PERSONNEL PSYCHOLOGY

1981, 34

SHIFT WORK RELATED TO JOB ATTITUDES,

SOCIAL PARTICIPATION AND WITHDRAWAL
BEHAVIOR: A STUDY OF NURSES AND
INDUSTRIAL WORKERS'
MUHAMMAD JAMAL
Concordia University

The present study investigates the relationship between shift

schedules and mental health, job satisfaction, social participation,
organizational commitment, anticipated turnover, absenteeism and
tardiness among nurses (iV=440) in two hospitals and among rank-
and-file workers (iV=383) in a manufacturing organization. Results
are generally supportive of the model which projected that workers
on fixed work schedules (high routine-oriented) would be better off
than workers on rotating work schedules (low routine-oriented) in
terms of mental health, job satisfaction, organizational commitment
and social participation. In addition, workers on fixed shift sched-
ules are found to be lower on anticipated turnover, absenteeism and
tardiness than workers on rotating shift schedules. Six potential
moderators: age, marital status, place of socialization, cultural back-
ground, seniority and respondent's sex, were measured and their as-
sociation with these relationships analyzed. Results are discussed in
the light of the previous empirical evidence.

APPROXIMATELY 25-26% of the labor force in the U.S. and Canada

is involved with shift work {Journal of International Management,
1977; Tasto and CoUigan, 1977). Despite the prevalence of shift work
practices in the U.S. and Canada, little is known about the effect of
shift work on workers' socio-psychological make-up and work and
non-work behavior. Some research has been conducted into worker
response to variation in the distribution of working hours across a
twenty-four hour cycle, with relationships reported between different
work shifts and individual characteristics both on and off the job (Aa-

' ' The author wishes to thank A. Ashour, V. Baba, R. Crawford and P. Frost for their
helpful comments on an earlier draft of this paper. Requests for reprints should be sent
to Muhammad Jamal, Department of Management, Concordia University, Loyola
Campus, 7141 Sherbrooke Street West, Montreal, Quebec, Canada, H4B 1R6.
535
536 PERSONNEL PSYCHOLOGY

nonsen, 1964; Colligan, Frockt, and Tasto, 1979; Frost and Jamal,
1979; Malaviya and Ganesh, 1976; Nicholson, Jackson, and Howes,
1979; Mott, Mann, McLaughlin, and Warwick, 1965). In recent years,
another stream of research has focused on worker response to varia-
tion in the distribution of hours in the model work week, typified by
40 hour, 9 a.m. to 5 p.m., five-day work schedule. Most variation and
attendant research has centered on the forty-hour, four-day work
week and on flextime work schedules which comprise a daily core
work period as well as fiexible starting and finishing times (Evans,
1975; Mahoney, Newman, and Frost, 1975; Nord and Costigan, 1973).
Despite differences in assumptions and purpose, one research find-
ing common to many studies of these two approaches to work sched-
uling is that worker response to the arrangement of work hours tends
to be in terms of its implications for the worker's emotional well-being
and non-work activities. Research on shift work has provided data
which suggest that worker perceptions of the degree to which work
hours impede the pursuit of leisure time activities are related to in-
dicators of emotional well-being such as anxiety, conflict, pressure
and self-esteem (Mott et al., 1965). One way to study relationships be-
tween hours of work and worker responses is the degree to which
work hours lead to a daily routine formation for the individual. Work
hours which tend to be high on routine formation provide individuals
with relativelyfixedhours for work and non-work activities every day.
Individuals on such work schedules can plan to fulfill family responsi-
bilities, take part in regularly scheduled social activities and cope with
physical and mental fatigue better than individuals who work on vari-
able work hours (low routine formation). Variable work hours (ro-
tated shift) lead an individual to expect to move from one work sched-
ule to another work schedule withfixedintervals, which could be after
every ten working days. Individuals on such work hours experience a
low degree of routine in everyday life and remain in a constant chang-
ing and adapting mode. By the time they have adapted to one work
schedule, they are expected to move to the next one.
It is suggested that individuals would view routine-oriented work
schedules more favorably and their responses would indicate higher
job satisfaction, emotional and psychological well-being, and organi-
zational commitment than individuals on low routine-oriented work
schedules. In addition, individuals on routine-oriented jobs would be
low in absenteeism and tardiness and would exhibit greater desire to
remain with the present employer than individuals on low routine-ori-
ented jobs.
 MUHAMMAD JAMAL 537

