Objective and Subjective Work Monotony: Effects on Job Satisfaction, Psychological Distress, and Absenteeism in Blue-Collar Workers

Abstract

The relation of objective work conditions (work underload, repetitive or varied work) and subjective monotony to job satisfaction, psychological distress, and sickness absence was examined in 1,278 male and female workers. Subjective monotony was moderately related to the objective work conditions. Hierarchical regression analyses showed that the effects on all outcomes were partially mediated by subjective monotony and were also directly related to repetitive work and work underload. Job satisfaction and psychological distress were mainly related to subjective monotony, whereas sickness absence was equally related to the work conditions and subjective monotony. The highest impact was observed for short-cycle repetitive work. Testing sex interactions revealed that sickness absence was related to the work conditions in women but not in men. The findings highlight the significance of noting the actual work conditions in predicting employee outcomes.

Full text

Journal
of
Applied
Psychology
1995,
Vol.
80,
No.
t,
29-42
Copyright
1995
by the
American Psychological Association, Inc.
002I-9010/95/S3.00
Objective
and
Subjective
Work
Monotony:
Effects
on Job
Satisfaction,
Psychological
Distress,
and
Absenteeism
in
Blue-Collar
Workers
Samuel
Melamed,
Irit
Ben-Avi,
Jair
Luz,
and
Manfred
S.
Green
Occupational
Health
and
Rehabilitation
Institute
at
Loewenstein
Hospital
The
relation
of
objective work conditions
(work
underload, repetitive
or
varied work)
and
subjective monotony
to job
satisfaction, psychological distress,
and
sickness absence
was
examined
in
1,278 male
and
female
workers. Subjective monotony
was
moderately
related
to the
objective
work conditions. Hierarchical regression analyses showed that
the
effects
on all
outcomes were partially mediated
by
subjective monotony
and
were also
directly
related
to
repetitive
work
and
work underload.
Job
satisfaction
and
psychological
distress were mainly related
to
subjective monotony, whereas sickness absence
was
equally
related
to the
work conditions
and
subjective monotony.
The
highest impact
was
observed
for
short-cycle repetitive work. Testing
sex
interactions revealed that sickness
absence
was
related
to the
work conditions
in
women
but not in
men.
The findings
high-
light
the
significance
of
noting
the
actual work conditions
in
predicting employee
outcomes.
The
negative impact
of
monotonous work
on
well-
being
has
been widely reviewed (Cox, 1985; Davis,
Shackleton,
&
Parasuraman,
1983;
O'Hanlon,
1981;
R.
P.
Smith,
1981;
Thackray,
1981).
However, several meth-
odological
problems
in the
research cited
may
make
in-
terpretation
of the findings
difficult.
The
definitions
of
the
different
types
of
work monotony
are
inconsistent,
especially
with regard
to
objective work
characteristics.
The
latter
have
been commonly classified into repetitive
work
and
work
underload; both
are
considered
stressful
(Frankenhaeuser&
Gardell,
1976;
Gardell, 1982, 1987).
Most
researchers share
a
similar conceptualization
Samuel
Melamed
and
Irit
Ben-Avi,
Behavioral Medicine
Unit,
Occupational Health
and
Rehabilitation Institute
at
Loewenstein
Hospital, Raanana, Israel; Jair
Luz and
Manfred
S.
Green, Cardiovascular Epidemiology Unit, Occupational
Health
and
Rehabilitation Institute
at
Loewenstein Hospital,
Raanana,
Israel.
This
study
was
supported
by the
Committee
for
Preventive
Action
and
Research
in
Occupational Health,
The
Ministry
of
Labor
and
Social
Affairs,
Jerusalem,
Israel.
A
portion
of
this
article
was
presented
at a
joint American Psychological
Association/National
Institute
of
Occupational
Safety
and
Health
conference,
"Stress
in the
'90s:
A
Changing
Workforce
in
a
Changing Workplace," Washington,
DC,
November
1992.
We
extend
our
sincere thanks
to
Shimon
Dolan
for
invaluable
comments
on
earlier versions
of
this
article.
Correspondence concerning this article should
be ad-
dressed
to
Samuel Melamed, Behavioral Medicine Unit, Occu-
pational
Health
and
Rehabilitation Institute
at
Loewenstein
Hospital,
P.O.
Box 3,
Raanana
43100,
Israel.
of
repetitive work, operationally
defined
as
"work
in
which
discrete sets
of
work activities
are
repeated
in the
same
order....
The
cycle time
for the set of
activities
may
be
measured
and
used
as an
index
of the
repetitiveness
of
the
work"
(Cox, 1985,
p.
86).
Jobs
characterized
by a
short cycle time
and
monotonous motor demands
are of-
ten
termed hectic
and
represent high workload
(Gardell,
1987).
Although
it has
been explicitly hypothesized that
work
stress
is
associated
with degree
of
repetitiveness
(i.e.,
an
inverse
function
of the
work-cycle time), surpris-
ingly,
this hypothesis
has
rarely been tested systematically
using
an
objective cycle-time measure. Some researchers
have
simply asked
the
study participants
to
indicate
the
degree
of
repetitiveness
of
their work
and
have used this
as a
predictor
of
stress outcomes (Cox,
1985).
Studies
in
which
cycle time
was
noted have
focused
largely
on
jobs
with
a
very short cycle time, demonstrating their negative
impact
on
psychological
and
physical health (e.g., Lun-
dberg, Granqvist, Hansson, Magnusson,
&
Wallin,
1989;
Wilkes,
Stammerjohn,
&
Lalich,
1981).
In
contrast
to
repetitive work, work underload
has
not
been clearly conceptualized.
To
illustrate,
Coburn
(1979)
defined
underload
as the
condition
in
which
workers
are
employed
in
jobs
beneath their capacities.
Gardell
(1982,
p.
34)
defined
it as
tasks with
too
narrow
and
one-sided
job
content, lacking stimulus variation,
and
with
no
demands
on
creativity, problem solving,
or
social interaction. Unfortunately,
the
latter definition
contains many elements
in
common with repetitive
work.
At
best,
GardelFs
definition
can be
considered con-
ceptual rather than operational
in
identifying
distinct
un-
derload tasks.
29

30
MELAMED,
BEN-AVI,
LUZ,
AND
GREEN
Industrial
tasks
typically
conceived
as
representing
work
underload
are
those involving vigilance, watch-
keeping,
monitoring, inspection,
or
guarding (see, e.g.,
Curry,
Jex, Levison,
&
Stassen, 1979; Fisher, 1993).
These tasks demand attention
yet
provide little stimula-
tion
in
return; over time they
are
likely
to
cause boredom
in
most people
(Fisher,
1993).
Johansson
(1989)
labeled
these situations
uneventful
monotony. What seems
to be
particularly
stressful
in
them
is the
coupling
of
arousal-
reducing
task characteristics with
the
opposing require-
ment
for
sustained alertness
(Thackray,
1981).
In the
present study, work underload
was
operationally
denned
as
work presenting
no
apparent cycle
and
making
no de-
mands
on
pace, necessitating sustained attention
throughout
the
working period,
and
requiring that
the
worker
be
ready
to
respond
to
certain predetermined
events.
In
this type
of
work,
the
worker
is
largely physi-
cally
passive, except when occasionally required
to
take
action.
Repetitive work
and
work underload pose
different
work
demands. Besides
the
difference
in
physical work-
load, Johansson
(1989)
listed
differences
in
work condi-
tions,
such
as the
unpredictability
of
events
and the
pos-
sibility
of
social interactions. Studies
on the
impact
of
work
underload
on
well-being
are
scarce. Most
of
these
studies
were conducted
in the
laboratory
and
were
de-
signed
mainly
to
disclose
the
impact
of
work conditions
on
the
secretion
of
stress hormones.
A
notable exception
is
the
National Institute
of
Occupational
Safety
and
Health
(NIOSH)-funded
field
study (Caplan,
Cobb,
French,
van
Harrison,
&
Pinneau,
1975)
in
which
2,010
men
from
23
occupations were investigated
for
stress,
strain,
and
health.
In
that study, workers
in
occupations
representing
either repetitive work
or
work underload
had
higher than average perceived-boredom
scores
and
expressed
higher
job
dissatisfaction
and
greater underuti-
lization
of
skills. Some workers also displayed higher lev-
els
of
distress.
A
drawback
of
that study
was
that
infer-
ences regarding work characteristics were based
on
work-
ers' reported occupational titles.
No
indication
was
given
of
the
actual work conditions (e.g.,
the
degree
of
work
repetitiveness,
if
any).
Moreover,
the
lack
of
similar large-
scale studies that also included women prevents
us
from
drawing
firm
conclusions regarding
the
stressfulness
of
work
underload compared with that
of
repetitive work
or
varied blue-collar work.
The
present study aimed
to
expand
this
field of
knowledge.
Another
major methodological shortcoming
in the
research literature
is the
lack
of an
attempt
to
contrast
objective
and
subjective measures
of
work
monotony
when
predicting stress outcomes.
In
many survey-based
studies,
the
independent variable
is
defined
exclusively
by
workers' perceptions
of
their
jobs,
using participants'
de-
scriptions
of
their work
as
monotonous, hectic,
and so
forth,
irrespective
of
objective work characteristics (e.g.,
Alfredsson,
Karasek,
&
Theorell,
1982; Coburn, 1979;
Karasek,
1979). Yet,
as
Kasl
(1984)
convincingly
argued,
When
only
a
subjective measure
of
environmental
exposure
is
included
as a
risk
factor,
we
have
two
primary
problems:
(a) we
cannot
be
sure
to
what objective
exposure
the
subjective
measure
can be
linked
(and
how
strongly),
thus making
it
risky
to
plan ameliorative inter-
vention;
and
(b)
we
cannot
be
sure that
we are
dealing with
a
reaction
to an
exposure
in the first
place,
as
opposed
to an
enduring
characteristic
of the
person that manifests itself
irrespective
of any
level
or
type
of
environmental chal-
lenge,
(p.
325)
Thus,
if
researchers wish
to
draw well-founded conclu-
sions regarding
the
implications
of
negative work condi-
tions
for
well-being, they must also include objective
measures
of the
work environment,
by
applying,
for in-
stance, job-analytic methods
(Aldag,
Barr,
&
Brief,
1981).
Indeed,
a
number
of
studies
focusing
on
objective
work
conditions have demonstrated high physiological
and
psychological stress reactions,
as
well
as a
high prev-
alence
of
ill-health symptoms, among workers engaged
in
various monotonous tasks, such
as
assembly-line work
(paced
or
unpaced),
piecework,
and
monitoring
(e.g.,
see
Caplan
et
al.,
1975; Ferguson, 1973; Johansson, Arons-
son,
&
Lindstrom, 1978; Lundberg
et
al.,
1989). Such
accumulating evidence
has
reinforced
GardelFs
(1982)
conclusion
that certain objective work conditions
may be
directly related
to
psychological
and
physiological ill-
health.
Nevertheless, this does
not
preclude
the
possibil-
ity
that these stress outcomes were mediated
by the
work-
ers' appraisals
of
their work conditions
as
being stressful.
In
fact,
support
for
this assumption
has
come
from
stud-
ies in
which
the
objective work characteristics were con-
trolled
(by
selecting workers
from
specific
job
categories,
such
as
assembly-line
or
piece-rate workers).
In
these
studies, subjective reports
of
work characteristics
did
pre-
dict stress outcomes (Ferguson, 1973; Lundberg
et
al.,
1989;Wilkesetal.,
1981).
The
present study
was
designed
to
address some
of
these methodological issues
by
collecting
data
from
a
large
sample
of
male
and
female blue-collar workers
on
objective
work conditions (objective monotony), per-
ceived
monotony,
and
subjective
and
objective
stress-out-
come measures.
To the
best
of our
knowledge, such
an
endeavor
had not
been attempted before.
The
objective
work
conditions were
(a)
three levels
of
repetitive work,
(b)
work
underload,
and
(c)
varied
work.
The
classifica-
tion
of the
work conditions
was
derived
from
a job
analy-
sis
performed
by an
independent observer. Convergent
validity
was
obtained
by
combining
data
from
three sep-
arate
sources:
the
observer, workers
in a
given job,
and the

