See

discussions, stats, and author profiles for this publication at:

https://www.researchgate.net/publication/232502668

A Meta-Analysis of the Relation of

Job Characteristics to Job
Satisfaction
ARTICLE in JOURNAL OF APPLIED PSYCHOLOGY APRIL 1985
Impact Factor: 4.31 DOI: 10.1037/0021-9010.70.2.280

CITATIONS

READS

300

1,277

4 AUTHORS, INCLUDING:
Raymond A. Noe
The Ohio State University
60 PUBLICATIONS 6,007 CITATIONS
SEE PROFILE

Nancy Moeller
Association of American Medic
2 PUBLICATIONS 300 CITATIONS
SEE PROFILE

Available from: Raymond A. Noe

Retrieved on: 17 January 2016

Journal of Applied Psychology

1985, Vol. 70, No. 2, 280-289

Copyright 1985 by the American Psychological Association, Inc.

0021-9010/85/S00.75

A Meta-Analysis of the Relation of Job Characteristics

to Job Satisfaction
Brian T. Loher and Raymond A. Noe
Michigan State University

Nancy L. Moeller

Michael P. Fitzgerald
Arthur Young and Company Detroit, Michigan

University of Maryland

Job enrichment is one method that has been used to increase employee satisfaction
and work motivation. Hackman and Oldham's (1976) job characteristics model
has served as the foundation for many job enrichment efforts. In particular, a
considerable amount of research has been devoted to the study of the job
characteristics-job satisfaction relation. The purpose of this study was to statistically
determine, using meta-analysis procedures, the "true" relation between job
characteristics and job satisfaction. The role of growth need strength (GNS) as a
possible moderator of this relation was also investigated. Results indicated a
moderate relation between job characteristics and job satisfaction. This relation
is stronger for employees high in GNS. Situational characteristics appear to be
more important in determining satisfaction for employees low in GNS.

Many theorists and researchers have argued

that one way to increase employee performance and satisfaction is by "enriching" the
employee's job. Job enrichment seeks to improve both employee performance and satisfaction by building greater scope for personal
achievement and recognition and greater opportunity for individual achievement and
growth into employees' jobs. Thus, job enrichment can be viewed as an organizational
intervention designed to restructure jobs with
the intent of making them more challenging,
motivating, and satisfying to the individual.
The current emphasis on job enrichment is
caused, in part, by the fact that today's
employees tend to bring more abilities, higher
expectations, and a greater desire for selfresponsibility to the workplace than did their
predecessors (Walton, 1972). To match the
characteristics of the job to the needs of the
individual, job enrichment efforts have generally applied theories that describe how the
The authors would like to thank Neal Schmitt, Dan
Ilgen, and two anonymous reviewers for their constructive
comments.
Requests for reprints should be sent to Brian T. Loher,
Department of Psychology, Psychology Research Building,
Michigan State University, East Lansing, Michigan 488241117.

interaction of job characteristics and individual characteristics are related to organizationally desirable outcomes.
The theoretical basis for many current
enrichment efforts is the Hackman and Oldham (1975, 1976) job characteristics model.
In the model (see Figure 1), specific job
characteristics such as skill variety (the degree
to which a job requires a variety of activities
to carry out the work) and task significance
(the degree to which the job has a substantial
impact on the lives of other people) affect the
individual's experienced meaningfulness of
the work, experienced responsibility for work
outcomes, and knowledge about the results
of his or her work activities (Hackman &
Oldham, 1976). These three "critical psychological states" have, in turn, been linked to
such outcome variables as internal work motivation, job satisfaction, absenteeism, turnover, and work quality (Ford, 1969; Hackman,
Oldham, Janson, & Purdy, 1975; Wanous,
1974). As originally conceived, the job characteristics model also included the individual
characteristic of growth need strength (GNS)
as a moderator of the relation between the
characteristics of the job and the outcome
variables (Hackman & Oldham, 1976).
Hackman and Oldham assumed that one of
the most important work values is the job

280

JOB CHARACTERISTICS-JOB SATISFACTION RELATION

Five Core
Characteristics
Skill variety
Task Identity

Psychological
States
"~?elings of

eanlngfulness

Task significance

Outcomes
-High Intrinsic
motivation
High quality work
High satisfaction

responsibility
i

results

Low absenteeism
I and turnover

Employee Growth Need Strength

Figure 1. Relations between core job dimensions, critical

psychological states, and on-the-job outcomes (from
Hackman & Oldham, 1976).

