What Matters in College? Four Critical Years Revisited. by Alexander W.

Astin
Review by: Kenneth A. Feldman
The Journal of Higher Education, Vol. 65, No. 5 (Sep. - Oct., 1994), pp. 615-622
Published by: Ohio State University Press
Stable URL: http://www.jstor.org/stable/2943781 .
Accessed: 23/11/2014 14:29

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp

.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact [email protected].

Ohio State University Press is collaborating with JSTOR to digitize, preserve and extend access to The Journal
of Higher Education.

http://www.jstor.org

This content downloaded from 131.156.157.31 on Sun, 23 Nov 2014 14:29:45 PM

All use subject to JSTOR Terms and Conditions
 ReviewEssay

WhatMattersin College?Four CriticalYearsRevisited,

by AlexanderW. Astin.San Francisco:Jossey-Bass,1993.
xx + 482 pp. $30.95

KENNETH A. FELDMAN

The earlyyears of the 1990s have been good

onesforthestudyoftheimpactsofcollegeson students.In 1991Pasca-
rellaand Terenzini[14] publishedtheirmonumental reviewand synthe-
sis of the researchon collegeeffectsthathad accumulatedfrom1967
throughthe 1970sand 1980s.And now,two yearslater(as of thiswrit-
ing),comes AlexanderAstin'sreporton his revisitto the"fourcritical
years"of college.His initialvisit[1] was widelyhailedby researchers,
educators,policymakers,and othersforitscontribution to theirunder-
standingof how undergraduates are affectedby the kindsof colleges
theyattendand thekindsofexperiences theyhavethere.This newvisit,
if anything,is even moreproductivethanthe firstone. Not onlywas
Astinable to confirm, reject,modify, and otherwiseupdatehis earlier
findings,buthe has also expandedhispreviousstudybyincorporating a
numberof newvariables:measuresof thestudent'speergroupcharac-
of thefaculty's
teristics; values,attitudes,and preferred teachingmeth-
ods; of thefeaturesof theinstitution'scurriculum; and of thestudent's
performance on collegeadmissiontestsand on varioustestsused for
professional and foradmissionto graduateand professional
certification
schools.
Data forthenewstudywerecollectedfromnearlytwenty-five thou-
sand students(in 1985whentheywereentering collegeas freshmen and
thenabout fouryearslater)at overtwo hundredinstitutions of higher
educationas wellas fromfacultymembers (in 1989-90)at thesesamein-
stitutions.The studydesignincludedprovisions fordetermining (roughly)
what part of particularstudentoutcomescould be attributed to the
experienceof collegeratherthanto the maturation of the student or to
societal-wide change(seculartrends).

KennethA. Feldmanisprofessor
ofsociology ofNewYork
at theStateUniversity
at StonyBrook.
October1994)
Journalof HigherEducation,Vol. 65, No. 5 (September/
Copyright? 1994bytheOhio State University Press

This content downloaded from 131.156.157.31 on Sun, 23 Nov 2014 14:29:45 PM

All use subject to JSTOR Terms and Conditions
616 Journalof HigherEducation

For thecollegepopulationunderstudy,Astin'sprimary was in

interest
explainingcollegeeffects(controlledfor the of
characteristics entering
students)on someeighty-two studentoutcomes,whichrequiredthathe
workwitha coreofeighty-two regressionanalyses.Becauseofthelarge
numberofvariablesconsidered, thestudyran"theriskofsubjecting the
readerto information overload"(p. xiv). Astin,therefore, decidedon
certainsimplifying proceduresin an attemptto makethe reader'stask
moremanageable.One ofthesewas thedecisionto leaveouta greatdeal
of technicaldetailnot crucialforthe understanding of the basic find-
ings.Thus,fullregression tableshavenotbeengiven,and selectedbeta
and correlationcoefficientsare providedonlywhenthe actual magni-
tudeof an effectseemedto be of centralinterest.Althoughit wouldbe
usefulto knowR2foreach oftheregression analyses,thisstrategyofse-
oftechnicaldatadoesworkoutfairlywellin practice.
lectivepresentation
A moreproblematicdecision,perhaps,was to focus almostexclu-
sivelyon the undergraduate of traditionalage who enterscollegesoon
aftercompletinghighschool and to limitthe studyto studentswho
begintheircollegestudiesfull-time (althoughtheymay becomepart-
timestudents at somelaterpoint).Astinvigorously defendsthisdecision
to ignorethe largenumberof collegestudentswho are adultsor part-
timestudents, stating,