OBJECTIVE
AND
SUBJECTIVE WORK MONOTONY
31
workers'
supervisor. Frese
and
Zapf
(1988)
considered
convergent validity
desirable
for
overcoming
a
possible
subjective
bias
in
observers' ratings.
The
subjective out-
come measures were
job
satisfaction
and
psychological
distress, which have been
found
to be
greater among
workers
in
repetitive work
and
work underload
(Caplan
et
al.,
1975).
The
objective
outcome
measure
was ab-
sence
from
work, which
has
been shown
to be
high
among employees
in
monotonous work
(Coburn,
1979;
Cox,
1985;
Ferguson, 1973; Johansson
etal.,
1978).
On the
basis
of the
work-stress literature,
we
exam-
ined three
hypothetical
pathways
for
the
impact
of
objec-
tive
and
subjective monotony
on the
outcome variables.
These pathways
are
depicted
in
Figure
1.
The first
possibility
(A
in
Figure
1)
is
that
the
out-
come variables
are
directly related
to
objectively monoto-
nous work.
This
is
similar
to the
linkage found
in
many
of
the
manipulation studies
and field
studies that have exam-
ined
the
association between
job
characteristics
(and
their
changes)
and
various employee outcomes (Fried
&
Ferris,
1987;
Spector, 1992). Besides testing this possibility,
we
also aimed
to (a)
compare
the
direct
effect
of
repetitive
work
and
work underload
on
psychological distress
and
sickness absence
and
contrast
it
with that
of
varied work
and
(b)
test whether
the
stressfulness
of
repetitive work
is
related
to the
degree
of its
repetitiveness.
The
second
possibility
(B
in
Figure
1)
is
that
the im-
pact
of
work
monotony
is
completely mediated
by the
appraisal
of the
work
as
being monotonous.
To
date,
there
has
been
no
direct evidence
of the
mediatory role
of
subjective monotony.
As
already noted,
the
evidence
linking subjective monotony
to
objective work
conditions
and to
stress symptoms
and
ill-health
has
come
from
different
studies.
The
third possibility
(C
in
Figure
1)
is
partial medi-
ation,
which
implies that part
of the
variance
in the
out-
come variables
is due to
objective work
conditions,
and
part
is
transmitted through
the
workers' appraisal
of
mo-
OBJECTIVE
MONOTONY
OUTCOME
VARIABLES
*
JOB
SATISFACTION
*
PSYCHOLOGICAL
DISTRESS
»
SICKNESS
ABSENCE
« JOB
SATISFACTION
*
PSYCHOLOGICAL
DISTRESS
«
SICKNESS
ABSENCE
C.
*
JOB
SATISFACTION
»
PSYCHOLOGICAL
DISTRESS
»
SICKNESS
ABSENCE
Figure
1.
The
hypothetical pathways
of
impact
for
objective
and
subjective
monotony
on the
outcome variables.
notony
at
work.
A
more detailed explanation
of the
par-
tial-mediation
model
and the
analytical
considerations
for
distinguishing between complete
and
partial media-
tion were given
by
Baron
and
Kenny
(1986)
and
James
and
Brett
(1984).
Finally,
there
is
evidence
of sex
differences
in
subjec-
tive monotony
and
boredom
for
similar
tasks.
Cox
(1985)
found
that women typically appraised these tasks
as
less boring
and
monotonous than
did
men.
On the
other hand, women engaged
in
repetitive work reported
lower
well-being than
men
(Cox, Thirlaway,
&
Cox,
1984).
This
discrepancy suggests
that
the
relationship
among objective work conditions, perceived work monot-
ony,
and
employee outcomes
may be
different
for men
and
women.
So
far, this possibility
has not
been system-
atically
investigated.
The
present study addresses this
is-
sue as
well.
Method
Participants were 1,278 blue-collar workers (885
men and
393
women)
in the
CORDIS (Cardiovascular Occupational
Risk
Factors
Determination
in
Israel)
study.
This
is a
multidis-
ciplinary,
cross-sectional,
and
longitudinal
epidemiological
project aimed
at
identifying
occupational risk factors
for
car-
diovascular
disease.
The
screening
was
offered
free
of
charge
on
company time
and was
accepted
by
about
60% of the
employ-
ees.
The
data
collected
for
each participant covered
a
multitude
of
medical, biochemical, ergonomic, environmental,
and
psy-
chological variables. This study
is
based
on
data collected dur-
ing
Phase
1 of
CORDIS
(1985-1987)
from
a
total
of
4,473
blue-collar workers
in
21
manufacturing
plants.
A
detailed
de-
scription
of the
types
of
plants that were sampled
and the
defi-
nition
of
blue-collar work
are
presented
elsewhere
(Melamed,
Luz,&
Green,
1992).
Exclusions
A
total
of
3,195
workers were excluded
from
this study:
1,334 workers (29.8%) performed work
that
could
not be
clearly
classified into
one of the five
predetermined work types
(short-,
medium-,
and
long-cycle repetitive work, work
un-
derload,
and
varied work), 1,726 workers
did not
complete
the
psychological
questionnaires (mainly because
of
language
difficulties
or
because they were absent
at the
time
of
administration),
25
were pregnant women,
100 men and
women
were outside
the age
range
of
20-64
years,
and
10
work-
ers had
incomplete demographic data.
Compared
with
the
participants
in the final
study sample,
the
excluded
men and
women were older,
and the
excluded
men
were less
educated.
A
higher percentage
of
excluded partici-
pants
of
both sexes were married, engaged
in
physical work,
had
greater
job
seniority,
and
were
of
European origin.
A
descrip-
tion
of the
study sample
is
provided
in
Table
1.
Job
Analysis
Procedure
Following
a
pilot study (described below),
job
analyses
were
performed
by an
experienced observer
who
observed