incumbent's need for personal growth and

development through his or her job. Employees with high GNS should respond more
positively to jobs that have high levels of the
five core dimensions (task significance, skill
variety, etc.) than employees with low GNS.
Various other individual characteristics such
as locus of control, knowledge, and skill have
also been hypothesized to moderate the outcomes of enriched jobs (Oldham, Hackman,
& Pearce, 1976; Wanous, 1974).
The results of research studies attempting
to evaluate the relation between job characteristics and employee outcomes have been
equivocal at best (White, 1978), possibly due
to the impact of situational variables that
vary across organizations (e.g., climate, work
group norms). Also, individual characteristics
besides GNS (e.g., authoritarianism or locus
of control) and demographic variables (e.g.,
urban/rural socialization, educational level)
may act as moderators (Pierce & Dunham,
1976; Roberts & Click, 1981; White, 1978).
The typical procedure used to study moderators of the relation of job characteristics
and employee outcomes involves subgroup
analysis; participants are split into two or
more groups based on their standing on the
moderator of interest. Correlations between
job characteristics and outcome variables are
then calculated and compared across groups
(Stone, 1976). In the Hackman and Oldham
(1975) model, the individual characteristic of
GNS was intended to function as a moderator.
Indeed, GNS has emerged as something of a
moderator in that GNS has "generally influ-

281

enced the magnitude rather than the sign of

the focal relationship" (Aldag, Barr, & Brief,
1981, p. 427). However, inconsistent findings
have plagued this facet of research. Because
of these inconsistent findings, Roberts and
Glick (1981) strongly questioned the utility
of including GNS as a moderator of the job
characteristics-employee outcome relationship.
From a practical standpoint, the evidence
does not appear sufficient to unequivocally
support the view that increasing the presence
of core job dimensions (i.e., enriching the
job) results in beneficial organizational and
employee outcomes. Job enrichment may
result in desirable outcomes for one group of
employees (high GNS) and may be more
successful in organizations with certain characteristics.
The purpose of this study is to apply metaanalytical techniques to determine (a) the
strength of the relation between job characteristics and one employee outcome, job satisfaction, and (b) whether the relation between
job characteristics and job satisfaction is
moderated by GNS.
Schmidt and Hunter (1977) have demonstrated how meta-analysis can be used to
estimate the "true" relation between variables
through identifying the extent to which variance in observed correlation coefficients
across studies is due to statistical artifacts
such as sampling error and unreliability in
measurement. Furthermore, these procedures
have also been applied to subgroup analyses.
First, meta-analysis procedures are applied
to the entire sample. If a "substantial" amount
of residual variance remains after corrections
for statistical artifacts, the sample is separated
into two or more subgroups on the basis of
their score on the moderator variable of
interest. Schmidt and Hunter (1977) defined
a "substantial" amount of residual variance
as being when 25% or more of the total
observed variance is not accounted for by
statistical artifacts. Mean weighted correlations and variance estimates are then calculated for each subgroup. Statistical evidence
can be taken as support for the moderating
effect when the average correlation varies
between subgroups and residual variance in
the correlation coefficient approaches zero

282

LOHER, NOE, MOELLER, AND FITZGERALD

within the subgroups (Hunter, Schmidt, &

Jackson, 1982).
Method
Compiling the Distribution
of Observed Correlations
The first step in the cumulative analysis was to establish
criteria to define which studies should be included.
Hackman and Oldham (1976) developed the Job Diagnostic Survey (JDS) to measure the facets of jobs that
are present in their job characteristics model. Therefore,
the first inclusion criterion was that only studies that
used the JDS or a JDS-like measure of perceived task
characteristics were included in the analysis. A second
criterion pertained to the measure of job satisfaction.
Only job satisfaction scales designed to measure the
feelings or subjective reactions of respondents were included. In addition, only studies reporting individual
responses rather than aggregates and zero-order correlations between job characteristics and job satisfaction were
included.
Academic and practitioner-oriented journals in the
areas of psychology, human resource management, and
organizational behavior were reviewed for studies that
met our inclusion criteria. The studies included in the
meta-analysis are listed in the Appendix. The data recorded from each study included: (a) sample size, (b) a
brief description of the sample, (c) the name and type of
satisfaction measure used, (d) reliabilities (when reported)
of the JDS task dimensions and satisfaction measure,
and (e) the reported correlations between the JDS task
dimensions and job satisfaction. Studies that met our
inclusion criteria were classified as (a) studies of the
direct relation between job characteristics and job satisfaction (Type O or overall studies), and (b) studies of the
relation between job characteristics and job satisfaction
as moderated by GNS (Type M or moderated studies).