I believethatit would be a seriousmistaketo lumpthese"nontraditional"

studentstogetherwithtraditional-age full-time studentsin a singlestudy.
Anyonewho has workedwithadultsand part-timers (as I haveand as I as-
sumemanyreadershave) knowsfullwellthattheissuesand problemscon-
frontingtheadultsand thepart-time studentsare quitedifferent fromthose
confronting thetraditional-age
full-time student.It is also reasonableto as-
sume thattheenvironmental variablesthataffectone groupmaybe quite
differentfromthosethatare important to theother.By combiningthetwo
groups,we runa seriousriskof confounding thesedifferent effectson stu-
dentdevelopment. In otherwords,itis farbetterto obtainclear-cutfindings
on an important and well-definedpopulation(thefull-time undergraduate of
traditionalcollegeage) thana watered-down set of conclusionsbased on a
muchmoreheterogeneous sample(pp. xviii).

The studentoutcomesin WhatMattersin College?are groupedinto

the followingsevencategories(withthe discussionof the findingsfor
each categoryforming a separatechapter):politicalidentification
ofthe
student;personality,selfconcept,and psychological well-being;
attitudes,
values,beliefs,and lifegoals; patternsofbehavior;academicand cogni-
tivedevelopment; careerdevelopment; withvariousas-
and satisfaction
pectsof college.Approximately 325 studentinputand collegeenviron-
mentalvariablesare used - eithersinglyor combinedintomulti-item

This content downloaded from 131.156.157.31 on Sun, 23 Nov 2014 14:29:45 PM

All use subject to JSTOR Terms and Conditions
 ReviewEssay 617

scores- to predicttheseoutcomes.Theseinputand environmental var-

iables are introducedinto the regressionanalysesin four"blocks" of
variables:(1) characteristics of the entering freshmen understudy;(2)
so-called"bridge"variablesthatcan be consideredcharacteristics ofthe
entering student(inputs)as wellas characteristics ofthestudent's educa-
tionalexperience (environments) - namely, thefreshman's financialaid
package,place of residence, and initialchoiceof a majorfieldof study;
(3) between-collegecharacteristics ofthestudent's environment (charac-
teristicsof the institution attended,curricularequirements, student's
peer environment, and facultyenvironment); and (4) measuresof the
quantityand quality of student involvement within thecollege(involve-
mentwithacademicwork,withfaculty, withstudentpeergroups,with
workforpay,and withmiscellaneousothernonacademicmatters).The
basic idea behindthetypeof regression analysesused in thestudywas
to controlthesedifferent blocksofvariablesaccordingto theirknownor
assumedorderof occurrence.
Each of the sevenchapterson studentoutcomespresentsdozens of
individualfindings, whichare also listedin easy-to-follow summaryta-
bles. These individualfindingsare thendrawntogether and interpreted
in two globalsummarychapters- one on "environmental effects" and
one on the"effects ofinvolvement." Even thefinalchapterof thebook
on the implications of thefindings foreducationaltheoryand practice
servesin partas a summarychapter.
One ofthemajorfindings oftheresearchis thepervasiveeffect ofthe
peer groupin the individualstudent'sdevelopment.Astinputsit this
way:"Viewedas a whole,the manyempiricalfindings fromthisstudy
seem to warrantthe followinggeneralconclusion:the student'speer
group is thesinglemostpotentsourceof influenceon growthand de-
velopmentduringthe undergraduate years"(p. 398, emphasisin the
original).Astinnotesthateveryaspectof thestudent's development -
cognitive, affective,
psychological, and behavioral- is affected in some
waybypeergroupcharacteristics. Moreover,thevalues,beliefs,and as-
pirationsofindividualstudentsgenerally tendto changein thedirection
of the dominantvalues,beliefs,and aspirationsof the peergroup-
whichAstinsees as a confirmation of notionsof "progressive confor-
mity"[2] and "accentuationof initialdifferences" [8, 9]. In elaborating
on thesenotionsand in developinga theoryofpeergroupeffects, Astin
buildson theviewofa peergroupas a groupofindividuals"withwhom
the individualidentif7es and afJiliates and fromwhomthe individual
seeksacceptanceor approval"(p. 400, emphasisin original).Although
Astindoes notstatethis,his theoryof peergroupeffects, includingthe
severalhypotheses he lists(on p. 402) to be testedin futureresearch,fits