32
MELAMED,
BEN-AVI,
LUZ,
AND
GREEN
Table
1
Description
of
the
Study Sample
and
Excluded
Participants
Study
sample
Excluded
participants
Characteristic
Men
Women
Men
Women
«
Age
(mean years)
Education
(%)
<8
years
9-
1
2
years
>
1
3
years
Married
(%)
Physical
work
(%)
Job
seniority
(mean
years)
Ethnic
origin
(%)
Asia
Africa
Europe
Yemen
Israel
(Jews)
Israel
(Arabs)
885
37.20
59.32
32.54
8.14
81.24
10.96
8.90
23.73
25.76
35.80
5.99
0.68
9.02
393
33.50
62.34
33.08
4.58
67.94
4.07
5.90
31.04
32.32
36.64
0.00
0.00
0.00
2,489
44.70
75.61
17.96
6.43
89.99
13.99
10.85
20.01
28.49
36.84
5.66
0.48
8.51
706
41.10
68.70
14.59
16.71
74.36
5.81
7.48
20.40
32.29
43.90
0.99
0.14
2.28
workers
in the
same
job for 1
day.
A
total
of 480
different
jobs
were
identified
in the 21
plants sampled; these jobs were classi-
fied
into
one of the five
work categories listed below. Jobs that
could
not be
distinctly
classified
into
any of
these types were
labeled
other.
The
convergent validity
of the
observer
classifi-
cation
was
obtained
by
asking workers
and
supervisors
to
clas-
sify
the
same work.
Work
Categories
Repetitive
work
(3
categories) referred
to
work
with
an
identifiable
work cycle
of no
longer than
1 hr.
Although some
workers
occasionally rotated between tasks
or
workstations,
they
retained
essentially
the
same
work-cycle
time.
Work cycles
were
classified
as (a)
short
(less
than
1
min),
(b)
medium
(1-
30
min),
or (c)
long
(30 min to 1
hr). Work underload (see
definition
in the
introduction) included monitoring, watch-
keeping,
sorting,
and
guarding. Varied work included
jobs
with
no
predetermined order
of
activities, entailing task rotation
and
decisions regarding
the
nature
and
order
of
work activities.
Workers
were
classified
into
the
different
work categories
as
follows:
100 men and 140
women
in
short-cycle repetitive
work,
248 men and 83
women
in
medium-cycle repetitive work,
147
men and 43
women
in
long-cycle repetitive work,
91
men
and
44
women
in
work underload,
and 299 men and 83
women
in
varied work.
classifications
made
on
separate days
for the
same
job
were
practically
the
same,
in the
present study
the
observations
for a
given
job
were made
on the
same day.
Psychological
Measures
Subjective
monotony. Subjective monotony
was
assessed
on
the
basis
of
workers' descriptions
of
their work, using
a
short
adjective
checklist. This approach
was
adapted
from
Cox
(1985),
who
reported
its
wide usage
in
psychological research.
Such
job
perceptions have been shown
to
predict stress out-
comes
in
several studies (e.g., Alfredsson
et
al.,
1982; Cox,
1985; Karasek, 1979).
The
four adjectives
used
in our
scale
were
routine, boring, monotonous,
and not
varied enough.
Workers
indicated which adjectives described their work
by
rat-
ing
each adjective
as yes = 3, ? =
1,
or no = 0. The
score ranges,
means,
standard
deviations
and
Cronbach
alphas
for
this
and
other scales
are
given later
in
Table
3.
Job
satisfaction.
Job
satisfaction
was
measured
by the
Satisfaction
With Work
facet
of the Job
Description Index
(JDI)
ofP.
C.
Smith, Kendall,
and
Hulin
(1969).
Thissubscale
had the
highest correlation with other well-being measures
(Meir
&
Melamed, 1986)
and
with
withdrawal behavior
(Muchinsky,
1977).
Aldag
et al.
(1981)
argued that
in
many
studies,
inflated
correlations between perceived
job
characteris-
tics
and the JDI
resulted
from
domain overlap
and
suggested
that redundant items
in the
corresponding scales
be
deleted.
Following
this suggestion,
we
omitted
the
routine
and
boring
items
from
the
JDI.
An
additional item, hot,
was
also omitted
following
item analysis. Thus,
the final
scale
consisted
of 15
items,
with
a
response scale
of
yes = 3, ? =
1,
or no = 0.
Psychological
distress. Psychological distress
was
gauged
by
using
the
symptoms
of
somatic complaints, anxiety, irrita-
bility,
and
depression
as
measured
by the
scales
described
below.
As
these symptoms were
found
here
to be
highly
intercorrelated
(mean
r =
.51),
the
unweighted
sum of all of
their respective
scores
was
used
as a
composite measure
of
psychological
distress.'
Somatic complaints were assessed
by 10
items
on the
basis
of
the
index reported
by
Caplan
et al.
(1975).
This index mea-
sured
the
frequency
with
which certain symptoms
(e.g.,
dizzi-
ness, shortness
of
breath, headaches,
and
clammy
hands)
were
experienced
during
the
month prior
to
participation
in the
study.
The
response scale ranged
from
never
(1)
to
very
often
(4).
Anxiety
was
assessed
with
a
scale
of five
items based
on
Zung's
(1971)
Self-Rating Anxiety Scale
(SAS);
the
items mea-
sured
the
frequency
with
which certain symptoms were experi-
enced during
the
past week. Responses were
on a
4-point scale
ranging
from
none
or
seldom
(1)
to
most
or
always
(4).
Pilot
Study
and
Interobserver
Agreement
The
reliability
of the
work classification
was
tested
by ap-
plying
it to 48
jobs
in two
plants
from different
industries.
Clas-
sifications
were made
by
three independent observers
who
ob-
served
workers
in the
same
job on
different
occasions,
often
on
separate days. Interobserver agreement, assessed
by the
Kappa
statistic
(Cohen, 1960), ranged
from
.88 to
.93. Because
the
1
We
have previously analyzed
the
psychological distress
symptoms separately
(Kushnir
&
Melamed,
1991),
and all of
the
results
were
essentially
in the
same direction. Hence,
as no
new
information
was
obtained
by
considering these symptoms
separately,
we
combined them into
one
aggregate measure. This
procedure follows common practice
in
many studies
of
occupa-
tional stress (e.g., Adams,
1981).

OBJECTIVE
AND
SUBJECTIVE
WORK
MONOTONY
33
Irritability
was
measured
by two
items (Kushnir
&
Mel-
amed,
1991)
scored
on a
5-point scale ranging
from
not
true
at
all
(I)
to
very
true
(5).
The two
items were:
"At
the end of the
workday,
are you
irritable, impatient, want
to be
left
alone,
or
very
much bothered
by
noises around
you
(e.g.,
TV)?"
and
"When
you
come home
from
work,
does
it
take
you a
long
time
to
relax
and do the
things
you
like
or
have
to
do?"
Depression
was
measured
by
using
five
items adapted
from
Zung's (1965) Self-Rating Depression Scale (SDS).
The
fol-
lowing
depression symptoms during
the
month prior
to
testing
were
measured:
fatigue,
loss
of
appetite, depressed mood
and
sadness, sleep disturbances,
and
loss
of
interest
in
work.
The
response scale ranged
from
never
(1)
to
very
often
(4).
Sickness-absence
data. Certified sickness-absence
data
for
the
years 1985-1987 (when
the
medical
and
psychological
data were collected), including
1-day
absences, were compiled
from
original medical sick-leave certificates
or
from
the
insur-
ing
institute's
database.
Twenty-four
months represented
sufficient
time
to
reduce skewness
and
kurtosis (Hammer
&
Landau,
1981).
Indexes
for
spells
(frequency)
and for
number
of
workdays lost were computed.
One
hundred
or
more work-
days
lost
was
coded
as
100. However, despite
the
long period
sampled,
the
absence data remained considerably skewed.
The
skewness
and
kurtosis levels
for men and
women
are
presented
in
Table
2.
These exceeded
the
accepted levels
for
skewness
(not
approaching
2)
and
kurtosis
(not
greater than
5;
Kendall
&
Stu-
art,
1958).
Hence,
we
applied
a log
transformation
to the
data,
thereby
reducing
the
skewness
and
kurtosis
to
acceptable levels
for
analysis
(see
Table
2).
After
the
transformation,
the
spells
and
workdays-lost measures were highly correlated
in
both
men
(r=
.78)
and
women
(r =
.89). Given that spells were
found
to
be
a
more reliable (stable) measure than workdays lost
(Hammer
&
Landau,
1981;
Muchinsky, 1977),
in the
subse-
quent
analyses
only
log
spells were used
as the
sickness-absence
measure.
Potential
Confounding
Variables
It
is
acknowledged that
the
self-selection
or
forced selection
of
work conditions
may be
affected
by
age, sex, education,
and
ethnic
origin (Wallace, Levens,
&
Singer,
1988).
These factors
may
also
influence
the
impact
of
these work conditions
(Jex
&
Beehr,
1991;
R. P.
Smith,
1981)
and the
outcome variables
measured
in
this study, such
as job
satisfaction, emotional-dis-
tress symptoms,
and
sickness absence
(Jex
&
Beehr,
1991;
Wal-
lace
et
al.,
1988). Therefore,
we
also collected demographic
data
on
age, education,
and
ethnic
(country)
origin. Education
was
coded
into three levels,
and
ethnic origin
was
coded
into
six
categories
(see
Table
1).
Ethnic origin
was
defined
according
to
the
worker's birthplace. Israeli-born workers were classified
as
such only
if
their fathers
had
also
been born
in
Israel.
Age
(continuous
variable),
education,
and
ethnic
origin
(coded
as
dummy
variables) were controlled
for in the
subsequent
analyses.
Results
Objective
and
Subjective
Monotony
The
significance
of the
differences
in
mean
subjec-
tive-monotony
scores
of men and
women
in
various
work
conditions
was
tested
through
two-way
(Work
Conditions
X
Sex)
analyses
of
variance.
Significance
was
defined
as
p
<
.05.
The
results
indicated
a
significant
main
effect
for
work
conditions,
F(4,
1268)
=
24.50,
p <
.005,
but not
for
sex
differences,
F(4,
1268)
=
1.54,
ns,
or for the
Work
Conditions
X Sex
interaction,
F(4,
1268)
=
1.41,
ns.
Thus,
both
men and
women
in
objective-monotony
work
conditions
equally
perceived
them
as
monotonous.
The
pooled
mean
subjective-monotony
scores
by
work
conditions
were
7.31
for
short-cycle
repetitive
work,
6.50
for
medium-cycle
repetitive
work,
5.76
for
long-cycle
re-
petitive
work,
7.07
for
work
underload,
and
4.49
for
var-
ied
work.
These
scores
indicate
that
for
workers
engaged
in
repetitive
work,
subjective
monotony
was
positively
re-
lated
to
repetitiveness
(an
inverse
function
of
cycle
duration).
Post
hoc
contrasts
revealed
no
significant
difference
between
workers
in
short-cycle
repetitive
work
or
work
underload;
all
workers
equally
reported
high
mo-
notony.
The
lowest
subjective
monotony
was
reported
for
varied
work
(when
compared
with
the
other
means,
all
probabilities
were
significant
at
the
p < .05
level,
at
least).
To
assess
the
degree
of
association
between
the
vari-
ous
types
of
objective
and
subjective
monotony,
we
formed
a set of
dichotomous
variables
(used
as
dummy
variables
in the
subsequent
regression
analyses)
that
con-
trasted
each
type
of
objective
monotony
against
varied
work.
Pearson
correlations
between
these
variables
and
subjective
monotony
are
included
in the
intercorrelation
Table
2
Skewness
and
Kurtosis
for
Sickness-Absence Data
Untransformed
data
Skewness
Kurtosis
Transformed
data
Skewness
Note.
M =
men,
W =
women.
Kurtosis
Absence
measure
Spells
Workdays lost
M
2.77
2.59
W
1.79
2.03
M
11.81
6.89
W
2.22
4.54
M
1.18
0.62
W
0.69
0.35
M
0.20
0.00
W
-0.90
-0.01