Data Analysis
Both Type O and Type M studies were included in the
analysis for establishing the "true" relationship between
job characteristics and job satisfaction. If job characteristics-job satisfaction correlations were reported for highand low-GNS groups but not for a combined sample,
then each GNS group within the Type M studies was
treated as if it were a separate study. Only Type M
studies were included in the moderator analysis.
Initially, separate analyses were conducted of the relation
between each of the five task dimensions and job satisfaction. This was done to ascertain the "true" relationship
between each task characteristic and job satisfaction and
to determine which, if any, of the task characteristics
were more strongly related to job satisfaction. A second
analysis of the job characteristic-job satisfaction relation
was conducted using the average of correlations between
each of the JDS task dimensions and the job satisfaction
measure. The rationale for combining the task dimensions
was based on the findings that the intercorrelations
among task dimensions are generally high and positive
(Hackman & Lawler, 1971). Additionally, Aldag et al.
(1981) found that the dimensionality of the JDS remains
questionable in view of the number of studies reporting

factor structures that are inconsistent with the original

five factor solutions proposed by Hackman and Oldham.
The zero-order correlations between the task dimensions
and the job satisfaction measures are presented in Table
1 for the Type O studies. The resulting average correlation
between the job characteristics dimensions and job satisfaction (job characteristics index) is presented in the
last column of Table 1 for each study. The analogous
correlations for Type M studies are presented in Table
2. The mean estimates presented in Tables 1 and 2 are
likely to be underestimates of the job characteristic-job
satisfaction relation for two reasons. First, these values
are not corrected for unreliability and, second, the job
characteristics index is a mean rather than a composite
value, which does not take into account the intercorrelations among dimensions.
The first step in the meta-analysis was to calculate the
average sample-weighted correlation between job complexity and satisfaction using the Schmidt-Hunter procedures. The sample-weighted correlation coefficient gives
more weight to correlations that presumably have the
least sampling error (i.e., those with the largest sample
size), thereby giving the best estimate of the "true"
relationship between the variables of interest (Schmidt
& Hunter, 1977).
The calculation of the weighted variance of the sample
correlation coefficients and the calculation of the error
variance was the next step in the analysis. The weighted
variance of the sample correlation coefficients is the
observed variance in the distribution of the correlation
coefficients. The error variance is the variation in the
correlations across studies that would be expected to
occur as a result of sampling error.
Sampling error is only one of several artifactual sources
of variance that may account for differences in observed
correlations across studies (Schmidt, Hunter, Pearlman,
& Shane, 1979). Two other such artifactual sources that
were of interest in the present study were differences
among studies in predictor reliability and differences
among studies in criterion reliability. Error of measurement has the effect of lowering the observed correlation
between job characteristics and job satisfaction. Therefore,
correction for unreliability in the measures of the job
characteristics, the job characteristics index, and satisfaction represented the next step in the analysis. For the job
characteristics index, the reliability for each study was
calculated by averaging across the reliabilities of the five
task dimensions. The predictor and criterion reliabilities
for Type O and Type M studies are presented in Table
3. Not all of the studies in the analysis reported reliabilities
for both the JDS task dimensions and/or the satisfaction
measure (unmatched data). Therefore, the interactive
formula (Hunter, Schmidt, & Jackson, 1982) was used
to obtain the variance of the true score correlations after
correcting for unreliability in the JDS and job satisfaction
measures.
The final step in the analysis was to calculate the 95%
confidence interval for the true score correlation between
job characteristics and job satisfaction.

Results
A total of 28 studies that met the inclusion
criteria were found. All but one of the studies

JOB CHARACTERISTICS-JOB SATISFACTION RELATION

were published in the following journals:

Academy of Management Journal, Administrative Science Quarterly, Human Relations,
Journal of Applied Psychology, Journal of
Business Research, Organizational Behavior
and Human Performance, and Personnel Psychology. Table 4 presents the characteristics
of the 28 studies.
The number of studies in which the impact
of job enrichment as an organizational intervention was assessed were quite small. Only
three studies involved an actual change in
the job process or technology. A number of
other studies that reported the results of

283

actual job enrichment interventions were

found, but these articles did not report the
statistical information needed in order to be
included in the present study (e.g., correlations between task dimensions and job satisfaction). The studies were almost equally split
into those in which the sample was performing the same job (N = 12) and those in which
data were collected across a number of different jobs (N = 16). Whereas the JDS was
the prevalent measure of job characteristics,
no measure of job satisfaction was used consistently across all studies. Finally, only a
small number of studies examined the prob-

Table 1
Correlations Between Job Characteristics and Job Satisfaction for Type O Studies

Study

1.

Aldag& Brief (1975)

Job
Characteristics
Index

T.I.

T.S.

S.V.

Aut.

Fdbk.

