You are on page 1of 9

Copyright 1982 by the American Psychological Association, Inc.

0021-9010/82/6703-0341S00.75

Journal of Applied Psychology


1982, Vol. 67, No. 3, 341-349

Measurement of Job and Work Involvement


Rabindra N. Kanungo
Faculty of Management, McGill University, Montreal, Quebec, Canada
In view of the recent distinction between job and work involvement, this study
developed separate measures of the two constructs using three different techniques: semantic differential, questionnaire, and graphic. The conceptual basis
of the two constructs and the reasons for developing new measures of the constructs are discussed. Data collected from a heterogeneous sample of 703 employees are analyzed to establish reliability, construct validity, and criterionrelated validity of each measure. Relative effectiveness of the three techniques
used to measure the constructs are examined. Results reveal that questionnaire
and graphic measures pass the tests of reliability and validity. Semantic differential measures, however, have questionable validity for measuring work involvement. Possible uses of these new measures in future research are suggested.

Past psychological research (Lodahl &


Kejner, 1965; Rabinowitz & Hall, 1977;
Saleh & Hosek, 1976) in the area of job
involvement is fraught with problems of conceptual ambiguities and measurement inadequacies. The major source of conceptual
ambiguity lies in the use of the construct
"job involvement" that carries excess meaning. Consequently, the techniques developed
to measure the construct suffer from the
problems of construct validity. Without adequate construct validity, inferences based
on the data on job involvement provided by
existing instruments are often misleading
and difficult to interpret.
The excess meanings of the job involvement construct can be identified in four different ways. First, past conceptualizations
of the construct have confused the issue of
job involvement with the issue of intrinsic
motivation on the job (Gorn & Kanungo,
1980; Kanungo, 1981). The most widely
used measure of job involvement, developed
by Lodahl and Kejner (1965), combines
items representing the two issues. Some
items, such as, "I live, eat and breathe my

This study was supported by a grant from the Formation des Chercheures et d'Action Concertee, Government of Quebec. I wish to thank Rajan Natarajan
for his assistance in data collection and analysis.
Requests for reprints should be sent to Rabindra N.
Kanungo, Faculty of Management, Samuel Bronfman
Building, McGill University, 1001 Sherbrooke Street
West, Montreal, Quebec, Canada H3A 1G5.

341

job," represent a person's psychological


identification with the job. Other items, such
as, "sometimes I'd like to kick myself for the
mistakes I make in my work," represent a
person's intrinsic motivation at work for fulfilling self-esteem needs.
Second, in dealing with the construct, researchers have confused the issue of identifying the antecedent conditions of job involvement with the issue of identifying the
state of job involvement and its subsequent
effects (Kanungo, 1979). Saleh and Hosek's
(1976) scale, for instance, contains three
categories of items that describe (a) presumed causal conditions of job involvement
(e.g., "how much chance do you get to do
things your own way?"), (b) presumed effects of job involvement (e.g., "I avoid taking
on extra duties and responsibilities in my
work"), and (c) the state of job involvement
itself (e.g., "the most important things I do
are involved with my job"). A third way in
which the construct carried extra meaning
can be seen in the description of job involvement as both a cognitive and a positive
emotional state of the individual. Lodahl and
Kejner's (1965) scale contains items that
represent these two meanings. Items such as,
"the major satisfaction in my life comes
from my job" and "the most important
things that happen to me involve my work,"
are descriptions of affective and cognitive
states, respectively.
Finally, earlier conceptualizations of job
involvement have failed to distinguish two

