorse,' 66, 151

and replication of Phipps findings at Loysville, objectives in .esearch, importapce of, 6, 83, 103,
63,-73-75 104
SVIB item analyses, 36,89, 66, 144, 147 observation, systematic, 15, 16, 20, 74 83
nonrandom error, see bis . Omnibus' Personality Inventory, 113
normal volunteers, 21; 28-29, one-to-one therapy, 5, 16, 82 .

N °TER.. optimizing, 83, 84,- 85, 107

and AB Variables, 40 1.`
Organization Level at Loysville, 54, 56, 64
defined, xi, 41, -43 original validation group, see'-validation group,
mean and standard deviation, 149 original
NQEST1 OTHR D, OTHR N, OTHR S, see .D OICHR; N
and AB variable, 40, 154 OTHR; S OTHR
clinical interpretation, 73-75 outcome measures, see also criterion variables; .
Coefficient Alpha, 109 MPRV
and-contact with mental-flisorders when growing and diagnosis, 84-85
up, 152 impact on therapy, 93
defined; xi, 39, 44-45 at Loysville, 58, 58n
in factor analysis, 167 ' at Phipps, 16
and improvement in delinquents, 65 progress in definition of, 3, 13
means and standard deviations, 149, 157, 158 and records, 92
and MP variables, 47, 155 and time trend, 148
..NQEST2 outpatients, see inpatient/outpatient.
it and AB variables, 40, 154
Coefficient, Alpha, 109 participative involvement with schizophrenics, 2,
defined, xi, 36, 39, 44-45 16, 24, 75, 76;18, 94, 114, 161
in factor analysis,. 167 pathogenesis, 8 ,
and improvement in delinquents, 65 patient defined, 73
means and standard deviations, 149, 157, 158 patterns of correlatione2142, 45-; 51, 63,, 73, 76
and MP variables, 47, 155 . Pearsonian r, 105
NSOLVE Pennsylvania Juvenile Court Judges' Commission.
and AB variables, 40, 154, 55, 57, 169
and age of therapist, 152 personal adjustment of the therapist, 7, 8, 81, 94,
clinical interpretation, 7a-75, 94 108

4
 7- -i-s1;

!co .

SUBJECT INDEX 181

P-Continued bias. in selection of, hrough statistical rnati-ipu-
personality, see also plasticity of the therapist lations, 91, 106-107
defined, 11 description of, first study, 21
stability over time, 150-151 description of, orientations study, 164
personality disorders , orp description of, second study, 28-29 ',
clinical interpretation, 76-78 description of, third study, 54-55, 59
compared with other diagnoses, 77 need for careful description, 2, 83, 92, 95
and functional reversal, 78-79 in analogue studies,. 91
among juvenile delinquents, 55,162, 76 in relation to experience of therapists, 81,'91
and outcome measures, 93 . )-, in relation to types and severity of pathology,
as research objective54 .4 . 80, 81, 90
among VA outpatients, 19 - Phipps patients;34, 50-51, 78
personality measurement, 92, 95, 102-103, 105, Phipps residents, 31, 35, 46, 49, 150-151
see also social desirability predictor, definition of, 18, see also AB predictors .

distortion of responses, 18n, 151 Preference for Decision Making

impOrtance ft context in.,.102, 107, 109-110, 143
personality o the therapist, importance of, 4-12,
and AB variablesi, 47, 52
Coefficient Alpha, 27
14-16, 73, 93 104 - defined, 134 -