104

.34

.43

.32

.51

.37

.394

120
155

.37
.31

.21

.28
.43

.41
.26

.43
.26

.34

88
784

.41
.16

.27
.20

.34
.18

.51
.27

.43
21

.392
.204

343
129
171

.15
.38
.66

.16
.54
.69

.29
.35
.71

.24
.66
.66

.22

208

.20

.26

.38

.39

.28

.313

658

.22
.21
.22

.21
.24
.25

.32
.23
.23

.38
.28
.27

.38
.26
.26

.302
.244
.246

.36

.302
.332

2. Arnold & House

(1980)

3. Brief & Aldag ( 1978)

4. Caldwell & O'Reilly
(1982)

5. Dunham (1977)
6. Evans, Kiggundu &
House (1979)

7. Griffin (1981) I
II
8. Hackman & Lawler
(1971)

.315

.483

.68

9. Hackman & Oldham

(1976)

10. Katz(1978a)
11. Katz (1978b)
12. Katz & Van Maanen

3060
2094

13. Kiggundu (1980)

14. O'Reilly, Parletee, &

138

.24
.34

.23
.21

.28
.57

.40
.60

Bloom (1980)
Orpen (1979)
Rousseau (1977)
Rousseau (1978)
Schmitt, Coyle, White,
& Rauschenberger
(1978)
19. Schmitt & White1
20. Sims &Szilagyi( 1976)
21. Walsh, Taber, & Beehr
(1980)" I

76
36
199
271

.23
.36
.26
.05

.33
.35
.54
.18

.33
.27
.58
.38

.39
.28
.48
.47

.46
.32
.37
.37

.348
.316
.446

411
860
766

.14
.24
.30

.27
.33

.42
.31
.54

.40
.28
.26

.26
.17
.26

.298
.266
.340

486
96
232

.28
.32
.33

.32
.12
.20

.31
.24
.38

.12
.28
.20

.258

(1977)

3500

15.
16.
17.
18.

II
III

-.06

.29

.24
.278

Note. T.I. = Task Identity; T.S. = Task Significance; S.V. = Skill Variety; Aut. = Autonomy; Fdbk. = Feedback.
* Unpublished study via personal communication.
b
Three separate studies in one publication.

284

LOHER, NOE, MOELLER, AND FITZGERALD

lem of common method variance. The average relationship is moderated by other variables.
correlation between the measures of growth An examination of the 95% confidence interneed strength, task characteristics, and job vals indicated that no one task characteristic
satisfaction was .05, which suggests that the necessarily has a stronger relationship with
correlations were not significantly inflated job satisfaction than any other (see Figure 2).
because of the reliance on self-report mea- This gives further evidence for the lack of
sures.
dimensionality of the JDS, that is, the JDS
The results of the analysis of the relation may be a measure of the overall complexity
between each of the five skill dimensions and of the job rather than of specific job characjob satisfaction are presented in Table 5. The teristics. These results support combining the
correlations after correcting for sampling error task dimensions in order to establish the
and reliability in the measures range from relation between "job complexity" and job
.46 for Autonomy to .32 for Task Identity. satisfaction.
Controlling for statistical artifacts accounted
The results of the analysis of the job charfor less than 75% of the observed variance in acteristic-job satisfaction relation based on
all of the task characteristics, which suggests the average of correlations between each of
that the job characteristic-job satisfaction the JDS task dimensions (job characteristic
Table 2
Correlations Between Job Characteristics and Job Satisfaction for Type M Studies

Study
High growth need strength
1. Armenakis, Field, Holley,
Bedeian & Ledbetter
(1977)
2. Brief &Aldag( 1975)
3. Griffin (1982)
4. Griffin (1981) I

II
5. O'Reilly & Caldwell
(1979)
6. Orpen(1979)
7. Pokorney, Gilmore, &
Beehr (1980)
8. Umstot, Bell, & Mitchell
(1976)
9. Wanous(1974)
Low growth need strength
1. Armenakis, Field, Holley,
Bedeian, & Ledbetter
(1977)
2. Brief & Aldag ( 1975)
3. Griffin (1982)
4. Griffin (1981) I

H
5. O'Reilly & Caldwell
(1979)
6. Orpen (1979)
7. Pokorney, Gilmore, &
Beehr (1980)
8. Umstot, Bell, & Mitchell
(1976)
9. Wanous(1974)

T.I.

T.S.

S.V.

Aut.

Fdbk.