Method
Research Setting
The present study was conducted in two different work settings: a
hospital and a manufacturing organization. Two large hospitals serv-
ing the English population of the city of Montreal, Quebec, were in-
cluded in the study (henceforth called 'hospital sample'). Both hospi-
tals were under the direct administration of the Province of Quebec
and thus were expected to have similar administrative practices. The
manufacturing organization (henceforth called 'manufacturing
sample') was a medium-size organization employing 600 rank-and-file
employees. The organization was involved in producing small con-
sumer products.
Subjects
Hospital Sample. AU nurses in both hospitals were included in the
study. Attitudinal data were collected through a stmctured question-
naire. Approximately 580 nurses in both hospitals were given ques-
tiormaires, and 440 usable questionnaires were received. The response
rate of subjects working various shifts was proportionate to the actual
distribution of nurses on all shifts. Absenteeism and tardiness data
were obtained from hospital records for each respondent for 4
months, from the day the questionnaires were distributed. Fifty-three
percent of the respondents were single, 39% were married, and 8%
were widowed, divorced, or separated. The average age was 31 years;
average monthly income, $1065; and average length of service, 6
years. The majority of nurses were female (97%); were raised in large
cities (77%); and had English as a mother tongue (57%). All nurses in
both hospitals belonged to the same union. Distribution of the re-
spondents on demographic and background variables such as age, sex,
marital status, place of socialization, cultural background, and senior-
ity was examined according to the type of shift they worked. Except
for age, no statistically significant differences were observed between
nurses on two shifts. The average age of the nurses working on fixed
shift was 34 years and for those working on rotating shift it was 27
years (p < .01).
Manufacturing Sample. All rank-and-file employees were included
in the study and were given copies of the stmctured questionnaire.
With one follow-up, 383 usable questiormaires were returned. The re-
sponse rate of employees working on rotating shifts was slightly lower
(6%) than the actual distribution in the company. The average age of
538 PERSONNEL PSYCHOLOGY

the respondent was 35 years, and the average monthly income was
$1124. The majority of the respondents were married (60%); were
male (70%); belonged to a union (68%); and were raised in large cities
(76%). Forty-four percent of the respondents had grade 9 or less edu-
cation, 22% had grades 10-13 education, and only 36% had some col-
lege education. Distribution of the respondents with regard to age,
sex, education, skill level, marital status, place of socialization and
cultural background was examined according to the type of shift
worked. Except for sex, no statistically significant differences were
found among employees on fixed and rotating shifts. Proportionately
more male employees (93%) worked on rotating shifts than female
employees (7%) and the difference was significant (p < .01).
Measures
Hours of Work. The high routine-oriented work schedule was oper-
ationalized in terms of the fixed working hours at any of the work
shifts (day, evening, night) during the 24 hours. Employees working
onfixedday, aftemoon, and night shifts were thus combined to form a
single category called ^e^? work schedule. In both samples, the num-
ber of people working on permanent evening (Hospital, iV=31; Manu-
facturing, N=5) and permanent night (Hospital, A'^=14; Manufactur-
ing, N=\l) shifts was too smaU to be treated as separate categories.
The rotating work schedule was kept as a second category of work
hours representing a low routine-oriented work schedule.
Mental Health. Measures of emotional weU-being comprised
worker responses to a 21-item scale developed to tap Komhauser's six
indexes of mental health (Komhauser, 1965). Following Komhauser,
a composite index of mental health was constmcted by combining
scores on his six indexes of mental health. The composite index was
used in the present study. Development of the scale and a fuU dis-
cussion of reliability and vaUdity are presented elsewhere (Jamal,
Bamowe, and MitcheU, Note 1; Jamal and MitcheU, 1980).
Job Satisfaction. Job satisfaction was assessed by the Brayfield and
Rothe (1952) scale. The scale consists of 18 Likert-type items ranging
from "strongly agree" to "strongly disagree."
Social Involvement. Social involvement was operationaUzed in
terms of reported participation in voluntary organizations (e.g.,
church or ethnic group, civic or social clubs). Measures were obtained
for each individual of the number of meetings attended and the hours
each individual spent in voluntary organization activities.
Organizational Commitment. Organizational Commitment was
measured with Porter and Smith's scale (Note 2). This scale consists of
15 Likert-type items ranging from "strongly agree" to "strongly dis-
 MUHAMMAD JAMAL 539