34
MELAMED,
BEN-AVI,
LUZ,
AND
GREEN
matrix
for all of the
study variables, which
are
presented
separately
for men and
women
in
Table
3. The
highest
correlations were obtained
for
women
in
short-cycle
re-
petitive work
and
work underload
(rs
= .42 and
.34,
respectively).
In
men,
three
out of the
four
correlations
were
about
.30.
These
findings
indicate that objective
and
subjective
monotony
are
moderately related.
Relationship
of
Work
Monotony
to
Outcome
Measures
The
correlations
in
Table
3
also show that
the
sub-
jective
outcome variables
(job
satisfaction
and
psycho-
logical
distress)
were related
in the
expected direction
to
both objective
and
subjective monotony
in
both sexes.
These correlations were,
not
unexpectedly, much higher
with
subjective monotony than with objective monotony.
The
highest
correlations
were obtained with
job
satis-
faction
(rs
=
—.52
and
—.58
in men and
women,
respectively),
despite
the
elimination
of
item overlap
in
the
subjective-monotony
and
job-satisfaction scales
(see
the
Method
section).
Lower correlations were obtained
with
psychological distress
(rs =
.21
and .35 in men and
women,
respectively).
These
findings are
similar
to
those
of
French, Caplan,
and van
Harrison
(1982)
in the
NI-
OSH
study. French
et
al.
suggested
that
whereas
job
sat-
isfaction
is a
job-related
effect,
distress symptoms
may be
caused
by
other factors
not
necessarily related
to
work.
Job
satisfaction
was
also negatively related
to
objec-
tive
work conditions. With
the
exclusion
of
long-cycle
re-
petitive
work
in
women,
all
types
of
work monotony,
when
compared with varied work, were associated with
lower
job
satisfaction.
On the
other hand, only
the ex-
treme conditions
of
short-cycle repetitive work
and
work
underload
(in
women) were positively associated with
psychological
distress.
In
contrast, sickness absence
(an
objective
outcome)
was
more closely related
to
objective monotony
in
women
only.
Interestingly,
the
magnitude
of the
corre-
lations
obtained with short-cycle repetitive work
(r =
.45),
medium-cycle repetitive work
(r =
.35),
and
work
underload
(r
=
.56)
far
exceeded
the
correlations usually
obtained
for
perceived
job
characteristics
(see,
e.g., Fried
&
Ferris,
1987;
Spector
&
Jex,
1991).
No
significant
cor-
relations between objective monotony
and
sickness
ab-
sence were
obtained
in
men.
Examination
of the
other
correlations presented
in
Table
3
indicates that sickness
absence
was
moderately
and
positively correlated
in
both
sexes
with subjective monotony
and
psychological
dis-
tress
and was
negatively correlated with
job
satisfaction.
Testing
the
Hypothetical Pathways
of
Impact
on the
Outcome
Variables
To
determine which
of the
possible pathways
of im-
pact depicted
in
Figure
1 was
supported
for
each
of the
t5i
-0*
-Si
-d
'£*
;§
P.
|
1
«j
Intercorrelations
%
<u
1%
O
_Q
C/J
^
—
0
§
</)
<u
o
*^
o
'•n
>>
cC
c
_o
lj
i
•-I
^
°S
2
O
~
&
-o
c
3
V
C
O
00
(ii
o
g-3
O
"J
cj
E
•5
g
/-i
T3
O
^
ll
>
c"
'•£
O
<U
O
co
£
S
J&
"^
c
OH
*
CO
s
^
§
2
Variable
*
*
OO
(N
*J
O O C
1
*
* *
—
•
Tf
W
(N
—
C
vi
(N r
\
\
*
(N
*
#
*
O
*
*
(N
1
#
*O
(N
OC
tN
^
|
ri
>n
v
m
o c
r-
rfr
v
m
0 C
(N
^O
V
^0
O C
OS
<"*")
V
V)
O C
3
^
e"
**
•*
Subjective
monotoi
Short-cycle
RW vs.
\AoA{,,m
mrffV)\\f
*
D
r-
r-
:>
o c
l
*
#
"i
m
o
2
O C
1
q
— r
1
1
*
3
ON
T
1
~|
i/1
>/•
>
0 C
1
l/n
"1
>
o c
~i
r*^
r
>
O C
}
f^l
f^
>
o c
^3-'^
*
>•
;
i
"
i^l
^1
ft!
ill
#
#
* *
3
=
^
1
1
* *
* *
°
S
oq
—
-
VI
O O
-i
o\
^t
rs
1
1
* *
* * *
(N —
\O
•<±
(N
vi
j
#
oo
r--
oo
q q
—
r * r
#
* * *
(N —
f")
|
* #
#
* * *
oo
O
v->
rn
r^
rf
r
* * *
r
oo
V)
\o
I
V)
(^1
—
1
v-i
o
™!
1
I
1
)
O
<N
O
(N
1
Os
f-
—
>
0 0
—
ON
TJ-
00
>
O
ON
O
Tt
0 0
j
\O
r*^
—
^
O
oo
m
>
r-
r-
o
(N
rn
Job
satisfaction
Psychological
distre
Sickness absence
ii
S
g
4)
O
TD
2
c
<u
u-
10
3
s
C
§
to
u
3
a1
5
Jj>
J2
c
'o
£
u
o
u
g
,™
JD
O
1
1
oo
-a
a
^
'^
a
<u
y
•£
f.
0
'^
1
"S
«
I"
*
1
P
^
O
•"
>
'
,o
-^>
<~
o
c
"S
—
JJ
||
|
|||
_-'C
j
§
r
s
&0
l~.
u
^
'"6
•S
^'
p^
S
i
§
°
O
ra
._g
O
u.
'.£
C
S
E'S
CO
C
||
^
O
>
-Q
V"
^§<2c
M»(?.
Correlation
men,
W =
women,
*
In
each
category
o
*
n
^
nS
**
n
^