Job
Characteristics
Index

28
35
48
65
65

.47
.40
.48
.52
.73

.28
.47
.53
.56
.82

.80
.53
.74
.52
.72

.66
.36
.84
.44
.71

.553
.440
.648
.510
.745

37
18

.53
.29

.41
.27

.87
.19

.85
.46

.65
.28

.662
.298

44

.40

.64

.34

.42

.63

.486

50
37

.27

.70

.70
.50

.71
.59

.76
.41

.628
.450

27
35
38
64
64

-.10
.33
.13
.18
.57

.09
.35
.60
.53

.03
.35
.23
.11
.73

.38
.36
.13
.42
.63

.348
.115
.328
.615

37
18

.31
.31

.29
.18

.63
.14

.58
-.02

.31
.09

.424
.140

54

.17

.29

.09

.23

.15

.186

.67

.28

.73
.15

.41

.48
.10

.514
.023

50
37

.30

-.07

-.03

-.09

.10

Note. T.I. - Task Identity; T.S. - Task Significance; S.V. = Skill Variety; Aut. = Autonomy; Fdbk. = Feedback.

JOB CHARACTERISTICS-JOB

index) are presented in Table 6. The sampleweighted correlation coefficient was .39 after
correcting for attenuation due to unreliability
in the task characteristic and job satisfaction
measures. The observed variance in the correlations was .0059. After correcting for variance due to sampling error and unreliability
in the measures, the variance in the correlations was still .0028. These statistical artifacts
account for 53% of the observed variance in
the correlation coefficients. This indicates
that 47% of the observed variance is due to
other factors, one of which may be GNS. If
both job characteristics and job satisfaction
had been perfectly measured, we would expect
to find a "true" distribution centered on .39,

285

SATISFACTION RELATION

with a standard deviation of .0712 and a 95%

confidence interval ranging from .25 to .53.
The large amount of variance that remained unexplained in the overall analysis
indicated the necessity for subgroup analyses.
As a result, subgroup analyses were performed
for the high- and low-GNS groups (Type M
studies). Results of the analysis of the influence of GNS on the job characteristics-job
satisfaction relationship are presented in Table
7. The sample-weighted correlation for the
high GNS group was .57. The observed variance for the high GNS group was .0135. The
"true" relationship between job characteristics
and job satisfaction was .68 after correcting
for unreliability of measurement. Approxi-

Table 3
Reliability Data for Job Characteristics and Job Satisfaction Measures
for Type O and Type M Studies
Mean job
satisfaction

T.I.

Study

T.A.

S.V.

Aut.

Fdbk.

'yy

Type O studies

1.
2.
3.
4.

Aldag & Brief (1975)

Brief & Aldag (1978)
Dunham (1977)
Evans, Kiggundu, &
House (1979)
5. Griffin (1981) I

II

.63

.72

.63

.72

.62

.76

.69

.73

.63

.75

.64

.74

.87

.52
.91
.97

.50

.53
.89
.96

.53
.89
.96

.38
.86
.96

.49
.89
.96

.79
.93

.77
.72
.62

.73
.59

.90
.83
.78

.77
.75
.63

.75
.71
.70

.80
.75
.66

.76
.74
.88

.59

.36

.63

.54

.62

.55

.67

.76
.61

.70
.57

.74
.51

.75
.51

.77
.48

.74
.54

.79
.70

.53
.55
.69

.50
.41
.37

.53
.67
.64

.36
.73
.72

.48
.59
.61

.74
.73
.69

.63
.93

.62
.84

.69
.91

.63
.92

.64
.90

.94

.73

.72

.75

.74

6. Hackman & Lawler

(1977)

7. Katz(1978b)
8. Kiggundu (1980)
9. O'Reilly, Parlette, &
Bloom (1980)

10. Rousseau (1978)

11. Schmitt, Coyle, White, &
Rauschenberger (1978)

12. Schmitt & White

13. Walsh, Taber, & Beehr
(1980)

II
III

Type M studies
1. Brief* Aldag (1975)
2. Griffin (1982)
3. Pokorney, Gilmore, &
Beehr(1980)
4. Wanous (1974)

.76

.76

Note. T.I. = Task Identity; T.S. = Task Significance; S.V. = Skill Variety; Aut. = Autonomy; Fdbk. = Feedback.

286

LOHER, NOE, MOELLER, AND FITZGERALD

Table 4

limits for the 95% confidence interval were

.64 and .72.
The sample-weighted correlation coefficient
for the low-GNS group was .32, and the
observed variance was .0368. After correcting
for variance due to sampling error and unreliability, variance for the low-GNS group
was still .0169 (approximately 46% of the
observed variance). This indicates that for
the low-GNS group, other factors besides
GNS may be affecting the relation between
job characteristics and job satisfaction. After
correcting the sample-weighted correlation
for attenuation due to unreliability in the
measures of job characteristics and job satisfaction, the correlation between the two
variables was .38, with the 95% confidence
interval ranging from .08 to .68.