agree." This scale had previously been found to be reUable (Jamal,

1976; Porter, Steers, Mowday and BouUan, 1974).
Anticipated Turnover. Anticipated tumover was assessed by asking
each worker to state the probabiUty of his/her staying with the same
organization for 2 years, from the day the questionnaire was an-
swered. This measure has been reported to be highly correlated with
actual tumover (Kiaut, 1975).
Absenteeism/Tardiness. Both absenteeism and tardiness were as-
sessed by the actual incidents of absenteeism and tardiness reported in
organization records for each individual for 4 months, from the day
the questionnaires were distributed. In both cases, frequently of ab-
senteeism and tardiness was measured, rather than duration. Ab-
senteeism and tardiness data were available only for the hospital
sample.
Results
Given the large number of dependent variables used in the study,
product-moment correlations among the variables were computed and
analyzed for evidence of possible overlap. The intercorrelation matrix
is reported in Table 1. ReUabiUty coefficients are also reported in
Table 1. Reliability coefficients of the variables "mental health," "job
satisfaction," and "organizational commitment" are reasonably high
for both samples, indicating their usefulness for the present study. The
highest correlation was between meetings attended and participation
hours in voluntary organizations. This relationship is expected and
logical because of the concomitant dependence of one variable on the
other variable. Other correlations in the matrix vary from very low (r
= .04) to a moderate degree (r = .53), suggesting a low degree of over-
lap among variables. Thus, aU variables were retained for further
analysis.
Correlations of demographic and background variables with depen-
dent variables of the study were also computed and analyzed for po-
tential effects on the dependent variables other than the shift sched-
ules. These correlations are also reported in Table 1. The average
correlation of demographic and background variables with 8 depen-
dent variables was .085 for the hospital sample and .076 for the manu-
facturing sample. Although there were several significant correlations
in both samples, the magnitude and the direction of these correlations
seemed to be consistent with the previous research (Baba and Jamal,
1979; Hagedom and Sanford, 1968; Lenski, 1956; Muchinsky, 1977;
Porter and Steers, 1973; WUensky, 1960). Therefore, it was felt logical
to examine the relationships between shift schedules and various de-
pendent variables.
540 PERSONNEL PSYCHOLOGY

o
V

S
r? V
II)
bles
(

a O

§ w IX
«?
§ 8 2
^—.s

00

r r
-I
n1
e

^5
I w
. O p -^ p p "
60
C

loo— ooooooooo — ooo

g
o
1 s^
"^^
. I I 1 I
•a
I:
Int

S
Ction

1 3 I I
iables

ffi m
'G
3 oj
g I I <2 I M I
g *c
O
>->
en ^ iri "o r-'oooN O --^ <N en •^
 MUHAMMAD JAMAL 541