OBJECTIVE
AND
SUBJECTIVE
WORK
MONOTONY
35
outcome variables,
we
carried
out
three hierarchical
re-
gressions
for men and
women combined. Step
0 in all of
the
regressions included
the
potential confounding vari-
ables
(this
step
is
discussed
later).
Step
1
included
the
confounding
variables
and
subjective monotony (con-
tinuous
variable)
in
Regression
1,
objective-monotony
conditions (taken
as
dummy variables,
as
described
above)
in
Regression
2, and
subjective
and
objective
mo-
notony
in
Regression
3.
Regressions
1 and 2
separately
examined
the
association
of
each
predictor
with
the
out-
come variables. Regression
3
examined which
of
these
associations remained
significant
when
the
main
effects
were
simultaneously controlled.
In
Step
2, the
significant
(at p
<
.05)
sex
interactions with main
effects
were
entered.
The
results obtained
in
Step
1 of the
analyses
are
considered
first.
Results
for the
subjective outcomes
are
given
in
Table
4. In
Regression
1,
job
satisfaction
was
sig-
nificantly
and
negatively related
to
subjective monotony.
In
Regression
2, job
satisfaction
was
negatively
related
to
all
conditions
of
objective
monotony.
The
job-satisfac-
tion means (adjusted
for the
various confounders)
for
workers
in the
short-,
medium-,
and
long-cycle repetitive-
work
levels,
work
underload,
and
varied work were 25.5,
28.5,
30.6, 27.5,
and
33.9, respectively.
For
repetitive
work,
the
beta
coefficients
decreased
in
size, moving
from
short-,
through
medium-,
to
long-cycle duration. Given
that
the
ranges
of all
three independent variables were
equal,
this
finding
implies that
the
degree
of
job
dissatis-
faction
was
related
to the
degree
of
work
repetitiveness.
When
both objective
and
subjective monotony were
included
in
Regression
3,
they still remained
significantly
and
negatively associated with
job
satisfaction. However,
the
inclusion
of
subjective monotony
in the
regression
equation
reduced
the
size
and
significance
of the
coeffi-
cients
obtained
for
objective monotony
in
Regression
2.
Almost
no
change
(from
Regression
1)
in
either size
or
significance
of the
beta
coefficient
was
observed
for
sub-
jective
monotony. Thus, these results
satisfy
the
require-
ments
for the
partial-mediation
model
(see Baron
&
Kenny,
1986),
namely, pathway
C in
Figure
1.
Of the
objective-monotony
conditions, short-cycle repetitive
work
had the
highest negative impact
on
this outcome.
No
impact
was
found
for
long-cycle repetitive work.
A
comparison
of the
squared multiple correlation (ad-
justed
for
shrinkage)
in all
three regressions showed that
most
of the
explained variance
in job
satisfaction
was due
to
subjective
monotony.
The
results
in
Table
4
reveal that psychological dis-
tress
was
also
positively
related
to
subjective monotony
(in
Regression
1)
and to
repetitive work
and
work
un-
derload
(in
Regression
2).
However,
of the
three repeti-
tive-work
levels,
only
short-cycle repetitive work
was
sig-
nificantly
associated with this outcome.
The
adjusted
mean psychological-distress scores
for the
short-,
medi-
um-,
and
long-cycle repetitive-work levels, work
un-
derload,
and
varied work were
42.5,39.1,
38.9,40.3,
and
38.1,
respectively. Both objective
and
subjective monot-
ony
were related
to
psychological distress
in
Regression
3.
Thus,
the
partial-mediation model
was
also supported
for
this outcome.
An
inspection
of the
adjusted squared
multiple
correlation values
in all
three regressions again
indicated
that
the
contribution
of
subjective monotony
to
this
subjective outcome
was
much higher
than
that
of
objective
work conditions.
The
results
for
sickness absence
are
given
in
Table
5.
Sickness
absence
was
significantly
related
to
subjective
monotony,
repetitive work,
and
work underload. How-
ever,
of the
three repetitive-work levels, only
the
short
cy-
cle was
significantly
related
to
this outcome
in
Regression
2. The
adjusted mean sickness-absence scores
for the
short-,
medium-,
and
long-cycle repetitive-work levels,
work
underload,
and
varied work were 0.86, 0.66, 0.55,
0.81,
and
0.52, respectively.
All of the
significant
results
obtained
in
Regressions
1 and 2
were retained
in
Regres-
sion
3.
Again, support
was
obtained
for the
partial-medi-
ation model. Furthermore,
the
adjusted squared multiple
correlations obtained
in
Regressions
1 and 2
were
very
similar,
indicating that sickness absence
was
related
equally
to
subjective
and
objective monotony.
Testing
Possible
Gender
Differences
The
possibility
of
gender
differences
in the
pattern
of
association between objective
and
subjective
monotony
and the
outcome measures
was
tested
by
examining
in-
teractions between main
effects
and
sex, entered
in
Step
2
in all the
regressions, while retaining
all the
effects
in
Step
1.
Significant
interactions
are
given
in
Tables
4
and 5.
As
Table
4
shows,
a
significant
Subjective Monotony
X
Sex
interaction
was
found
for job
satisfaction
in Re-
gression
1 (p <
.05);
this interaction
was
also retained
in
Regression
3 (p <
.005).
No
interaction
with objective
monotony
was
found.
To
help describe
the
nature
of the
interaction
effect,
we
separately plotted
the
regression
lines
for men and
women
(see
top
left
of
Figure
2).
These
graphs
indicate that
high
levels
of
perceived monotony
were
associated with
low job
satisfaction
in
women more
than
in men and
that
low
levels
of
perceived monotony
were
associated
with
high
job
satisfaction
in men
more
than
in
women. Yet, this interaction contributed very lit-
tle to the
multiple squared correlation
(AT?2
=
.002),
even
though
it was
statistically
significant
(p <
.05).
Significant
results were also obtained
for
psycholog-
ical
distress
(see
Table
4).
A
Subjective Monotony
X Sex
interaction
was
obtained
in
Regression
1,
and a
Short-
Cycle
Repetitive Work
X Sex
interaction
was
obtained

36
MELAMED,
BEN-AVI,
LUZ,
AND
GREEN
Table
4
Results
of
Hierarchical
Regression Analyses
on Job
Satisfaction
and
Psychological
Distress
by
Objective
Monotony,
Subjective
Monotony,
and
Possible
Confounding
Variables
Unstandardized
regression
coefficient
(E)
Variable
entered
Age
Sex"
Education
<;8
years"
9-
12
years'"
Ethnic
origin
Asiac
Africa'
Subjective
monotony
Subjective
Monotony
X
Sexd
dfs
F
Adjusted
R2
Age
Sex"
Education
<8
years'
9-
12
years"
Ethnic
origin
Asiac
Africa0
Short-cycle
RW
Medium-cycle
RWe
Long-cycle
RWe
Work
underload"
Short-Cycle
RWe
X Sex
rf/s
F
Adjusted
R2
StepO
0.05
-0.95
-3.26*
-3.59**
0.85
0.93
6,1271
2.35*
0.0063
0.05
-0.95
-3.26*
-3.59**
0.85
0.93
6,
1271
2.35*
0.0063
Job
satisfaction
Step
1
Regression
0.01
-0.77
-0.74
-1.65
1.53*
1.97**
-1.67**
7,
1270
77.58**
0.2957
Regression
0.05
0.43
-2.71*
-3.55**
0.54
0.96
-7.56**
-4.56**
-3.07**
-6.16**
10,
1267
9.82**
0.0646
Psychological distress
Step
2
1
0.01
1.38
-0.70
-1.62
1.48*
2.04**
-1.22*
-0.35*
8,
1269
68.67**
0.2977
2
f
StepO
0.08**
5.33**
4.24**
3.16*
0.09
0.97
6,1271
17.92**
0.0739
0.08**
5.33**
4.24**
3.16*
0.09
0.97
6,1271
1
7.92**
0.0739
Step
1
0.10**
5.25**
3.20**
2.35*
-0.19
0.54
0.69**
7,
1270
28.57**
0.1313
0.08**
4.32**
3.91**
3.10*
0.33
0.95
4.67**
1.10
1.03
2.22*
10,
1267
14.38**
0.0948
Step
2
0.10**
1.87
3.14**
2.31*
-0.11
0.44
-0.01
0.56**
8,
1269
26.58**
0.1381
0.08**
3.52**
3.83**
3.03*
0.50
0.96
-0.43
1.12
1.03
2.31*
3.42*
11,1266
13.62**
0.0981
Regression
3
Age
Sex"
Education
<;8
years"
9-
12
years"
Ethnic
origin
Asia0
Africa0
Short-cycle
RWe
Medium-cycle
RWe
Long-cycle
RWe
Work
underload'
Subjective
monotony
Subjective
Monotony
X
Sexd
dfs
F
Adjusted./?2
0.05
-0.95
-3.26*
-3.59**
0.85
0.93
6,1271
2.35*
0.0063
0.01
-0.20
-0.65
-1.72
1.36*
1.94**
-2.99**
-1.42*
-0.97
-1.99*
-1.58**
11,1266
51.06**
0.3013
0.01
1.73
-0.62
-1.70
1.32*
2.00**
-2.87**
-1.51*
-1.03
-1.99*
-1.17**
-0.32*
12,
1265
47.24**
0.3029
0.08**
5.33*
4.24**
3.16*
0.09
0.97
6,
1271
17.92**
0.0739
0.10**
4.58**
3.08**
2.37*
-0.01
0.55
2.83**
-0.17
0.18
0.54
0.64**
11,
1266
19.74**
0.1390
0.10**
1.67
3.03**
2.34*
0.06
0.46
2.65**
-0.03
0.27
0.54
0.02
0.49**
12,
1265
18.90**
0.1440
Note.
RW =
repetitive work.
• Men =
1,
women
= 2.
b
Compared
with
SL
13
years.
c
Compared
with Europe.
e
Compared
with
varied work.
f
No
interaction
was
found.
*/><.05.
**p<.005.
1
Significant interaction term added
to
effects
entered
in
Step
1.

OBJECTIVE
AND
SUBJECTIVE WORK MONOTONY
37
Table
5
Results
of
Hierarchical
Regression Analyses
on
Sickness Absence
by
Objective
Monotony,
Subjective
Monotony,
and
Possible
Confounding
Variables
Unstandardized
regression
coefficient
(B):
Sickness absence
Variable
entered
Age
Sex"
Education
<8
years"
9-12yearsb
Ethnic
origin
Asiac
Africa'
Subjective
monotony
Subjective
Monotony
X5exd
dfe
F
Adjusted
R2
Age
Sex"
Education
<8
years"
9-
12
years"
Ethnic
origin
Asia0
Africa'
Short-cycle
RW'
Medium-cycle
RW'
Long-cycle
RWe
Work
underload'
Long-Cycle
RWe
X
Sex"
Work
Underload'
X
Sex"
Short-Cycle
RW'
x
Sex"
Medium-Cycle
RW'
X
Sex"
dfs
F
Adjusted
R2
Age
Sex"
Education
<8
years"
9-
12
years"
Ethnic
origin
Asiac
Africa0
Short-Cycle
RW'
Medium-cycle
RWe
Long-cycle
RW'
Work
underload'
Subjective
monotony
Long-Cycle
RW'
Work
Underload'
X
Sex"
Short-Cycle
RW' X
Sexd
Medium-Cycle
RW'
X
Sexd
dfs
F
Adjusted
R2
StepO
0.00
0.46**
0.00
0.01
-0.16*
-0.04
6,1271
14.18**
0.0583
0.01*
0.46**
0.00
0.01
-0.16*
-0.04
6,1271
14.18**
0.0583
0.01*
0.46**
0.00
0.01
-0.16*
-0.04
6,
1271
14.81**
0.0583
Step
1
Regression
0.01*
0.45**
-0.04
-0.02
-0.17*
-0.06
0.03**
7,
1270
14.83**
0.0705
Regression
0.01
0.38**
-0.02
0.01
-0.14*
-0.05
0.34**
0.09
0.06
0.31**
10,
1267
11.56**
0.076
Regression
0.01*
0.39**
-0.04
-0.02
-0.15*
-0.06
0.28**
0.05
0.03
0.25**
0.02**
11,1266
11.34**
0.0818
Step
2
1
0.01*
0.31*
-0.04
-0.02
-0.17*
0.07
-0.01
0.03*
8,
1269
13.65**
0.0734
2
0.00
0.51**
-0.00
0.00
-0.13*
-0.05
0.29**
0.09
1.30**
0.29**
-1.02**
11,1266
14.59**
0.1048
3
0.01*
0.52**
-0.03
-0.02
-0.14*
-0.06
0.24**
0.05
1.26**
0.24**
0.02*
-1.00**
12,
1265
14.09**
0.1095
Step
3
0.00
0.41**
-0.00
0.00
-0.13*
-0.06
0.33**
0.09
1.18**
-0.64*
-0.92**
0.71**
12,
1265
15.07**
0.1168
0.00
0.42**
-0.03
-0.02
-0.14*
-0.07
0.27**
0.05
1.14**
-0.68**
0.02*
-0.91**
0.71**
13,
1264
14.56**
0.1213
Step
4
0.00
0.23**
-0.02
-0.01
-0.10
-0.06
-0.54*
0.10
0.95**
-0.87**
-0.73**
0.90**
0.59**
13,
1264
15.65**
0.1298
0.00
0.24**
-0.04
-0.03
-0.11
-0.07
-0.58*
0.06
0.92*
-0.91*
0.02*
-0.73*
0.89*
0.58*
14,
1263
15.08**
0.1338
StepS
0.00
-0.15
-0.03
-0.02
-0.09
-0.05
-1.00**
-0.85**
0.49*
-1.33**
-0.36*
1.28**
0.97**
0.77**
14,
1263
16.80**
0.1476
0.00
-0.15
-0.05
-0.04
-0.10
-0.07
-1.05**
-0.92**
0.44*
-1.38**
0.02**
-0.34*
1.28**
0.97**
0.79**
15,
1262
16.32**
0.1525
Note.
RW =
repetitive work.
"
Men
=
1,
women
= 2.
b
Compared
with
;»
13
years.
c
Compared
with Europe.
"
Significant
interaction term
added
to
effects
in the
previous
step.
'
Compared
with
varied work.
*p<.05.
**p<.005.