Characteristics of Studies Used in Meta-Analysis

Intervention
Yes
No
Sample
Multiple jobs
Same job
Measure of task characteristics
JDS
JCI
Measure of satisfaction
Alderfer ERG Scale
Brayfield & Rothe
Hackman & Lawler
JDIWork
JDSGeneral Satisfaction
Kunin Faces Scale
Minnesota Satisfaction Questionnaire
Other (e.g., Index of Organization Reactions,
Yale Job Inventory, Lawler & Hall)
Assess common method variance
Yes
No

3
25
16
12

23
5
2
3
3
2
4
2
2
10

4
24

Discussion
The results of the analysis involving studies
of the job characteristics-job satisfaction relation appear to answer our question regarding
the level of the relation between job characteristics and job satisfaction. The correlation
between the job characteristics index and job
satisfaction is about .39. The relation between
each of the task characteristics and job sat-

Note. JDS = Job Diagnostic Survey; JCI = Job Characteristic Inventory; JDI = Job Descriptive Index; ERG =
Existedness, Relatedness, Growth.

mately 97% of the observed variance was

accounted for by sampling error and measurement unreliability. The lower and upper

Table 5
Results of the Meta-Analysis by Skill Dimension
Variable
Sample-weighted correlation
coefficient
Observed variance
Variance due to sampling
error
Variance corrected for
sampling error
Variance due to
unreliability in the
measures
Sample-weighted correlation
after correcting for
unreliability in the
measures
Variance of distribution of
true score correlations
Upper 95% confidence limit
Lower 95% confidence limit

T.I.

T.S.

S.V.

Aut.

Fdbk.

.24

.25

.30

.34

.29

.006403

.003923

.013691

.009454

.011311

.002177

.001983

.002030

.001912

.002041

.004226

.002440

.011661

.007542

.009270

.000663

.000835

.001624

.001346

.001968

.32

.38

.41

.46

.41

.006389

.003587

.018664

.011106

.013863

.48
.16

.50
.26

.68
.14

.67
.25

.64
.18

Note. T.I. = Task Identity; T.S. = Task Significance; S.V. = Skill Variety; Aut. = Autonomy; Fdbk. = Feedback.
Total sample size was 15,542.

287

JOB CHARACTERISTICS-JOB SATISFACTION RELATION

Table 6
Results of the Overall Mela-Analysis
Sample-weighted correlation
coefficient
Observed variance
Variance due to sampling error
Variance corrected for sampling
error
Variance due to unreliability in the
measures
Sample-weighted correlation after
correcting for unreliability in the
measures
Variance of distribution of true score
correlations
Upper 95% confidence limit
Lower 95% confidence limit

Task Identity

Task Significance

.29
.005882
.002058

.26i-

Skill Variety

Autonomy

.003824
.001046

Feedback

-1.64
.OO

.39
.005068
.53
.25

Note. Total sample size was 15,542.

isfaction ranges from .32 (task identity) to

.46 (autonomy). Some unexplained variance
in the correlations between studies remains
after correcting for variance due to the artifacts of sampling error and unreliability in
the predictor and criterion measures. This
level of correlation should certainly offer
some support for efforts to increase job satisfaction through the use of job enrichment.
However, the results from the moderated
studies warn that simply enriching a job will
not necessarily hold the same amount of
benefit for everyone.
The results for the studies with GNS as a
moderator are particularly intriguing. We can
now state with some confidence that growth
need strength (GNS) acts as moderator of the
relation between job characteristics and job
satisfaction. The correlation between job
characteristics and satisfaction is .68 for per-

.10

.20

.30

.40

.SO

.60

.70

Figure 2. 95% confidence intervals for task characteristics.

sons who are high on GNS and about .38 for

persons who are low on GNS. The variance
in the correlations for the high GNS group
is essentially eliminated after controlling for
sampling error and measurement unreliability. However, the variance in the correlations
between job characteristics and job satisfaction is still large for the low-GNS group, even
after correcting for sampling error and unreliability in the JDS and job satisfaction
measures. Apparently, factors that do not
affect the relation between job characteristics
and job satisfaction for persons with high
growth need strength do come into play for
persons who are low on growth need strength.
The model in Figure 3 illustrates how such
a relation migh operate.
The more complex and enriched a job is,
the more likely the high-GNS person who
possesses a high need for personal growth
and development (Hackman & Oldham,
1976) is to be satisfied with that job. In
contrast, for low-GNS employees, who have
less need for growth and development, the
presence of certain external situational char-

Table 7
Results of the Moderator Analysis
Variable
Sample-weighted correlation coefficient
Observed variance
Variance due to sampling error
Variance corrected for sampling error
Variance due to unreliability in the measures
Sample-weighted correlation after correcting for unreliability in the measures
Variance of distribution of true score correlations
Upper 95% confidence limit
Lower 95% confidence limit

High GNS

Low GNS

.57

.32

.013517
.010721
.002796
.002458

.036792
.019083
.017709
.000764

.68

.38

.000482

.024179

.72
.64

.68
.08

Note. Total sample size for high-GNS group was 427. For low-GNS group, sample size was 424.