Personal Standards . in factor analysis, 167,

. and AB vatiables, 47, 52,.105 helpers', in relation to age of youtlis, ,160
defined, 13§ and improvement in delinquents, 64
in faotor analysis, 167 and manipulators vs. schizophrenia, 77
and improvement in delipquents, 64 means and standard deviations, 157, 158
. means and standard deviations, -157-158 and MP variable's, 153, 1561 159
and MP variables, 153, 156, 159 and SXPRSA, 166
and situationals, 77 Preference for Intellectual Challenge.
Personal Tendencies Questionnaire and AB. variables, 47, 52
First version: Coefficient Alpha, 27
construction of, 21, 108 defined, 140
6 objectivesof, 21 in factor analysis, 167
Revised version: and improvement in delinquents, 64
AB items, 35, 41-42, 143-147 means and standard deviations, 157, 158
construction of, 107-110 . and leP variables, 153, 156, 159
objectives of, 26, 61, 111-116 and spending,aspect of the Protestant Ethic,
MP items, 27, 128-142 164
reproduced', 117-127 preference for work, see Work Activity
Second revision, 164-165 premedical students, 29, 32-33, 91
phenotype, 11, 103, 115,, 164 presenting problem, evaluation research implica-
Phipps Clinic, Henry, of the Johns Hopkina-Hos- done of, 56n, 88
pita', 2 -4 for question, see research strategies
,advantages of, as a researciv'setting, 16, oblem solving, rejection of, sel NSOLVE
further analyses of data from, 26, 34-45, 4651, ess therapeutic
143 -149 alt mate. approaCh to, 'in studying personality
psycho-social orientation of, 14, 32 of the therapist, 1, 4, 5-6, 10, 84, 94, 114-115
PTQ respondents compared with residents at, witj delinquent's, 169-171
21, 28, 150-151 with depressives, 80
PTQ response rates, 28, 30, 31' with manipulators, 77'
pphobia, 3 .
with neurotics, 74-75
hysician, personality of, 162-164 with psychotics, 79,-
placebo controla, 85, 87, 89-90 with schizophrenics, 16, 75-76
plasticity of the therapist's personality with subcultural identifiers, 78
critical importance of, 11, 12, 7879, 94 .
process schizophrenics, see schizophrenics
evidence regarding, 111, 150 -151. proof, scientific, 13, 84, 85, 89, 104
as general characteristic of effective therapist, Protestant Ethtc, 107-108, see also spending
pleasure-seeking, see spending aspect of the aspect; Work Ethic
Protest-ea Ethic pseudotherapy, 19, 29, 91, 111
polarities, 9, 30 -psychoanalysts, 5-6, 75
Populations, research -psychobiology, 14 to
bias in selection of,- through experimental 'con- psychological treatment orientation, see psycho-.
. trols, 88-89 social treatment orientation
bias in selection of, through nonresponse, 30- sythopathology
31, 150-151 fetiation in, 9, 80-81
 wy
182 SUBJECT INDEX.
psychopathology (cont.) and improvement in delinquents, 64 .

individual reactions to, 21, 22-23 means and standard deviations, 157-158
psychOpaths, see manipulators and MP_ variables, 153, 156, 159
psychopharmacolbgy, interest in, 30, 31 and success with neurotic..delinquente, 63
psychoses, see functional psychoses and success with neurotics, 151
psycho-social therapies, see also matching; milieu replication'
therapy; process, therapeutic; relationship, A-B records research, successes of, 19-0,.24,
therapeutic 75, 76
darnaging, 7, 8, 10-11,14n "A-B Scale," failures of, 16;19-20, 111,..
effecti/Peness of, 13-14 . "A-B faihires of,.explained, 32, 50, 143
measurement of, 1, 3, 6,.13, 88, 91 "A-B Scale," successes of, 19, 111 0.
.sociaraspeets of, g, 5,12; 76 and.control, groups, 88
variance; increase in, 14, 14n diffieulty. in, 1, 16,.84
psycho-social treatment orientation need for, 1, 18, 56, 7a,
"andmA-B Seale," 22, 24, 3'2, 33 of personality-variabtes regardless of other
and MP variables, 165' characteristics of therapists, 73
at the Phipps Clinic, 14-15 present. replicated findings summarized; 2, 94-95
among respondents, 28, 29-31 and therapists' success with neurotics 330,- 78-.
andNA outpatient clinics, 78 and therapists' success with schirtoPlyeriics,
49, 51, .75-70
repression, 3, 74 :. - .