The association between shift schedules and various criterion vari-

ables was analyzed using the one-way analysis of variance (ANOVA),
appropriate for unequal ceU frequencies. Results are reported in
Table 2 (i^ values were estimates, given unequal ceU sizes). Significant
differences existed between fixed shift employees and rotating shift
employees on aU of the criterion variables for the hospital sample.
Nurses working on fixed shifts were higher on the average in mental
health, job satisfaction, social involvement and organization com-
mitment than nurses on rotating shifts. They were lower on antici-
pated tumover, absenteeism and tardiness than nurses working on ro-
tation shifts.
In the manufacturing sample, significant differences existed be-
tween fixed shift and rotating shift employees on three of the five cri-
terion variables. Rank-and-file workers on fixed shifts were higher on
the average in mental health and in participation hours in voluntary
organizations than workers on rotating shifts. Fixed shift employees
were less desirous of leaving their present employer than rotating shift
workers. No significant differences were found between fixed shift and
rotating shift workers in terms of job satisfaction and meetings at-
tended in voluntary organizations.
Two-way analyses of variance were performed to test interactive ef-

TABLE 2
Shift Schedules Related to Mean Scores on Mental Health, Job Satisfaction, Participation
in Voluntary Organizations, Anticipated Tumover, Organizational Commitment,
Absenteeism and Tardiness

Fixed' Rotating
N = 245" N=ISO
Criterion Variables (JV = 289) (N= 104) F P
1. Mental Health 87.06 83.46 9.76 .01
(76.03) (73.09) (6.79) (.01)
2. Job Satisfaction 72.50 68.35 12.70 .01
(60.31) (58.81) (2.16) (ns)
3. Meetings Attended" 1.76 1.04 7.09 .01
(2.39) (1.93) (1.09) (ns)
4. Participation Hours'^ 2.88 1.79 4.64 .05
(3.81) (2.47) (5.60) (.05)
5. Anticipated Tumover 2.31 2.79 17.23 .01
(2.60) (2.93) (9.04) (.01)
6. Organizational Commitment 49.91 46.72 10.75 .01
7. Absenteeism 1.39 1.60 4.10 .05
8. Tardiness 1.75 2.18 7.19 .01
* Includes people working onfixedday, evening and night shifts.
^ Niimbers without parentheses are from the hospital sample, and numbers in parentheses are from the manufac-
turing sample.
*^ Meetings attended and participation hours in voluntary organizations per '4-week period.
542 PERSONNEL PSYCHOLOGY

fects of shift schedules and six potential moderators (age: younger vs.
older; marital status: married vs. single; place of socialization: rural vs.
urban; cultural background: French-speaking vs. English-speaking;
seniority; and sex: male vs. female) on various criterion variables. The
results are presented in Table 3. In the hospital sample, out of a pos-
sible forty interaction effects, only four effects were significant beyond
.05 level. Older nurses on fixed shifts attended more meetings of vol-
untary organizations than did the younger nurses on fixed or rotating
shifts. Married nurses on fixed shifts showed a higher level of organi-
zational commitment than single nurses on fixed or rotating shifts.
Married and single nurses on fixed shifts showed the same level of ab-
senteeism, but single nurses on rotating shifts showed the highest level
of absenteeism as compared to married nurses on rotating shifts.
French-speaking and English-speaking nurses on fixed shifts showed
an identical level of organizational commitment, but the English-
speaking nurses on rotating shifts showed a significantly lower level of
organization commitment than the French-speaking nurses on rotat-
ing shifts.
In the manufacturing sample, six out of 25 possible interactive ef-
fects were significant beyond the .05 level. Younger workers on fixed
shifts reported higher participation hours in voluntary organizations
and were less desirous to leave their present employer than older
workers on fixed or rotating shifts. Single workers on fixed shifts re-
ported greater number of participation hours in voluntary organiza-
tions than did married workers on fixed or rotating shifts. Single
workers on rotating shifts reported the least number of hours spent in
voluntary organizations. French-speaking workers onfixedshifts were
less desirous to leave their present organization than English-speaking
workers on fixed or rotating shifts. Male and female workers on fixed
shifts had the same level of mental and emotional health, but female
workers on rotating shifts (N = 8) enjoyed better mental health than
male workers. Both male and female workers on fixed shifts showed
almost the same level of desire to leave the present organization, but
females on rotating shifts (N = 8) were the least desirous to leave their
present organization.
As previously reported, significant differences were found between
fixed shift and rotating shift employees with regard to age in the hos-
pital sample (F = 12.9, p < .01) and sex in the manufacturing sample
(F = 8.9, p < .01). Analyses of covariance between fixed vs. rotating
shift and each of the dependent variables in the study, with age as a
covariate in the hospital sample, and sex as a factor in the manufac-
turing sample, provided results no different from those observed ear-
Uer.
 MUHAMMAD JAMAL 543