288

LOHER, NOE, MOELLER, AND FITZGERALD

the job characteristics-job satisfaction relation

will help to determine the extent to which
common method variance results in spuriously inflated correlation coefficients.
References

Figure 3. A model of the proposed relation between job

complexity, GNS, and job satisfaction.

acteristics (such as work group or management support for enrichment activities) may
be necessary if the core job dimensions are
to increase employee satisfaction. That is, the
opportunities an enriched job offers, in and
of themselves, may not be recognized or
cared about by low-GNS employees. What
may be more important is how the employee's
work group views these opportunities. If the
work group is supportive of enriched work,
this may help to enhance the employee's
satisfaction with a more complex job. Future
research is necessary to investigate hypotheses
about moderating situational characteristics
for low-GNS persons.
An alternative explanation is that satisfaction influences how individuals describe their
jobs (the authors thank an anonymous reviewer for this suggestion). Satisfied individuals may see their jobs as more "complex"
than dissatisfied individuals. The direction of
arrows in Figure 3 may be reversed.
Finally, the characteristics of the studies
reported in the literature are particularly
disturbing. Few empirical studies of actual
job enrichment interventions have been reported in the professional literature. More
studies evaluating the impact of actual changes
in job characteristics on employee attitudes
and performance need to be conducted. Additionally, more attention needs to be directed
to the impact of common method variance
on the job characteristics-job satisfaction relation. The use of multiple methods (e.g.,
interviews, unobtrusive observation) to assess

Aldag, R. J., Barr, S. H., & Brief, A. P. (1981). Measurement of perceived task characteristics. Psychological
Bulletin, 90, 415-431.
Ford, R. N. (1969). Motivation through the work itself.
New \brk: American Management Association.
Hackman, J. R., & Lawler, E. E. (1971). Employee
reactions to job characteristics. Journal of Applied
Psychology, 55, 259-286.
Hackman, J. R., & Oldham, G. R. (1975). Development
of the job diagnostic survey. Journal of Applied Psychology, 60, 159-170.
Hackman, J. R., & Oldham, G. R. (1976). Motivation
through the design of work: Test of a theory. Organizational Behavior and Human Performance, 16, 250279.
Hackman, J. R., Oldham, G. R., Janson, R., & Purdy,
K. (1975). A new strategy for job enrichment. California
Management Review, 57-71.
Hunter, J. E., Schmidt, F. L., & Jackson, G. B. (1982).
Meta-analysis: Cumulating research findings across
studies. Beverly Hills: Sage.
Oldham, G. R., Hackman, J. R., & Pearce, J. L. (1976).
Conditions under which employees respond positively
to enriched work. Journal of Applied Psychology, 61,
395-403.
Pierce, J. L., & Dunham, R. B. (1976). Task design: A
literature review. Academy of Management Review, 4,
83-97.
Roberts, K. H., & Click, W. (1981). The job characteristics
approach to task design: A critical review. Journal of
Applied Psychology, 66, 193-217.
Schmidt, F. L., & Hunter, J. E. (1977). Development of
a general solution to the problem of validity generalization. Journal of Applied Psychology, 62, 529-540.
Schmidt, F. L., Hunter, J. E., Pearlman, K., & Shane,
G. S. (1979). Further tests of the Schmidt-Hunter
Bayesian validity generalization. Personnel Psychology,
32, 257-281.
Stone, E. F. (1976). Moderating eifect of work-related
values in the job scope-job satisfaction relationship.
Organizational Behavior and Human Performance, 15,
147-167.
Walton, R. E. (1972). How to counter alientation in the
plant. Harvard Business Review, 50, 70-81.
Wanous, J. P. (1974). Individual differences and reactions
to job characteristics. Journal of Applied Psychology,
59, 616-622.
White, J. K. (1978). Individual differences in job qualityworker response relationship: Review, integration, and
comments. Academy of Management Review, 3, 267280.