342

RABINDRA N. KANUNGO

different contexts in which an individual can


show personal involvement (Kanungo, 1981).
The two contexts are (a) specific or particular job context and (b) generalized work
context. Involvement in a specific job is not
the same as involvement with work in general. The former is a belief descriptive of the
present job and tends to be a function of how
much the job can satisfy one's present needs.
But involvement with work in general or the
centrality of work in one's life is a normative
belief about the value of work in one's life,
and it is more a function of one's past cultural conditioning or socialization. Thus, job
involvement is a descriptive belief that is
contemporaneously caused whereas work involvement is a normative belief that is historically caused.
Job involvement as a specific belief regarding one's relationship with one's present
job is also different from organizational commitment (Porter, Steers, Mowday, & Boulian, 1974), which refers to a general attitude toward an organization as a whole.
Work involvement should also be distinguished from the Protestant Ethic. Belief in
the centrality of work may result from Protestant-Ethic-type socialization, but the two
are not identical. Protestant Ethic may not
even be a necessary condition for work involvement to develop. It is conceivable that
work involvement may result from socialization that is not of the Protestant Ethic type.
Researchers have confused work involvement with the Protestant Ethic in the same
manner as they have confused job involvement with intrinsic motivation. A lack of
clear conceptual distinction between job and
work involvement is reflected in the previously developed scales of both job involvement (Lodahl & Kejner, 1965; Saleh &
Hosek, 1976) and work values (Blood, 1969).
These scales have used the words job and
work interchangably, and one is not sure
whether the respondents view them synonymously. The validity and usefulness of a
conceptual distinction between involvement
in a particular job and identification with
work in general have been demonstrated recently by Gorn and Kanungo (1980).
In view of the above construct validity
problems that are associated with past research, Kanungo (1979) argued for a refor-

mulation of the construct of involvement


that eliminates the problems of excess meaning. According to such reformulation (Kanungo, 1979, 1981), involvement either in the
context of a particular job or with work in
general can be viewed as a cognitive or belief
state of psychological identification. An individual's psychological identification with
a particular job (or with work in general)
in turn depends on (a) the saliency of his or
her needs (both extrinsic and intrinsic) and
(b) the perceptions he or she has about the
need-satisfying potentialities of the job (or
work). Viewed in this way, job involvement
and work involvement cannot be measured
with the existing instruments (Blood, 1969;
Lodahl & Kejner, 1965; Saleh & Hosek,
1976). This necessitates the development of
valid and reliable new measures of job and
work involvement for use in future research.
The present study is an attempt in this direction.
Method
Item Constructions for Involvement Scales
For the purpose of obtaining distinct measures of specific job involvement and of general work involvement,
three different measurement formatsquestionnaire,
semantic differential (Osgood, Suci, & Tanncnbaum,
1957), and graphic techniqueswere used. Questionnaire items that directly reflected a cognitive state of
psychological identification were judged and compiled
by 10 graduate students after a thorough search of the
existing measures of involvement and alienation in both
the psychological and sociological literature (e.g., Blauner, 1964; Clark, 1959; Dubin, 1956; Lodahl & Kejner,
1965; Saleh & Hosek, 1976; Wollack, Goodale, Wijting,
& Smith 1971). There was complete agreement by the
10 judges on 12 items for inclusion in the Job Involvement Questionnaire (JIQ) and on 9 items for inclusion
in the Work Involvement Questionnaire (WIQ). For the
JIQ and WIQ items, 6-point agree-disagree response
formats were used. Subsequent items analyses resulted
in dropping 2 items from the JIQ and 3 items from the
WIQ scales because of their low interitem and itemtotal correlations. Thus the final scales contained 10 JIQ
items (i.e., "The most important things that happen to
me involve my present job"; "To me, my job is only a
small part of who I am"; "I am very much involved
personally in my job"; "I live, eat and breathe my job";
"Most of my interests are centered around my job";
"I have very strong ties with my present job which would
be very difficult to break"; "Usually I feel detached from
my job"; "Most of my personal life goals are job-oriented"; "I consider my job to be very central to my
existence"; "I like to be absorbed in my job most of the
time"). In addition, there were 6 WIQ items (i.e., "The