38
45
=
35
J>
(0
f
$'
15
MELAMED,
BEN-AVI,
LUZ,
AND
GREEN
Job
Satisfaction
Psychological
Distress
45
=
35-
Men
Women
12
Subjective
Monotony
25
f
Women
Men
12
Subjective
Monotony
Sickness
AbsenceSickness
Absence
2.4
.
"1.2
0.6-
0.0
Women
Men
Varied
Work
R.W.-Short
Cycle
Objective
Monotony
2.4
"1.2
ra
3
0.6
0.0
Women
Men
Varied
Work
Work
Underload
Objective
Monotony
Figure
2.
Interactions between
sex and the
independent variables
in the
prediction
of
employee
outcomes. R.W.
=
repetitive work.
in
Regression
2.
Only
the
former
remained
significant
in
Regression
3.
This interaction appears
on the top
right
of
Figure
2. It
shows
a
closer relationship between perceived
work
monotony
and
resultant psychological distress
in
women
than
in
men. Women also
had
higher psychologi-
cal-distress levels than men. This
finding is
consistent
with
similar
findings
reported
by Cox et
al.
(1984).
How-
ever,
adding this interaction
to the
equation
in
Regression
3
again yielded
a
significant
(p <
.05)
but
small change
in
the
squared multiple correlation
(AJ?2
=
.005).
An
interesting
and
different
sex-dependent
influence
of
the
predictor variable
was
observed
for
sickness
ab-
sence (see Table
5).
In
Regressions
1 and 2, sex in-
teracted
with
both objective
and
subjective monotony.
However,
in
Regression
3, sex
interacted only with objec-
tive
work
conditions.
Significant
interactions
were
ob-
tained with short-
and
medium-cycle repetitive work
and
with
work underload.
Two of
these interactions
are
shown
at
the
bottom
of
Figure
2,
which
clearly shows
that
in
men,
objective work conditions were
not
associated with
sickness
absence. However,
in
women, engagement
in
short-cycle
repetitive work
or
work underload
(as
com-
pared with varied work)
was
positively associated with
sickness
absence.
The
same trend
was
obtained
for me-
dium-cycle
repetitive work
(not
shown).
The
inclusion
of
the
interaction terms
in
Regression
3
appreciatively
increased
the
squared multiple correlation,
(A/?2
=
.068),-F(3,
1263
) =
35.33,
p<.
005.
Assessing
the
Contribution
of
the
Confounders
to
the
Outcomes Studied
To
test whether
the
potential confounding variables
indeed contributed
to the
variance
of the
outcomes stud-
ied,
we
entered them into
the
regression equation
in a
preliminary run, presented
as
Step
0 in
Tables
4 and
5.
The
results
of
Regression
2
indicate that
the
confounders
explained less
than
1%
of the
variance
in job
satisfaction.
On the
other hand, they explained nearly
8% of the
vari-
ance
in
psychological distress
and 6% of the
variance
in
sickness absence. Thus, over
53% of the
total explained
variance
in
psychological distress
and
nearly
40% of the
total explained variance
in
sickness absence were
ac-
counted
for by
variables unrelated
to
work.
Discussion
The
common practice
in
large studies
of the
effects
of
work monotony
and
other negative
job
characteristics
on
employee outcomes
has
been
to
infer
their existence
from
incumbents'job
titles.
In
this study
of a
large cohort
of
male
and
female blue-collar workers, work conditions
were
objectively assessed through
a job
analysis con-

OBJECTIVE
AND
SUBJECTIVE WORK MONOTONY
39
ducted
by an
experienced observer. This procedure
enabled
us to
identify
workers
in
distinct types
of
monotonous
work—repetitive
work
and
work underload
(objective
monotony)—and
a
control group
in
varied
work. Workers were asked
to
indicate
the
extent
to
which
they
perceived their work
as
monotonous (subjective
monotony).
This design permitted
the
testing
of the
pos-
sible pathways
of
impact
of
objective
and
subjective
mo-
notony
on
employee objective
and
subjective outcomes.
Three hypothetical pathways were examined:
(a) a
direct
effect
of
objective monotony
on the
outcome vari-
ables,
(b)
complete mediation
by
subjective monotony,
and (c)
partial mediation
by
both predictors.
For
subjec-
tive
monotony
to
play
a
mediatory role,
it
should
be re-
lated
to
both objective work conditions
and the
outcome
measures. Thus,
we first
examined whether objective
work
conditions
and
workers' monotony ratings were
re-
lated. Results indicated that, indeed,
in
both repetitive
work
and
work underload,
the
monotony ratings were sig-
nificantly
higher than
in
varied work. These results
are
consistent with other
findings
(Caplan
et
al.,
1975; Cox,
1985;
Johansson
et
al.,
1978; Wilkes
et
al.,
1981).
In ad-
dition,
the
subjective-monotony ratings
by
workers
in re-
petitive work were related
to the
degree
of
work repeti-
tiveness (indexed
by
length
of the
work
cycle).
This
ex-
tends
the findings of
previous studies relying
on
workers'
perceptions
of
cycle time
and the
degree
of
repetitiveness
(Cox, 1985).
Despite
the
apparent correspondence between
the
various types
of
objective
and
subjective monotony, most
of
the
correlations
for
both sexes were about .30. Thus,
objective
and
subjective monotony were only moderately
related. This concords with
the
generally reported trend
concerning
the
relationship between self-reports
and the
objective
work environment (Jex
&
Beehr,
1991;
Kasl,
1984;
Spector,
1992).
Our
correlations were even higher
than those previously obtained. Spector
and Jex
(1991),
for
example,
found
a
lower convergence between
self-re-
ports
of job
characteristics
and
job-analysis data.
The
lower
convergence
may
have been because
the
job-analy-
sis
data
were
not
based
on
direct observation
of
employ-
ees'jobs
but
rather were derived
from
job-description rat-
ings
and the
Dictionary
of
Occupational
Titles (DOT).
The
pattern
of the
relationship between
the
predic-
tor
variables
and the
employee
outcomes
was
tested
through
a
series
of
hierarchical regression analyses.
The
findings
obtained
for
job
(dis)
satisfaction, psychological
distress,
and
sickness absence were remarkably similar.
Both objective work conditions
and
subjective monotony
were
independently
and
positively related
to all
three out-
comes
in the
entire sample
(men
and
women),
thus sup-
porting
the
partial-mediation model. These
findings are
in
line
with
Garden's
(1982)
conclusion that there
is a
direct relationship among objective work conditions, psy-
chological stress,
and
impaired well-being. They also sup-
port
and
replicate earlier
findings
regarding
the
associa-
tion between
job
perceptions
(as
monotonous, hectic,
boring, etc.)
and
stress outcomes
(Alfredsson
et
al.,
1982;
Coburn,
1979;
Cox, 1985; Karasek, 1979). However,
an
examination
of the
proportional contribution
of the
pre-
dictor variables
to the
variance
in the
outcomes studied
showed
job
(dis)
satisfaction
and
psychological distress
to
be
associated mainly with perceived monotony. Sickness
absence,
on the
other hand,
was
equally influenced
by ob-
jective work conditions
and
perceived monotony.
The findings
here demonstrate
the
importance
of
noting both
the
objective work conditions
and the
work-
ers' appraisals
of
their work because both measures inde-
pendently predicted important employee outcomes.
Re-
liance
on
only subjective perceptions,
as in
many earlier
studies, only partly captures
the
impact
of the
work envi-
ronment
on
workers.
We
wish
to
emphasize that although
the
impact
of
objectively
determined repetitive work
and
work
underload
was
theoretically expected,
to the
best
of
our
knowledge, this
is the first
time
it has
been demon-
strated
in a
large sample
of men and
women. Further-
more, this
finding was
obtained
in a
comparison with
workers engaged
in
varied work
in
similar work environ-
ments.
Such
a
comparison
group
was
missing
in
some
previous studies.
The
study
of
workers engaged
in
repetitive work
of
different
cycle lengths enabled
us to
test
the
hypothesis
that
the
stressfulness
of
this type
of
work
is
related
to its
repetitiveness. Other studies have suggested that
the de-
gree
of
work repetitiveness
is
related
to
generalized stress
outcomes, such
as
subjective well-being (Johansson
et
al.,
1978),
job
pleasantness, postshift stress,
and
self-re-
ported symptom level
(Cox,
1985).
Support
for
this
hy-
pothesis
was
gained here only
for the
attitudinal outcome
of
job
satisfaction. Thus,
the
relationship between work
repetitiveness
and
generalized stress outcomes remains
equivocal.
Psychological distress
and
sickness absence,
on the
other hand, were positively
affected
only
by
short-cycle
(hectic)
repetitive work. Medium-
and
long-cycle repeti-
tive
work were
not
sufficiently
stressful
to
influence
these
outcomes. Both outcomes were shown here
and
else-
where
to be
greatly influenced
by
factors
not
related
to
the
work environment.
The
fact
that
these
outcomes
were
significantly
affected
by
hectic repetitive work, even
after
controlling
for
potential
confounders,
replicates
and
fur-
ther validates
its
negative impact.
It
reinforces
the
con-
ceptualization
of
hectic work
as
highly
stressful
and
threatening
to
both physical
and
psychological health
(Alfredsson
&
Theorell, 1983; Lundberg
et
al.,
1989;
Wilkes
etal.,
1981).
An
important contribution
of
this study
is the
evi-
dence
of a
relationship between work underload
and