This content downloaded from 131.156.157.31 on Sun, 23 Nov 2014 14:29:45 PM

All use subject to JSTOR Terms and Conditions
618 Journalof HigherEducation

nicelywithexistingtheoryin thefollowingareas: membership groups;

referencegroups;socialization;interpersonalinfluence, attraction,and
affiliation;
interpersonalenvironments and peer cultures;and subcul-
turesin general[see,forexample,5, 6, 10, 11, 12, 13, 15, 17, 18].
The characteristicsand behaviorsof facultyalso have important im-
plicationsforstudentdevelopment.For instance,attendinga college
whosefacultyis heavilyresearchorientedincreasesstudentdissatisfac-
tionwithvariousaspectsof collegeand has negativeimpactson most
measuresofcognitiveand affective development (althoughcertainstan-
dardized tests- GREs, LSATs - are positivelyaffected).Attendinga
collegethatis strongly orientedtowardstudentsand theirdevelopment
showstheoppositepatternofeffects.
Anotherimportant generalfinding is that,althoughinstitutional type
and controlare associatedwitha numberofstudentoutcomes,suchin-
stitutionalcharacteristicshavelittledirecteffect on theseoutcomesonce
peergroupand facultycharacteristics are takeninto account.That is,
mosteffects of institutionaltypeare indirect;theyare mediatedby fac-
ulty,peer group,and studentinvolvement variables.Moreover,and
somewhatsurprisingly, the formof the institution's generaleducation
curriculumhas a littledirectimpact on student development.Only a
"true-core curriculum" thatrequiresall students to takeexactlythesame
courseappearsto havedistinctive effects on studentoutcomes(in par-
ticular,highsatisfaction withvariousaspectsof collegeand positiveef-
fectson leadership).
Whatstudentsarelikewhentheyentercollegeis hardlyunimportant,
of course. Astinreportsparticularly interestingdifferentialcollegeef-
fectson studentsaccordingto theirgender,race, and socioeconomic
status.He finds,forexample,thatcollegesin theyearsof hisstudydid
notserveto eliminateor evenreducemanyofthestereotypic differences
betweenmale and femalestudents.Womenenteredcollegealreadydif-
fering considerably frommenon self-rated emotionaland psychological
health,standardizedtestscores(such as GREs), grade-point averages,
politicalattitudes,
personality characteristics,and careerplans;and most
of thesedifferences widenedduringtheundergraduate years.Likewise,
whiteand African-American studentsshowedcontrasting patternsof
changeon a numberofaffective variables.Theygrewapartin theirpolit-
intheiragreement
ical attitudes, withtheproposition thatracialdiscrim-
inationis no longera problem,in theircommitment to the goal of
promotingracial understanding, and in theirlikelihoodto engagein
campusprotests. Finally,students fromfamiliesofhighersocioeconomic
background, comparedto otherstudents, enjoyeda varietyofmorepos-
itiveoutcomesin college,regardlessoftheirabilities,academicprepara-