MEASUREMENT OF JOB AND WORK INVOLVEMENT


most important things that happen in life involve work";
"Work is something people should get involved in most
of the time"; "Work should be only a small part of one's
life"; "Work should be considered central to life"; "In
my view, an individual's personal life goals should be
work-oriented"; "Life is worth living only when people
get absorbed in work").
Another six graduate students, using available literature and dictionaries for synonyms and antonyms,
made an extensive search for key words that clearly
reflected the notion of psychological identification. This
process yielded 11 bipolar items on which all the six
judges agreed. These items with a 7-point response format were used to construct Job Involvement Semantic
Differential (JISD) and Work Involvement Semantic
Differential (WISD) scales. Three items were rejected
from each scale on the basis of interitem and item-total
correlations. Thus each scale contained 8 items (involving-noninvolving; important-unimportant; fundamental-trivial; essential-nonessential; identifled-not identified; attached-detached; integrated-nonintegrated;
united-disunited). Finally, three graphic items representing the notion of psychological identification were
prepared for use in each of the job and work contexts.
Two of these items were finally selected after item analyses. In one item, two circles representing self and job
or work were presented with varying degrees of overlap
(no overlap to complete overlap). In the other item, a
human figure (representing self) and an office desk (representing job or work) were presented with varying distances between them. These two items formed the Job
Involvement Graphic (JIG) scale in the job context and
the Work Involvement Graphic (WIG) scale in the work
context. Both JIG and WIG items used a 7-point response format.

343

in three different universities in Montreal. These employees belonged to various industrial and governmental
organizations in and around Montreal. The respondents
were told that participation in the study was optional
and that they could be assured of the confidentiality of
the data. The questionnaire was completed during the
class hour in groups of varying sizes ranging from 40
to 100. The final count revealed that 703 completed
questionnaires (184 in French and 519 in English) were
returned.
A parallel study was conducted in two of the universities to establish the test-retest reliabilities of the measures included in the questionnaire. One evening extension course (with approximately 50-55 full-time
employees enrolled in it) that was offered in each of the
universities was used for this purpose. The questionnaire
was administered twice, 3 weeks apart. The respondents
were asked to put their identification numbers on the
questionnaire each time they were tested. Matching of
identification members revealed that data from 63 repondents could be used in the test-retest analysis.

Results
Demographic Data

Both the samples in the original and in the


test-retest study were heterogeneous in composition. In the original sample, employees
belonging to public and private sector organizations were equally represented. Almost half of the employees came from large
organizations (with more than 700 employees) and the other half came from small- or
medium-size organizations. Fifty-seven perDesign of the Questionnaire
cent of the respondents were male and 43%
1
A questionnaire containing three parts was designed
were
female, with a mean age of 28 years
for the purpose of testing the reliability and validity of
the newly constructed job and work involvement scales. (SD = 6.66) for the total sample. There were
Part 1 of the questionnaire contained JISD, JIQ, and 37% French Canadian and 41% English
JIG scale items. In addition, this part included two other Canadian subjects, and the remaining 22%
instruments. One instrument measured the perceived belonged to other ethnic groups. Forty perimportance of 15 job outcomes by asking the respondents' satisfaction with the same 15 job outcomes and cent of the respondents were married, and
overall satisfaction with their present job. The validity 60% were single. Their education levels
and reliability of these two instruments were established ranged from high school to advanced gradin earlier studies (e.g., Corn & Kanungo, 1980; Kan- uate degrees, and their income levels ranged
ungo, Gorn, & Dauderis, 1976), and the instruments from less than $10,000 to more than $40,000
were used to test the criterion-related concurrent validity of the newly developed involvement scales. Part 2 per year. Almost half the sample had orof the questionnaire contained the three work involve- ganizational tenure of 2 to 5 years. Of the
ment scales (WISD, WIQ, and WIG). Part 3 of the other half of the sample, approximately 20%
questionnaire was designed to determine the demo- had less than 2 years, and 30% had more
graphic characteristics of the respondents.
than 5 years of organizational experience.
The test-retest sample closely resembled the
Subjects and Procedure
original sample in its composition.
The questionnaire was written in both French and
English following the translation-retranslation procedure (Brislin, Lonner, & Thorndike, 1973) and was
administered to 900 full-time French- and Englishspeaking employees enrolled in evening extension courses

The complete questionnaire is available upon request from the author.