O
< I P I f^ I t . I O\
- ^ o '"^r-i "-^O •—'O

I
1

S. ^vo
1 -
-^a .-vvo z.^—
I p I -. I q
^1 o ^ •o ^"^ ^^o
O SI

I
I >
— vo — O O O O <

• *O 00 O »/^ •^ O »A

.1 II

I 3S
II
I
I 1
S o
5 -S f iS^

X »X'gX
55 "5
1 t X'L
•g U -g "^
CO en
544 PERSONNEL PSYCHOLOGY

Discussion
The results of the present study indicate that workers on fixed
shifts, the measure of a high routine-oriented work schedule, appeared
to be "better off" than workers on rotating shifts (a low degree of rou-
tine-oriented work schedule) in terms of mental and emotional health,
job satisfaction, social involvement, tumover intentions, and organi-
zational commitment in the hospital sample. Similar results were ob-
tained in the manufacturing sample with regard to mental health, an-
ticipated tumover, and number of hours spent in participation in
voluntary organizations.
The fact that in both samples results are generaUy in accordance
with our earlier expectations lends support to the notion that the de-
gree of routine orientation in work scheduling is related to workers'
attitudes and behavior both in work and non-work spheres. This sup-
port seems to be more cmcial because of the nature of the two samples
which represented two entirely different work settings. However, be-
fore results are discussed any further, a note of caution is warranted
conceming routine formation in shift schedules. The degree of routine
as experienced or perceived by employees was not directly assessed in
the present study. Employees were not asked to express whether or
not shift schedules lead to routine formation in their Uves. It was im-
plied that employees on fixed shifts would have a high degree of rou-
tine formation, while employees on rotating shifts would have a lower
degree of routine. A fruitful area of future research would be to exam-
ine the relationship between shift schedules and employees' per-
ception of routine formation.
The findings of the present study, in general, support the con-
clusions of previous studies: that workers assigned to rotating shifts
are prone to greater health problems than workers who are per-
manently assigned to day, aftemoon, or even night shifts. Taylor
(1967) concluded that workers onfixed(night) shifts suffered fewer ill-
nesses than those on rotating shifts. A recent study of hospital employ-
ees by CoUigan, Frockt and Tasto (1979) concluded that "relative to
nurses on permanent shifts, rotating shift nurses exhibited a signifi-
cantly higher rate of clinic visits and took more sick days for serious
illness." In their study, upper Gl tract distress and respiratory infec-
tions were the two greatest causes of illness among nurses on rotating
shifts, which differed significantly from the figures for fixed shift
nurses. Wyatt and Marriott (1953) in a study of rotating shift workers
in Britain concluded "that the change from day to night work has dis-
turbing effects on worker appetite and digestion." A higher incidence
of ulcers among shift workers was also found by Thiis-Evensen (1958)
in both his Danish and Norwegian sample. However, Mott et al.
 MUHAMMAD JAMAL 545