Quay Correctional Preference. Survey, 71, 72 reprochicibility in measurement, 107-108, 143

Quay and Parsons, BC1 to BC4, 61-62, 64 'research, therapists' rejection. of, 10, 74, 75,88, 90
question, scientific, see blankety-blank question; research 'strategies in evaluation, 83; see also
research strategies blankety-blank. question; control group;
. objectives in research; populations, research
r, Pearsonian, 105 .control variables illustrated, 34-35, 55130, 87
random assignment, 85, 86-87, 165 cost effectiveness.otvarioUs, 1, 6, .10, 106, 114
nonrandom at Loysville, 56, 60, 160 . and diagnostic specificity, .12 -13
.

random error, 103

3.
'empiricism, "dust bowl," 20, 102-103
random 'sampling, 10,'" favorable 'conditions for, 16, 87, 90
rapport, see respOndents' motivation in human sciences, 86ff
reanalysis of Phipps data, 30: :a5-45,. 46- 50,:.143 independent variables ill defined, 3-4, 6, 13, 13n
recidivism, 58n prior questions, 83-84, 93
records;Tyst of, 16, 20, .92 reporting of difficulties encountered, 86, 88: 89
referee systems, 83, .86 respondents' Totivation,.107, 108-1.09
Regression in the Service of the Ego' respondents to the PTQ, see populations, research
and AB variables, 47, 5-2 ., response rates, 28; 30-31, 92, 95, 107
and age of therapist, 152' re once set, 108-109
Coefficient Alpha,.27 responsibility of the therapist, 10-11, 16, 82
defined, 129 reversal, see functibnal reversal; semantic reversal
and drug prescriping cluster, 165 rigidity, see plasticity .

in factor analysis, 167. rigor, scientific, 83,90

and helpers' preferences for older youth, 160 rod and frame test, 114
and improvement in delinquents, 64 rules. of research, 89, 103
means and standard deviations, 157-158 . "little. rules," 102, 109-110
and MP variables, 153, 156, 159
and success with schizophrenics, 79 S (predictor)
rejection of problem solving, 2, 73-75, see also and AB variables, 40, 154
NSOLVE . clinical interpretation, 75
relationship;: therapeutic, 4-5, 8-9, 16, 29, 49, see defined, xi, 35, 42, 44
also psycho-social therapies and drugs, 50,175
reliabilitY measures, 102i 106, 107, '109, see also and improvement in delinquents, 65
Coefficient Alpha; internal consistency means and standard deviations, 149, 157, 158
interrater agreement, 160 and MP variables, 47, 155
long:term stability, 150-151. sample size, 91, 106
test-retest; 92 sampling methods, 95, see also clustered samples
Remorse scale, definition of, 18, 41, 42,'45 -
and AB variables, 47, 52 schizoid, 9
Coefficient Alpha, 27 Schizophrenics, see also ABDRUG; "A-B. Scale;"
defined, 135 authoritarianism; Empathic Interest; par-
in .factor analysis, 167 ticipative involvement; pathogenesis; S; SD;
 . SUBJECT INDEX 183
SQUEST; SXPRSA; SXPRSS; TOM- S; improvement of, 59
WB-22 and MP, variables, 64
and the "evil eye," 15 Quay Survey, item analysis, 72
and hyperglyceia, 15 sevefitYof, 59.
and neurosis, 19, 78-79, 114-115 SVIB item analysis, 70
and personality disorders; 77-79 skewness, 105, 136
process vs: reactive, 3,41, 92 snakes, 37, 92
,PTQ'flesigned primarily for, 26, 111-115 sociability, 27, 75, 140, 1'62, see also Extroversion
and similarity-complementaritY in therapist, 112 social class of therapists' fathers, 127, 152
schools, therapeutic, 4-6, 13 , social competence; 84; 113, 148
Science and Moral Values .- social concern, 43, 80, 80n
and AB variables, 47, 62, 105 social desirability, 108-109
Coefficient Alpha, 27 social distance, 114
defined, 139 social expression, 2, 22, 103, see also SXPRSA,
in factor, analysis, 167 SXPRSS
and improvement in delinquents, 64 social group effects, 86
. Means ani standard deviations, 157-158 social treatment orientation, 30, 165
and measurement 'domains, 109:-110 socio-economic status, see social class
arid MP.- variables; 47-52 , sociopaths, 76, see also manipulators
and success with neurotics, 75 sodium amtytal; 15
scientific questions, see rei3earch.strategiee somatic orientation
scoring of soles., 17, 32, 58, 106 and the "A-B Scale," 22, 50 .'
SD (cluster) as an aspect of therapist's personality in the
and the " B Scale,"42, 161 therapeutic encounter; 161-165
and A-B riables, 40, 154 and MP variables, 165
and age O therapiste:152 among-PTQ respondents, 28, 29-31
clinical interpretation of, 79 somatology of schizophrenia, 15'
Coefficient-Alpha, 109 S OTHR, xi, 41, 42, 149
correlates compared with correlates ABDRUG, spending aspeCt of the Protestant Ethic, 107-108,
50, 161 113-114, 132, 162, 164
defined, xi, 36, 41-42 SQUE ST
in factor analysis, 47 and A-B variables, 40,.44, 154
and improvement in delinquents, 65 Coefficient Alpha, 109
means and standard. deviations, 149, 2r57, 158 defined, 37, 43
and MP variables, 47/155 and extroversion, 47, 102
and sex of therapist, 160 in factor analysis, 167 ,
and success When drugs are preicribed, 49-50 and improvement in delinquents, 65
selection of test items, see item selection means and standard deviations, 49, 157, 158
selection tests and MP variables, 47) 165 .
"A-B" Scale as, t8, 112, 115 and sex of therapist, 160
Criteriourbased,.103 and success when drugs are prescribed, 50
self-knowledge of the therapist, 7, 12 standardized tests, 102, 103
semantic reversal of A-B predictors and scales standards, personal, see Personal Standards
of;the "A-B Scale," 2,.22, 32-33, 78, 92, 161 statistical significance
of SD' and DBUSNS, 47, 155, 161 for clustered samples, 106-107
severity, 56, 56n, 57, 58-60, 78, 79, 160 for correlation matrix as a whole, 60n, 104
sex of clients, 41,-78, 92 of correlations between predictOrs with over-
sex of therapists. , tapping items, 35
anda.he "A-B Scale," 21, 29, 34, 41, 115 for`drug-free.predictors and MP variables in
and other A-B predictors, 152, 158, 160 , hospital and nonhosiital therapists, 45
shamanism, 10 for item analysis in high and low empathic in-
showing off, 164 terest groups, 23
side-show freaks,.38,1112, 92 , Strong Vocational Interest Blank; ii
significance,,see statistical significance Form M, for Men, 29, 34
similarity 'of therapists and clients, ,112-113, see long-term stability, 150
also matching some items in context, 143
situationals (delinquent) strutting for an audience, 164
and the "A-B Scale," 78, 79 subcultural identifiers (delinquent)
and A-B variables, 65 and A-B variables, 65.
defined, 57, 169-171 defined, 57, 170-171
helpers' improvement scores correl te with helpers' improvement scores correlated with
those for subcultural identifiers, 60-61 those for situationals, 60-61. ,
 SUBJECT INDEX