This content downloaded from 131.156.157.31 on Sun, 23 Nov 2014 14:29:45 PM

All use subject to JSTOR Terms and Conditions
 ReviewEssay 619

tion,or othercharacteristics.Astinarguesthatthesedifferential effects

forstudentsof different gender,race, and socioeconomicstatusmay
havebeenattributable to peergroupeffects - due in essenceto the"dif-
ferentialassociation"(myuse of thisphrase,not his) of thesedifferent
groupsof students.
As forstudentinvolvement in college,Astin'sresearchonce again
pointsto its unusuallystrongpotentialforenhancingmostaspectsof
theundergraduate student'scognitiveand affective development. Thus,
learning,academicperformance, and retention associated
are positively
withacademicinvolvement, involvement withfaculty, and involvement
withstudentpeergroups.Moreover,a widespectrumof cognitiveand
affectiveoutcomesis negativelyaffectedby formsof involvement that
eitherisolatethe studentfrompeersor removethe studentphysically
fromthecampus(livingat home,commuting, beingemployedoffcam-
pus,beingemployedfull-time, watchinga lot oftelevision,and thelike).
I should point out - as does Astin himself- that the findingsfor
studentinvolvement are not withoutinterpretiveuncertaintiesand am-
biguities.This is so becausevariablesin thisresearchweremeasuredat
onlytwo pointsin time- at collegeentranceand at the end of four
yearsofcollege.Thus,one setofinvolvement variables- theaforemen-
tioned"bridge"variables has to do withformsof involvement
- that
wereascertainedat thepointwhenthestudentinitiallyenteredcollege:
freshmanplace of residence,financialaid package,initialchoice of a
major fieldof study(upon collegeentrance).The student,of course,
mayhavechangedhisor herplace ofresidenceafter(or evenduring)the
freshmanyearor maynot have majoredin his or herinitialchoice of
majorfields,butthisstudywas not able to takethesechangesintoac-
countin a clear-cutway.The secondsetof involvement measurescame
fromthefollow-upquestionnaires given to studentsafter fouryearsof
college.The difficulty hereis thatone cannotbe sureexactlywhenthe
studentexperienced a particularformofinvolvement as recordedin the
follow-upquestionnaire. Thus,evenassumingthatstudents'retrospec-
tionsabout the natureand frequencyof theirinvolvement in college
wereaccurate,itmaystillbe thatsomechangeduringcollegein thestu-
dents'attitudes, values,or othercharacteristicsled to changesin his or
herparticularsortsof involvement in collegeratherthan thatcertain
involvements in collegeproduceda changein thestudents, whichthen
registeredin thefollow-up Thatthesesortsofpossibilities
questionnaire.
and concomitantcausal uncertainties existin the presentstudyunder-
scoresthe continuingneed forresearchthatgathersinformation from
studentsduringtheircollegeyearsand implicitly suggeststheusefulness
ofqualitativeand ethnographic studieswiththeirfocuson socialprocess

This content downloaded from 131.156.157.31 on Sun, 23 Nov 2014 14:29:45 PM

All use subject to JSTOR Terms and Conditions
620 Journalof HigherEducation

and individuals'interpretations of thisprocess[see, forexample,3, 4,

16].
In thelastchapter,entitled"Implications forEducationalTheoryand
Practice,"Astin suggestsseveralpossibleapproachesto ameliorating
some of the problemsconfronting undergraduate educationin the re-
searchuniversity and in otherkindof institutions thatoveremphasize
researchor haveweakpeergroupstructure. He also suggestshowpeda-
gogicalpracticesand generaleducationprograms couldbe strengthened
so as to increasetheirpositiveinfluence on students' cognitive and affec-
tivedevelopment. In one sectionof thischapter,he discussestheimpli-
cationsof the weightof the empiricalevidencethatshows a college's
emphasison diversity and on studentparticipation in diversityactivities
to have broadlypositiveeffects on students'cognitiveand affective de-
velopment.In anothersection,he appliesthe findingsof his studyto
variousissuesofresourceallocationin colleges.Most ofthesuggestions
foundin thischapterforimproving theeffectiveness ofpolicyand prac-
ticein highereducationare sensible;some are provocative;all of them
are groundedin empiricalevidence.
Astindoes notattempta full-scaleintegration ofhisfindings withthe
existingliterature in thefield:"I havetriedto makereference to earlier
researchwhenthe resultseitherreplicateor contradictmyresults,but
suchreferences are restricted forthemostpartto earlierstudiesthatare
nationalin scope"(p. xix). Moreover,althoughexplicitconnections are
madeto themorecomprehensive reviewsand syntheses inthearea [7, 8,
14],thesetie-insare morein theformofbroad-grained referencing than
finelyhonedarticulations. Fullerintegration willcome in thefutureas
researchers and scholarsincorporate theideas and findings ofthebook
intotheirown researchand writing.
This book is written in a straightforward, forthright, and unpreten-
tiousmanner.Astinis a particularly honestwriter.He does not tryto
coverup or getaroundissuesin thefield,problemsofinterpretation, or
limitations in thedata. Whenan interpretation is purelyspeculative,he
saysso. Whenhe is notable to figureoutwhya correlation betweentwo
variablesexists,he also saysso (and letsit go at that).He intendsthis
bookfora wideaudience:on p. xii he listsfaculty, studentaffairs profes-
sionals,administrators, policy-makers, and prospective college students
and theirparents(curiously, he does not includeresearchers of higher
educationon thelist).He has done hisbestto makethebook readable
and understandable forthosereaderswho mightnotbe particularly fa-
miliarwiththecurrent languageand formatofsocial scientific research
and writing, although,conceivably, thebook mayrequirea readerto be