(1965) found a higher incidence of ulcers among workers on per-

manent day and evening shifts than among those working night and
rotating shifts.
The results on absenteeism and tardiness indicate that nurses on
fixed shifts were, on the average, lower in absenteeism and tardiness
than nurses on rotating shifts. While the previous research on shift
schedules and absenteeism and tardiness has not been equivocal, the
results of the present study seem to be logical and vaUd. Almost all
previous studies on work schedules and absenteeism had primarily
male subjects. It could be the case that there is no consistent relation-
ship between work schedules and absenteeism for males (Sergean &
Brierley, 1968; Taylor, 1967; Taylor and Pocock, 1972; Thiis-Evensen,
1958), but there may be a relationship between work schedules and
absenteeism for females. Some support for this argument comes from
study of CoUigan et al. (1979) who found a significant difference on
the duration of sick absence between fixed shift nurses and rotating
shift nurses. However, only future research will be able to clarify this
relationship.
Our results from moderator analyses indicate the complexity of in-
terpretation of the reported findings. Not a single criterion variable
out of eight variables was found to have the same interactive effects in
both samples. This might be attributed to the very nature of the two
samples, which represented two different work groups. However, our
results indicate that within each sample, some moderators seem to be
more crucial than others. In the manufacturing sample, age appeared
to be the most relevant moderator, significantly affecting tumover in-
tention and participation hours in the voluntary organizations and
also had a moderating effect of approaching significance on mental
health and meetings attended in voluntary organizations. This in-
dicates that among industrial rank-and-file workers, as age increases,
the relationships among work schedule and mental health, participa-
tion in voluntary organizations, and tumover intention also change.
Sex was another moderator in the manufacturing sample which mod-
erated the relationship between work schedules and mental health and
anticipated tumover. However, the observed findings that female
workers on rotating shifts had higher mental health and were lower on
anticipated tumover than male workers seemed to be obscured. Since
there were only 8 female workers on rotating shifts, these results
should be interpreted cautiously.
In the hospital sample, the variable "marital status" appeared to be
relevant, moderating the relationship between work schedules and or-
ganizational commitment and absenteeism. The result that married
nurses on fixed shifts showed higher organizational commitment than
546 PERSONNEL PSYCHOLOGY

single nurses was consistent with Becker's (1960) theory of com-

mitment which suggests that sidebets (in this case, married status)
play an important role in enhancing an individual's commitment to an
organization. The finding that single nurses on rotating shifts had a
higher frequency of absenteeism than married nurses may be because
of the nature of the measurement of absenteeism in our study. We as-
sessed only the frequency of absences and did not assess the duration
of absences. It may be true that single nurses get sick more often, but
their duration might not be as lengthy as those of the married nurses.
Only future research could shed more Ught on this issue.
REFERENCE NOTES
Jamal, M., Bamowe, J. T., Mitchell, V. F. Surveying worker mental health: Aspirin for
measurement headache. Paper presented at the meeting of the American Psychologi-
cal Association, San Francisco, 1977.
Porter, L. W. and Smith, F. Etiology of organizational commitment. Unpublished pa-
per. Graduate School of Administration, University of California, Irvine, 1971.