40
MELAMED,
BEN-AVI,
LUZ,
AND
GREEN
stress outcomes. There
is
considerable
confusion
regard-
ing
the
operational
definition
of
work underload.
Its
definition
here clearly distinguished
it
from
repetitive
work
in
terms
of
both
physiological requirements
and
work
demands
(except
for the
frequent
need
for
continu-
ous
attention
in
both types
of
work).
Nevertheless, work
underload
was
independently
and
negatively related
to
job
satisfaction
and was
positively related
to
sickness
ab-
sence. Furthermore, work underload
was
rated
as
monot-
onous
to the
same
extent
as
hectic
repetitive
work. These
findings
lend
support
to the
assumption, rarely tested
under
field
conditions,
of the
stressfulness
of
work
underload
(Frankenhaeuser
&
Gardell, 1976; Gardell,
1982,
1987)
and to
similar
findings in the
NIOSH
study
(conducted among
men
only; Caplan
et
al,
1975).
They
may
also suggest
that
what contributes
to the
negative
impact
on
well-being
and
behavior
are the
psychosocial
stressors common
to
both types
of
work rather than
the
physiological
load characterizing hectic repetitive work.
An
additional
aim of
this
study
was to
examine
pos-
sible gender
differences
in the way in
which objective
and
subjective
monotony
affect
stress outcomes. This
was
achieved
by
entering
all
possible
sex
interactions with
the
main predictor variables
in the
regression equations.
The
nature
of the
significant
interactions obtained
was
depen-
dent
on the
outcome studied.
For the
subjective out-
comes—job
satisfaction
and
psychological
distress—sex
interacted
with
subjective monotony. Yet,
the
increment
in
explained variance
after
the
inclusion
of
those interac-
tions
was
rather
small,
albeit
statistically
significant
(mainly
because
of the
large sample
size).
This leads
to
the
conclusion that
the
direction
of the
association
be-
tween
the
independent variables
and the
subjective out-
comes
was
essentially similar
for
both sexes.
More pronounced gender
differences
were observed
for
sickness absence.
Significant
sex
interactions were
found
with objective work conditions. Compared with
varied
work, engagement
in all
types
of
monotonous
work
(except
for
long-cycle repetitive work, which
was
the
least stressful)
was
associated
with significantly
higher
absences
in
women.
In
men,
if
anything,
the
oppo-
site
was
observed (see regression lines
in
Figure
2). The
inclusion
of
these interactions
in the
regression equations
nearly
doubled
the
explained variance
in
sickness
absence.
The
explanation
for
this gender
difference
in
sick-
ness
absence
is not
clear.
It may be
possible
that
women
are
more distressed
by
unfavorable
job
conditions
than
men
and are
consequently absent more. Alternatively,
women
and men may be
equally negatively affected,
but
as
women generally tend
to
have higher sickness-absence
rates,
as
found
here
(p <
.005)
and
elsewhere
(Hackett,
1989;
Zaccaro, Craig,
&
Quinn,
1991),
they
may use
sickness
absence
as a way of
coping with stress
more
than
men
do.
Further studies
are
warranted
to
elucidate
the
reasons
for
these
findings.
The
proportion
of
blue-collar workers, particularly
women,
employed
in
objectively monotonous work con-
ditions
was
very high
(78.9%
of the
women
and
66.2%
of
the men in the final
sample). Such conditions were
shown
to be
stressful,
thus suggesting
the
need
for
some
form
of
intervention
to
reduce their impact
on
workers.
As
organization-level intervention appears
to be
more
effective
than individual-level intervention (Burke,
1993),
efforts
should
be
directed
at
changing work
conditions (e.g., introducing greater rotation between
workstations).
Special attention should
be
given
to
short-
cycle
(hectic) repetitive work
(in
which
45.2%
of the
women
here were
engaged);
perhaps
the
number
of
hours
spent
in
this type
of
work should
be
curtailed.
This study, which included
a
large sample
of
blue-
collar workers,
had a
number
of
limitations—many
of
them
in
common
with
other epidemiological studies:
1.
A
relatively high
proportion
of
participants
(29.8%)
had to be
excluded
from
the
study because
of the
difficulty
in
classifying
their work conditions into
one of
the
five
distinct, operationally predefined categories.
The
relatively
high exclusion rate illustrates
a
basic problem
encountered
in
occupational epidemiological studies:
Many
workers
are
typically subjected
to
mixed condi-
tions.
This
highlights
the
probability
of
classification
er-
rors arising
from
the
assignment
of
job
characteristics
on
the
basis
of
occupational titles alone.
The men and
women excluded from
the
analysis were
older
and
less
educated,
and
they tended
to be
employed
in
heavier
physical
work, making them likely
to
have been more
affected.
Thus,
the
exclusion,
if
anything,
may
have
di-
luted
the
effects.
2.
Workers under monotonous work conditions
may
be
exposed
to
other occupational stressors
as
well.
How-
ever,
because
our
data
were aggregated across workers
from
21
factories with
different
environmental
and
work
conditions,
it
seems
fair
to
assume
that
the
contribution
of
additional
stressors
was
leveled out.
3. We did not
record
how
much
of the
short-,
medi-
um-,
and
long-cycle repetitive work
was
machine
paced.
Machine pacing
is
considered
an
additional source
of
pressure
(Salvendy
&
Smith,
1981),
so it may
have con-
tributed
to the
stressfulness
of the
repetitive work.
In fu-
ture studies
of the
impact
of
objective work conditions,
this variable should
be
included
to
assess
its
unique
contribution.
The
scarcity
of
large-scale studies comparable
to
this
one
leaves many questions unanswered.
We
focused pri-
marily
on a
single perceived negative
job
characteristic—
subjective
monotony—as
this
has
been
the
most fre-
quently examined
in
epidemiological studies
of the im-
pact
of
monotony
at
work. However, other negative
job