344

RABINDRA N. KANUNGO

Table 1
Reliability Coefficients for Involvement and
Job Satisfaction Scales

Dimensionality of the scales. The scores


on job and work involvement items were factor analyzed separately for each of the three
methods (semantic differential, questionCoefficient
naire, and graphic). The principal-compoInternal
Testnent analysis followed by a varimax rotation
consistency
retest
Scale
was used to arrive at factor solutions. Each
Job involvement
analysis yielded two clear interpretable factors of Job Involvement and Work Involve.74
JISD
.81
.85
.87
ment. For semantic differential items, the
JIQ
.82
JIG
.70
first factor loaded highly on WISD items
(item loadings ranged from .64 to .81 for
Work involvement
WISD and from .01 to .15 for JISD), and
,83
.78
WISD
the second factor loaded highly on JISD
WIQ
.75
.67
items (item loadings ranged from ,42 to .81
.68
.67
WIG
for JISD and from .01 to .15 WISD). The
Job satisfaction
two factors, with eigenvalues of 5.64 and
15-item scale
.89
.73
3.40, explained 56.5% of the total variance
Overall job satisfaction
(and 93.3% of common variance). For ques(single item)

.87
tionnaire items, the first factor reflected job
Note. Internal consistency (Cronbach a) coefficients are involvement (item loadings ranged from .44
based on data from 703 respondents. Test-retest coef- to .77 for JIQ and from .07 to . 17 for WIQ)
ficients are based on data from 63 respondents. JISD =
and the second factor reflected work involveJob Involvement Semantic Differential scale; JIQ = Job
ment
(item loadings ranged from .40 to .73
Involvement Questionnaire; JIG = Job Involvement
Graphic scale; WISD = Work Involvement Semantic for WIQ and from .00 to .31 for JIQ). The
Differential; WIQ = Work Involvement Questionnaire; two factors, with eigenvalues of 5.15 and
WIG = Work Involvement Graphic scale.
2.39, explained 47.2% of the total variance
(and 93.6% of the common variance). Finally, for the graphic items, again Job InEmpirical Properties of the Involvement
volvement
and Work Involvement emerged
Scales
as the first (item loadings were .95 to .70 for
Item analyses. The 8 items included in JIG and .09 and .19 for WIG) and the secthe JISD and WISD scales had median ond (item loadings were ,70 and .89 for WIG
item-total correlations of .75 (range = .64- and .21 and .06 for JIG) factors, respec.82) and .74 (range = .71-.82), respectively. tively. The eigenvalues were 2.11 and 1.23,
The median item-total correlation for the 10 explaining 83.5% of the total variance (and
items in the JIQ scale was .68 (range = .59- 100% of the common variance). Following
.74). For the 6 items in the WIQ scale, the separate analyses, item scores from all six
median item-total correlation was .67 scales put together were again factor ana(range = .54-.74). The intercorrelations for lyzed. In spite of the introduction of method
items in the JIG and WIG scales were .70 variance, a two-factor solution clearly reand .68, respectively. The means and stan- vealed differential factor loadings on job and
dard deviations for each of the six involve- work involvement items. The first factor
ment scales were as follows: JISD: M = loadings ranged from .70 to .40 on job and
23.94, SD = 10.07; WISD: M = 20.30, SD from .14 to .03 on work involvement items.
= 8.28; JIG: M= 31.31, 57)= 10.61; WIQ: The second factor loadings ranged from .79
M= 20.70, SD= 5.97; JIG: M= 8.39; to .34 for work and from .25 to .00 for job
SD = 3.01; and WIG: M = 9.04, SD = 2.69. involvement items. The eigenvalues were
In the case of the JISD and WISD, lower 9.68 and 5.18, explaining 41.3% of the total
scores represented higher involvement. For variance. These results clearly suggest disall other scales higher scores represented tinctiveness and unidimensionality of job
higher involvement.
and work involvement constructs.