REFERENCES
Aanonsen, A. Shift work and health. Oslo: Universitetsforlaget, 1964.
Baba, V. and Jamal, M. On Becker's theory of commitment: An empirical verification
among blue-collar workers. Industrial Relations (Laval), 1979, 34, 123-139.
Becker, H. C. Notes on the concept of commitment. American Journal of Sociology,
I960, 66, 32-40.
Brayfield, A. H. and Rothe, F. H. An Index of job satisfaction. Journal of Applied Psy-
chology, 1951, 35, 307-311.
Colligan, M. J., Frockt, L J., and Tasto, D. L. Frequency of sickness absence and work-
site clinic visits among nurses as a fimction of shift. Applied Ergonomics, 1979, 10.2,
79-85.
Dunham, R. B. Shift work: A review and theoretical analysis. Academy of Management
Review, 1977, 2, 626-634.
Evans, M. G. A longitudinal analysis of the impact offlexibleworking hours. Studies in
Personnel Psychology, 1976, 6, 140.
Frost, P. J. and Jamal, M. Shift work, attitudes and reported behavior: Some associa-
tion between individual characteristics and hours of work and leisure. Journal of Ap-
plied Psychology, 1979, 64, 77-81.
Hagedom, R. and Labovitz, S. Participation in community associations by occupation:
A test of three theories. American Sociological Review, 1968, 33, 272-283.
Jamal, M. Task specialization and organization and organizational commitment: An
empirical examination among blue-collar workers. Industrial Relations (Laval), 1975,
30, 612-627.
Jamal, M. and Mitchell, V. Work, nonwork and mental health: A model and a test. In-
dustrial Relations (Berkeley), 1980, 19, 88-93. Journal of International Management,
May 1977.
Komhauser, A. Mental Health of the industrial worker: A Detroit study. New York: John
Wiley, 1965.
Kraut, A. L Predicting tumover of employees from measured job attitudes. Organiza-
tional Behavior and Human Performance, 1975, 13, 233-243.
Lenski, G. E. Social participation and status crystallization. American Sociological Re-
view, 1956, 21, 458-464.
 MUHAMMAD JAMAL 547

Mahoney, T. A., Newman, J. M., and Frost, P. J. Workers' perceptions of the four-day
week. Califomia Management Review, 1975, 18, 25-35.
Malaviya, P. and Ganesh, K. Shift work and individual differences in the productivity
of weavers in an Indian textile mill. Journal ofApplied Psychology, 1976,61,774-776.
Mott, P. E., Mann, F. C, McLaughlin, Q. and Warwick, D. P. Shift work: The social
psychological and physical consequences. Ann Arbor, Michigan: The University of
Michigan Press, 1965.
Muchinsky, P. M. Employee absenteeism: A review of the literature. Journal of Voca-
tional Behavior, 1977, 10, 316-340.
Nicholson, N., Jackson, P., and Howes, G. Shift work and absence: An analysis of tem-
poral needs. Journal of Occupational Psychology, 1978, 51, 127-137.
Nord, W. R. and Costigan, R. Worker adjustment to the four-day week: A longitudinal
study. Journal of Applied Psychology, 1973, 58, 60-66.
Porter, L. N. and Steers, R. M. Organizational, work and personal factors in employee
turnover and absenteeism. Psychological Bulletin, 1973, SO, 151-176.
Porter, L. W., Steers, R. M., Mowday, R. T., and Boulian, P. V. Organizational com-
mitment, job satisfaction and tumover, among psychiatric technicians. Journal of Ap-
plied Psychology, 1974, 59, 603-609.
Sergean, R. and Brierley, J. Absence and attendance under non-continuous three-shift
systems of work. Nature, 1968, 219, 536.
Tasto, D. L. and Colligan, M. J. Shift work practices in the United States: DHEW Pub-
lication NO (NIOSH), 77-148. Washington, D.C: U.S. Government Printing Office,
1977.
Taylor, P. J. Shift and day work: A comparison of sickness, absence, illness, lateness
and other absence behavior at an oil refinery from 1962 to 1965. British Journal of
Industrial Medicine, 1961, 24, 93-102.
Taylor, P. 1. and Pocock, S. J. Mortality of shift and day workers. British Journal of In-
dustrial Medicine, 1972, 29, 201-207.
Thiis-Evensen, E. Shift work health. Industrial Medicine, 1958, 27, 493^97.
Wilensky, H. Work, careers, and social integration. Intemational Social Science Jour-
nal, 1960, 12, 543-560.
Wyatt, S. and Marriott, R. Night work and shift changes. British Journal of Industrial
Medicine, 1953, 10, 164-172.