OBJECTIVE
AND
SUBJECTIVE WORK MONOTONY
41
characteristics, including
skill
underutilization, lack
of
opportunity
for
learning
new
things
or
exercising
control,
time pressure,
and
unmeaningful
jobs
(Cox, 1985; Gar-
dell, 1987; Johansson, 1989) have also been implicated.
Further studies
are
warranted
to
determine
the
contribu-
tion
of
each
job
characteristic
to the
stressfulness
of the
objective
conditions
of
work monotony.
The
present evidence supporting
the
stressfulness
of
work
underload,
in
combination with
its
prevalence
in
modern industry, suggests that this issue deserves
further
study.
The findings
here were pooled
from
various
un-
derload tasks
with
varying degrees
of
complexity. More
extensive
studies
are
needed
to
examine whether less
complex
tasks, such
as
sorting
faulty
items
on a
conveyor
belt,
are
potentially more stressful
than
more
complex
tasks, such
as
control-room
operations.
Further study
is
also needed
to
validate
the
concept
of
work underload.
Most scales used
in the
literature were designed
to
gauge
workload
(or
overload).
However,
low
workload does
not
necessarily imply high underload
(Fletcher,
1988).
Spe-
cific
scales should
be
constructed
to tap
work underload,
perhaps with subdivisions
to
distinguish between qualita-
tive
and
quantitative underload,
as
recently suggested
by
Fisher
(1993).
It
would
be
interesting
to
know whether
workers
engaged
in
underload tasks, such
as
those exam-
ined
here, would indeed report high underload.
References
Adams,
J. D.
(1981).
Health, stress
and
your
lifestyle.
Arling-
ton,
VA: NTL
Institute.
Aldag,
R.
J.,
Barr,
S. H., &
Brief,
A. P.
(1981).
Measurement
of
perceived task characteristics.
Psychological
Bulletin,
90,
415-431.
Alfredsson,
L.,
Karasek,
R.
A.,
&
Theorell,
T.
(1982).
Myocar-
dial
infarction risk
and
psychosocial work environment:
An
analysis
of the
male Swedish working force. Social Science
and
Medicine,
16,
463-467.
Alfredsson,
L.,
&
Theorell,
T.
(1983).
Job
characteristics
of oc-
cupations
and
myocardial infarction risk:
Effects
of
possible
confounding
factors. Social Science
and
Medicine,
17,
1497-
1503.
Baron,
R. M., &
Kenny,
D. A.
(1986).
The
moderator-media-
tor
variable distinction
in
social psychological research:
Con-
ceptual, strategic,
and
statistical considerations. Journal
of
Personality
and
Social
Psychology,
57,1173-1182.
Burke,
R. J.
(1993).
Review: Organizational-level interventions
to
reduce occupational
stressors.
Work
and
Stress,
7,77-87.
Caplan,
R.
D.,
Cobb,
S.,
French,
J. R. P.,
Jr.,
van
Harrison,
R.,
&
Pinneau,
S.
R.,
Jr.
(1975).
Job
demands
and
worker
health:
Main
effects
and
occupational
differences
(NIOSH
Publica-
tion
No.
75-160). Washington,
DC:
U.S.
Department
of
Health, Education
and
Welfare.
Coburn,
D.
(1979).
Job
alienation
and
well-being. Interna-
tional
Journal
of
Health Services,
9,
41-59.
Cohen,
J. A.
(1960).
A
coefficient
agreement
for
nominal
scales.
Educational
and
Psychological
Measurement,
20, 37-
46.
Cox,
T.
(1985).
Repetitive work: Occupational
stress
and
health.
In C. L.
Cooper
& M. J.
Smith
(Eds.),
Job
stress
and
blue collar work
(pp.
85-111).
Chichester, England: Wiley.
Cox,
T,
Thirlaway,
M.,
&
Cox,
S.
(1984).
Occupational well-
being:
Sex
differences
at
work. Ergonomics,
27,
499-510.
Curry,
R.,
Jex,
H.,
Levison,
W, &
Stassen,
H.
(1979).
Final
report
of
control engineering group.
In N.
Moray
(Ed.),
Men-
tal
workload:
Its
theory
and
measurement
(pp. 235-252).
New
York: Plenum Press.
Davis,
D.
R.,
Shackleton,
V.
J.,
&
Parasuraman,
R.
(1983).
Mo-
notony
and
boredom.
In R.
Hockey
(Ed.),
Stress
and
fatigue
in
human performance
(pp.
1-32).
Chichester, England:
Wiley.
Ferguson,
D.
(1973).
A
study
of
occupational
stress
and
health.
Ergonomics,
16,
649-663.
Fisher,
C. D.
(1993).
Boredom
at
work:
A
neglected
concept.
Human Relations,
46,
395-417.
Fletcher,
B.
(C).
(1988).
The
epidemiology
of
occupational
stress.
In C. L.
Cooper
& R.
Payne
(Eds.),
Causes,
coping
and
consequences
of
stress
at
work
(pp. 3-50).
Chichester,
England: Wiley.
Frankenhaeuser,
M.,
&
Gardell,
B.
(1976).
Underload
and
overload
in
working
life:
A
multidisciplinary
approach.
Journal
of
Human Stress,
2,
35-46.
French,
J. R. P.,
Jr.,
Caplan,
R.
D.,
& van
Harrison,
R.
(1982).
Mechanisms
of
job
stress
and
strain.
New
York:
Wiley.
Frese,
M.,
&
Zapf,
D.
(1988).
Methodological issues
in the
study
of
work
stress:
Objective
vs
subjective measurement
of
work
stress
and the
question
of
longitudinal studies.
In C. L.
Cooper
& R.
Payne
(Eds.),
Causes,
coping
and
consequences
of
stress
at
work(pp.
375-411).
Chichester, England: Wiley.
Fried,
Y, &
Ferris,
G. R.
(1987).
The
validity
of the job
char-
acteristics model:
A
review
and
meta-analysis.
Personnel Psy-
chology,
40,
287-322.
Gardell,
B.
(1982).
Scandinavian research
on
stress
in
working
life.
International Journal
of
Health Services,
12,
31-41.
Gardell,
B.
(1987).
Efficiency
and
health hazards
in
mecha-
nized
work.
In J. C.
Quick,
R. S.
Bhagat,
J. E.
Dalton,
& J. D.
Quick
(Eds.),
Work
stress
(pp.
50-11).
New
York:
Praeger.
Hackett,
R. D.
(1989).
Work attitudes
and
employee absentee-
ism:
A
synthesis
of the
literature. Journal
of
Occupational
Psychology,
62,
235-248.
Hammer,
T.
H.,
&
Landau,
J. C.
(1981).
Methodological issues
in
the use of
absence data. Journal
of
Applied
Psychology,
66,
574-581.
James,
L.
R.,
&
Brett,
J. M.
(1984).
Mediators,
moderators,
and
tests
for
mediation. Journal
of
Applied
Psychology,
69,
307-321.
Jex,
S.
M.,
&
Beehr,
T. A.
(1991).
Emerging theoretical
and
methodological issues
in the
study
of
work-related
stress.
Re-
search
in
Personnel
and
Human Resources Management,
9,
311-365.
Johansson,
G.
(1989).
Job
demands
and
stress
reactions
in re-
petitive
and
uneventful
monotony
at
work. International
Journal
of
Health Services,
19,
365-377.
Johansson,
G.,
Aronsson,
G.,
&
Lindstrom,
B. O.
(1978).
So-

42
MELAMED,
BEN-AVI,
LUZ,
AND
GREEN
cial
psychological
and
neuroendocrine stress reactions
in
highly
mechanized work. Ergonomics,
21,
583-599.
Karasek,
R. A.
(1979).
Job
demands,
job
decision latitude
and
mental strain: Implications
for job
redesign. Administrative
Science
Quarterly,
24,
285-308.
Kasl,
S. V.
(1984).
Stress
and
health. Annual Review
of
Public
Health,
5,
319-341.
Kendall,
M.
G.,
&
Stuart,
A.
(1958).
The
advanced theory
of
statistics.
New
York:
Hafner.
Kushnir,
T., &
Melamed,
S.
(1991).
Workload, perceived con-
trol
and
psychological distress
in
Type
A/
B
industrial work-
ers. Journal
of
Organizational Behavior,
12,
155-168.
Lundberg,
U.,
Granqvist,
M.,
Hansson,
T.,
Magnusson,
M.,
&
Wallin,
L.
(1989).
Psychological
and
physiological
stress
re-
sponses during repetitive work
at an
assembly line.
Work
and
Stress,
3,
143-153.
Meir,
E.
I.,
&
Melamed,
S.
(1986).
The
accumulation
of
per-
son-environment congruences
and
well-being. Journal
of
Occupational
Behaviour,
7,
315-323.
Melamed,
S.,
Luz,
J., &
Green,
M. S.
(1992).
Noise exposure,
noise annoyance
and
their relation
to
psychological
distress,
accident
and
sickness absence among blue collar workers
in
the
CORDIS study. Israel Journal
of
Medical Sciences,
28,
629-635.
Muchinsky,
P. H.
(1977).
Employee absenteeism:
A
review
of
the
literature. Journal
of
Vocational
Behavior,
10,
316-340.
O'Hanlon,
J. F.
(1981).
Boredom: Practical consequences
and
a
theory. Acta
Psychologica,
49,
53-82.
Salvendy,
G.,
&
Smith,
M.
(1981).
Machine pacing
and
occu-
pational stress. Chichester, England: Wiley.
Smith,
P. C.,
Kendall,
L.
M.,
&
Hulin,
C. L.
(1969).
The
mea-
surement
of
satisfaction
in
work
and
retirement. Chicago:
Rand McNally.
Smith,
R. P.
(1981).
Boredom:
A
review. Human Factors,
23,
329-340.
Spector,
P. E.
(1992).
A
consideration
of the
validity
and
mean-
ing
of
self-report measures
of
job
conditions.
In C. L.
Cooper
& I. T.
Robertson
(Eds.),
International review
of
industrial
and
organizational
psychology
(Vol.
7, pp.
123-151).
Chich-
ester, England:
Wiley.
Spector,
P.
E.,
&
Jex,
S. M.
(1991).
Relations
of job
character-
istics
from
multiple
data
sources with employee
affect,
ab-
sence, turnover intentions,
and
health. Journal
of
Applied
Psychology,
76,
46-53.
Thackray,
R. I.
(1981).
The
stress
of
boredom
and
monotony:
A
consideration
of
the
evidence. Psychosomatic Medicine,
43,
165-176.
Wallace,
M.,
Levens,
M.,
&
Singer,
G.
(1988).
Blue collar
stress.
In
C. L.
Cooper
& R.
Payne
(Eds.),
Causes,
coping
and
con-
sequences
of
stress
at
work
(pp.
53-76).
Chichester, England:
Wiley.
Wilkes,
B.,
Stammerjohn,
L.,
&
Lalich,
N.
(1981).
Job de-
mands
and
worker health
in
machine-paced
poultry inspec-
tion. Scandinavian Journal
of
Work
and
Environmental
Health,
7(Suppl.4),
12-19.
Zaccaro,
S.
J.,
Craig,
B.,
&
Quinn,
J.
(1991).
Prior absentee-
ism, supervisory style,
job
satisfaction,
and
personal
charac-
teristics:
An
investigation
of
some
mediated
and
moderated
linkages
to
work absenteeism. Organizational Behavior
and
Human Decision Processes,
50,
24-44.
Zung,
W. W. K.
(1965).
A
self-rating depression
scale.
Archives
of
General
Psychiatry,
12,
63-70.
Zung,
W. W. K.
(1971).
A
rating instrument
for
anxiety disor-
ders. Psychometrics,
12,
371-379.
Received
February
17,
1993
Revision
received
May 2,
1994
Accepted
June
6,
1994
•