This content downloaded from 131.156.157.31 on Sun, 23 Nov 2014 14:29:45 PM

All use subject to JSTOR Terms and Conditions
 ReviewEssay 621

somewhatbetterversedin the waysof social science(and quantitative

methodology) thanhe hoped.
Those of us workingin thefieldof highereducationhave longbeen
indebtedto theefforts of AlexanderAstin.Startingin theearly1960s,
he was a majorpioneerin themodernstudyofcollegeimpacts,withits
use of techniquesof multivariate analysisand multipleregression pro-
ceduresto controlforstudent"input"in determining theeffectsof col-
legeenvironments on student"output."Throughout thepastthirtyyears,
bothbyexampleand exhortation he has askedus to do morethantheo-
rize,philosophize, polemicize,and fantasizeabouthighereducationand
its effects.Rather,he has been indefatigable in insistingthatthe best
possibleempiricalevidencebe broughtto bear on findingout how col-
legesand universities affecttheirstudentsand on suggesting howthese
institutions of higherlearningmightimprovetheirown effectiveness.
Withthepublicationof WhatMattersin College?,ourdebtto himhas
increasedyetagain.

References
1. Astin,A. W. Four Critical Years: Effectsof College on Beliefs,Attitudes,and
Knowledge.San Francisco:Jossey-Bass,1977.
2. Astin,A. W., and R. J. Panos. The Educationaland VocationalDevelopmentof
CollegeStudents.Washington, D.C.: AmericanCouncilon Education,1969.
3. Attinasi,L. C. Jr."GettingIn: Mexican Americans'Perceptionsof University At-
tendanceand theImplicationsforFreshmanYear Persistence." Journalof Higher
Education,60 (May/June1989),247-77.
4. . "Phenomenological Interviewingin theConductof Institutional
Research:
An Argumentand an Illustration." Associationfor Institutional
ResearchProfes-
sional File, No. 38 (Fall-Winter1990-91), 1-8.
5. Becker,H. A., B. Geer,and E. Hughes.Makingthe Grade: TheAcademicSide of
CollegeLife.New York:Wiley,1968.
6. Berscheid,E. "Interpersonal In Handbook of Social Psychology.Vol.
Attraction."
2. Special Fields and Applications,editedby G. Lindzeyand E. Aronson,pp.
413-84. New York:RandomHouse, 1985.
7. Bowen,H. R. Investment in Learning:The Individualand Social Valueof Ameri-
can HigherEducation.San Francisco:Jossey-Bass,1977.
8. Feldman,K. A., and T. M. Newcomb.The Impact of College on Students.San
Francisco:Jossey-Bass,1969.
9. Feldman,K. A., and J. Weiler,"Changesin InitialDifferences amongMajor-Field
Groups:An Explorationofthe'AccentuationEffect."'In School and Achievement
in AmericanSociety,editedby W. H. Sewell,R. Hauser,and D. Featherman,pp.
373-407. New York:AcademicPress,1976.
10. Merton,R. K. Social Theoryand Social Structure(enlargededition).New York:
Free Press,1968.

This content downloaded from 131.156.157.31 on Sun, 23 Nov 2014 14:29:45 PM

All use subject to JSTOR Terms and Conditions
622 Journalof HigherEducation

11. Newcomb,T. M. The AcquaintanceProcess.New York:Holt,Rinehartand Win-

ston,1961.
12. Newcomb,T. M., R. H. Turner,and P. E. Converse.Social Psychology:TheStudy
ofHuman Interaction.New York:Holt,Rinehartand Winston,1965.
13. Newcomb,T. M., and E. K. Wilson(eds.). College Peer Groups: Problemsand
Prospects for Research.Chicago:Aldine,1966.
14. Pascarella,E. T., and P. T. Terenzini.How CollegeAffectsStudents:Insightsfrom
TwentyYearsof Research.San Francisco:Jossey-Bass,1991.
15. Singer,E. "Reference Groupsand Social Evaluations."In Social Psychology: Socio-
logical Perspectives,editedby M. Rosenbergand R. H. Turner,pp. 66-93. New
York:Basic Books, 1981.
16. Tierney,W. "An Anthropological Analysisof StudentParticipationin College."
Journalof HigherEducation,63 (November/ December1992),603-18.
17. Wallace,W. L. StudentCulture:Social Structure and Continuity in a LiberalArts
College.Chicago:Aldine,1966.
18. Wheeler,S. "The Structureof FormallyOrganizedSocializationSettings."In So-
cializationafterChildhood:Two Essays,by 0. C. Brim,Jr.and S. Wheeler,pp.
51-116. New York:Wiley,1966.

This content downloaded from 131.156.157.31 on Sun, 23 Nov 2014 14:29:45 PM

All use subject to JSTOR Terms and Conditions