su bculti. ral identifiers

entifiers (cont.) and MP variables, 153, 156, 159
improvement of, 59 and need-fear dilemma, 76
and MP.variables, 64 and other tolerance scales, 107'
Quay Survey item analysis, 71 tolerance for complexity, see Need for Closure
severity of, 59 tolerance of schizophrenic pathology, 2, 76, 78,
SVIB item analysis, 69 6 -7 ' 95, 114
Subliminal conimunication, 6, 84, 93, 113 Tolerance of the Unrealistic Experience.
"Super A" therapist, 78 '- and A-B variables, 24', 47, 52, 155
superego, 27, 63; 135; see also Remorse and Antidemocratit, 137 '
supportive function of leadership, 76 Coefficient Alpha, 27
SXPIZSA defined, 133
iand the "A-B" Scale," 103 in factor analysis; 167
and A-B variables, 40, 44, 154 and improvement in delinq ents, 64
clinical interpretation, 75 and manipulatore vs. schizophrenics, 77
CoefficientAlpha, 109 means and standard deviations, 157, 158
' defined, id, 37, 43 and MP variables, 153, 156, 159
in factor analysis, 166107 ,- and other tolerance scales, 107
and improvement .in delinquents, 65 i . and rejection of business. interests, 80
a manifest scale, 161, 166 TOTL D and TOTL N
and A-B variables, 40, 154
means anal standard deviations, 149; 157, 158
and MP variables, 47, 155
-