345

MEASUREMENT OF JOB AND WORK INVOLVEMENT

Reliabilities of the scales. The internal


consistency and test-retest reliabilities of the
six involvement scales and the job satisfaction measures are presented in Table 1. The
reliability coefficients ranged from .67 to .89,
suggesting that both reliability of repeated
measurements and of internal consistency of
items were adequate for these scales. The
correlation between the two job satisfaction
measures as parallel form tests was .78.
Convergent and discriminant validity of
involvement scales. Intercorrelations among
the six involvement scales are presented in
Table 2 in the form of a validational matrix
(Campbell & Fiske, 1959). From the validity diagonals (boxed correlations in Table 2)
it can be seen that all the correlations were
statistically significant (p < .01), suggesting
convergent validity of the scales. However,
the magnitude of the correlations suggest
that convergent validities of questionnaire
and graphic scales measuring job involvement (r = .80) and work involvement (r =
.69) were quite high. By comparison, semantic differential scales showed a moderate
to very weak relationship to other scales
measuring job and work involvement. The
JISD scale showed a moderate relationship
to both the JIQ (r = -.33) and the JIG
(r = -.44) scales. The WISD scale showed

a very weak relationship to both the WIQ


(r = -.12) and the WIG (r = -.24) scales.
Assessment of discriminant validities requires that monotrait-heteromethod values
(agreement between different ways of measuring the same trait) should exceed the heterotrait-heteromethod values (agreement
between different traits measured in different ways). Table 2 shows that every boxed
correlation representing a rnonotrait-heteromethod value is higher than the adjacent
noncircled correlation representing a heterotrait-heteromethod value.
A second but more stringent criterion for
the assessment of discriminant validity requires that monotrait-heteromethod values should exceed heterotrait-monomethod
(agreement between different traits measured the same way) values. This would indicate whether common trait variance is
greater than common method variance. This
criterion was satisfied in 67% of the cases.
A closer inspection of Table 2 reveals that
the semantic differential format did not meet
this criterion, particularly in measuring work
involvement. The validity of the WISD scale
is questionable because the correlations between the WISD and the WIQ and WIG
involvement measures (r& = -.12 and -.24,
respectively) did not exceed the correlations

Table 2
Multitrait-Muhimethod Matrix for Job and Work Involvement Scales
Semantic
differential
Scale

JISD

WISD

Questionnaire
JIQ

WIQ

Graphic
JIG

WIG

Semantic differential
^
JISD
^\
WISD
1 .28*\^
Questionnaire
JIQ
[~-.33*|
-.08
.29*""-\
WIQ
.01
| -.12* |
Graphic
JIG
| -.44* |
-.09
WIG
.02
Ljl-24*J
.33*
[~69*n
L-36*\^
Note. N = 703. Correlations enclosed in boxes represent validity diagonal or monotrait-heteromethod values;
correlations enclosed in triangles represent heterotrait-monomethod values. The remaining correlations represent
heterotrait-heteromethod values. Negative correlations are due to the reverse scoring of scales using semantic
differential format. JISD = Job Involvement Semantic Differential scale; WISD = Work Involvement Semantic
Differential scale. JIQ = Job Involvement Questionnaire; WIQ = Work Involvement Questionnaire; JIG = Job
Involvement Graphic scale; WIG = Work Involvement Graphic scale.

.so* i .21* r\.

346

RABINDRA N. KANUNOO

Table 3
Analysis of Variance of MultitraitMultimethod Matrix
Source

df

MS

VC

VC
index

Respondent (R)
R X Trait
R X Method

702
702
1, 404

2.94
.78
.48

4.45*
1.18*
.72

.33
.31
.19

.46
.44
.33

Error

1, 404

.66

,38

Note. VC = variance component.


*p < .01.

between the WISD and the JISD (/ = .28).


In the case of the JISD scale, on the other
hand, the validity criterion is met because
correlations between the JISD and the JIQ
and JIG scales (rs = -.33 and -.44, respectively) were higher than the correlation between the JISD and the WISD (r = .28).
However, because of the moderate relationship of the JISD with the other two job involvement measures, its use should be discouraged.
The convergent and discriminant validities of the questionnaire and graphic scales
seem to be adequate. In fact, if one removes
the two semantic differential scales (JISD
and WISD) from the matrix in Table 2, the
picture becomes clear. For the two job (JIQ
and JIG) and the two work (WIQ and WIG)
involvement scales, the monotrait-heteromethod correlations were .80 and .69, respectively. These were substantially higher
than the monomethod-heterotrait correlations of .36 and .29 and heteromethod-heterotrait correlations of .33 and .21, respectively.
The multitrait and multimethod matrix
presented in Table 2 was further reanalyzed
in a confirmatory fashion using the analysis
of variance (ANOVA) three-way classification model suggested by Kavanagh, Mackinney, and Wolins (1971). In this model, the
purpose of the analysis was to estimate three
effects: (a) respondent (R) effect or the degree to which the alternative methods and
traits yield similar involvement scores or
agreement within respondents; (b) Respondent X Trait (R X T) interaction effect or
the degree of rated discriminations on traits
by respondents; and (c) Respondent X