JOB CHARACTERISTICS-JOB SATISFACTION RELATION

289

Appendix
Studies Included in Meta-Analysis
Aldag, R. J., & Brief, A. P. (1975). Impact of individual
differences on employee affective responses to task
characteristics. Journal of Business Research, 3, 311321.
Armenakis, A. A., Field, H. S., Jr., Holley, W. H., Jr.,
Bedeian, A. G., & Ledbetter, B., Jr. (1977). Human
resource consideration in textile work redesign. Human
Relations, 30, 1147-1156.
Arnold, H. J., & House, R. J. (1980). Methodological
and substantive extensions to the job characteristics
model of motivation. Organizational Behavior and
Human Performance, 25, 161-183.
Brief, A. P., & Aldag, R. J. (1975). Employee reactions
to job characteristics: A constructive replication. Journal
of Applied Psychology, 60, 182-196.
Brief, A. P., & Aldag, R. J. (1978). The job characteristics
inventory: An examination. Academy of Management
Journal, 21, 659-670.
Caldwell, D. F., & O'Reilly, III, C. A. (1982). Task
perceptions and job satisfaction: A question of causality.
Journal of Applied Psychology, 67, 361-369.
Dunham, R. B. (1977). Reactions to job characteristics:
Moderating effects of the organization. Academy of
Management Journal, 20, 42-65.
Evans, M. G., Kiggundu, M. N., & House, R. J. (1979).
A partial test and extension of the job characteristics
model of motivation. Organizational Behavior and
Human Performance, 24, 354-381.
Griffin, R. W. (1981). A longitudinal investigation of task
characteristics relationships. Academy of Management
Journal, 24, 99-113.
Griffin, R. W. (1982). Perceived task characteristics and
employee productivity and satisfaction. Human Relations, 35, 927-938.
Hackman, J. R., & Lawler, E. E. (1971). Employee
reactions to job characteristics. Journal of Applied
Psychology, 55, 259-286.
Hackman, J. R., & Oldham, G. R. (1976). Motivation
through the design of work: Test of a theory. Organizational Behavior and Human Performance, 16, 250279.
Katz, R. (1978a). The influence of job longevity on
employee reactions to task characteristics. Human
Relations, 31, 703-726.
Katz, R. (1978b). Job longevity as a situational factor in
job satisfaction. Administrative Science Quarterly, 23(2),
204-223.
Katz, R., & Van Maanen, J. (1977). The loci of work
satisfaction: Job, interaction, and policy. Human Relations, 30, 469-486.
Kiggundu, M. N. (1980). An empirical test of the theory

of job design using multiple job ratings. Human Relations. 33, 339-351.
O'Reilly, C. A., & Caldwell, D. (1979). Informational
influence as a determinant of task characteristics and
job satisfaction. Journal of Applied Psychology, 64,
157-165.
O'Reilly, C. A., Parletee, G. N., & Bloom, J. R. (1980).
Perceptual measures of task characteristics: The biasing
effects of differing frames of reference and job attitudes.
Academy of Management Journal, 23, 118-131.
Orpen, C. (1979). The effects of job enrichment on
employee satisfaction, motivation, involvement, and
performance: A field experiment. Human Relations,
32, 189-217.
Pokorney, J., Gilmore, D., & Beehr, T. (1980). Job
diagnostic survey dimensions: Moderating effect of
growth needs and correspondence with dimensions of
job rating form. Organizational Behavior and Human
Performance, 26, 222-237.
Rousseau, D. (1977). Technological differences in job
characteristics, employee satisfaction, and motivation:
A synthesis of job design research and sociotechnical
systems theory. Organizational Behavior and Human
Performance, 19, 18-42.
Rousseau, D. (1978). Characteristics of departments,
positions, and industries: Contexts for attitudes and
behavior. Administrative Science Quarterly, 23, 521540.
Schmitt, N., Coyle, B., White, J. K., & Rauschenberger,
J. (1978). Background needs, job perceptions, and job
satisfaction: A causal model. Personnel Psychology, 31,
889-901.
Sims, H. P., & Szilagyi, A. D. (1976). Job characteristics
relationships: Individual and structural moderators.
Organizational Behavior and Human Performance, 17,
211-230.
Umstot, D., Bell, C. H., & Mitchell, T. R. (1978). Effects
of job enrichment and task goals on satisfaction and
productivity: Implications for job design. Journal of
Applied Psychology, 61, 379-394.
Walsh, J. I., Taber, T. D., & Beehr, T. A. (1980). An
integrated model of perceived job characteristics. Organizational Behavior and Human Performance, 25,
252-267.
Wanous, J. P. (1974). Individual differences and reactions
to job characteristics. Journal of Applied Psychology,
59, 616-622.
Received May 25, 1984
Revision received July 30, 1984