and age of therapist, 152 /

a nascentethistruct, 103, 161 , , defined, 35, 41, 44
negative for subcultural identifiers, 78 and improvement in delinquents, 65
and success with schizophrenics when drugs are means and "standard deviations, -149; 157, 159.
prescribed, ',50 and MPvariables, 47, 155
SXPRSS. 1/ stability, long-term, 150-151
and A-Nvariables, 40, 44, 154 TL S
Coefficient Alpha, 109 Band ABDRUG, 49, 5142, 76
defined, xi, 37, 43 "And A-13 variables, 40,-154
in factor analysis, 167 A lifefin'ed, 35, 41, 44 .,
and improvement in delinquents, 65
1,1
d improvement in delinquents; 65
means and standard. deviations, 149, 157;1158 eans and standard deviations, 149, 157, 158
and MP variables, 47, 155 , ' and MP variables, 47, 155
6,, and success when.drtigs are prescribed,. 50 "stability, long-term, 190-151
, symptom reduction, emphasis on, 16, 24, 76', 93, -and success whew drugs afire prescribed, 50
see also tolerance of schizophrenic pathology training of therapists, 5-9, 12, 73, 82, 95
systematic observation, see observation, syatematic traumas of childhood, 74'; ,
4 : ...
turning against the self, 9
-technique, therapeutic, see process, theiapeutic unconditional regard, 8, 114
temperament, 11 understanding the sehizophrenic, see also- Em-
test construction, 102-103, '107-110, see also pathie Interest, 2, 16, 76
: i personality, Measurement .
unidimensionality, see internal consistency
unrealistic experience, age Tolerance of the Un-
Thematic Apperception Test, 'xi, 5, 8
therapist, defined, 73, see also adjustment prob- realistic Experience-
- .lems; aptitudes;, expectations; 'multiple
traits;, Plasticity; psycho-slcial treatment validation group,,. original, of therapists, see also
Orientation; schools; self-knowledge; sex; populations, research
. somatic orientation; training inpatient vs. outpatient, 79'
"therapiet variable" criticized, 7, 9, 18t 92 Loysville delinquents described, 56 .

time trends, 60, 148 male vs: female, 78

tolerance of ambiguity, 132 neurotics at Phipps and at Loysville, 54, 94
Tolerance of Ambivalence neurotics at Phipps and at VA outpatient clinics,
and A-B variables, 24,.47, 52, 155 34, 78
and age of therapist, 152 validation group, original of therapists, see also
and Antidemocratic scale, 137 criterion subjects; populations,. research
Coefficient Alpha, 27 compariEion of PTQ respondents with Phipps
defined, 134 residents, 21, 26, 28, 31, 51, 91-92
in factor analyais, .167 variance, critical role of variables in, 94-85
and improvement in delinquents, 64 . vari ce, crease in -psychwaocial therapies, 14, 14n
internal consistency vs. validatiOn, 109 Vet Administration outpatient clinics, pa-
means-and standard deviationst 157-158 , lents of, 19, 32, 41, 78-79, 148
: warmth, 8 , Work Ethic
Waiihiligton University., St. Louis, 16 and A-B variables, 47, 52, 155'
W-B 22 and W-B 23, see also ":A.-B Scale" Coefficient Alpha;
and the "A-B Scales" 32h, 3t, 41, . defined, 132 ".

and the A-B variables, 40; 154 in factor analysis, 167

Correlation between, ,32n, 41 and helpers' preferences for older youths, 160
Means and standard deviatiOns, _149, 157, 158 and improvement in delinquents, 64
and MP variables, 47, 155 - and improvement in depressives, 151
and SD, 42, 161 and means and standard deviations,-,157, 168
Work Activity, and 'MP variables,; 153, 156, 159
and A-B variables, 47, 155 and spending agpect of the Protestant Ethic, 164
Coefficient Alpha, 27 and Work Activity, 79n, 142
defined; 142 .. working through neurotic problems, 73-74
not in factor analysis, 166 worsening of clients in psycho-:social therapies,
means and standard dOviii.@ns,c1.57--,168 7, 8, 10-11, 14,14n, 88, 93
and MP variables, 153, 156
skewed, 105
). and Work Ethic, 790,'142 yielding to 'social pressure, 80n, 138, .

*U.S. GOVERNMENT PRINTING OFFICE: 1978 0-248-058