Method (R X M) interaction effect or the


degree of disagreement on methods by respondents. The three effects corresponded to
an estimate of overall convergent validity,
discriminant validity, and method bias, respectively. The results of the ANOVA are presented in Table 3. Both the main effect of
R and the R X T interaction effect were statistically significant (p < .01), suggesting
confirmatory evidence for convergent and
discriminant validities of the instruments.
Following the suggestion of Kavanagh et
al. (1971), the variance component and variance component indexes for each of the three
effects were also calculated to compare their
relative impacts in the study. The results are
also presented in Table 3. Although the R X
M interaction (method bias) effect was not
statistically significant, its variance component index showed that its relative impact
was not too small to ignore, presumably due
to the use of semantic differential formats.
Separate estimations of variance components and variance component indexes for
matrices with and without semantic differential scales are presented in Table 4. The
results clearly suggest that method bias and
error components were substantially reduced; convergence and discriminant validities were clearly enhanced when semantic
differential scales were eliminated from the
matrix.
Criterion-related concurrent validity of
the scales. The concurrent validity of the
job and work involvement scales was examined by testing three theoretical predicTable 4
Comparison of Methods Used to Measure
Involvement
Scale
SD and
questionnaire

SD and
graphic

Questionnaire
and graphic

Source

VC

VC
index

VC

VC
index

VC

VC
index

Respondent (R)
R X Trait
R X Method

.25
.18
.24

.32
.26
.32

.33
.29
.27

.45
.42
.40

.53
.44
.05

.71
.67
.19

Error

.51

.39

.21

Note. SD = semantic differential; VC = variance component.

MEASUREMENT OF JOB AND WORK INVOLVEMENT

347

tions derived from the motivational frame- Table 5


work proposed by Kanungo (1979, 1981). Correlation of Involvement Scales With Job
First, it has been suggested that involvement Satisfaction Measures
in one's present job stems primarily from the
Overall job
Job satisfaction
perception of need-satisfying potential of the
Scale
scale
satisfaction
job, whereas involvement with work in genJISD
-.27*
-.56*
eral is more a matter of past socialization.
WISD
-.OS
-.01
Thus it was expected that measures of job
/(700)
5.96*
12.79*
involvement compared to measure of work
JIQ
.57*
.43*
involvement would be more strongly assoWIQ
.12*
.04
ciated with measures of job satisfaction.
12.18*
((700)
9.64*
Second, it has been proposed (Gorn &
JIG
.65*
.55*
Kanungo, 1980; Kanungo, 1979, 1981) that
.24*
.06
WIG
job involvement is more a function of sat((700)
12.61*
13.94*
isfaction of salient needs on the job. Thus
it was expected that the job involvement Note. JISD = Job Involvement Semantic Differential
measures would be more strongly associated scale; WISD = Work Involvement Semantic Differenwith the satisfaction of salient rather than tial scale; JIQ = Job Involvement Questionnaire; WIQ =
Work Involvement Questionnaire; JIG = Job Involvenonsalient needs. This pattern of relation- ment
Graphic scale; WIG = Work Involvement Graphic
ships was not expected in the case of work scale.
involvement because work involvement is a * p < .01.
culturally conditioned normative belief and
is not directly dependent on present job sat- ment is a function of salient need satisfacisfaction.
tion, was tested in the following manner. The
Finally, it has been argued that employees precetved importance rankings of the 15 job
with salient extrinsic needs are as likely to factors were analyzed to determine the sabe involved in their jobs as employees with liency of needs of the respondents. For every
salient intrinsic needs, provided they have respondent, salient need was defined as the
equal levels of job satisfaction (Gorn & Kan- two job outcomes that were ranked by the
ungo, 1980). Thus, controlling for the levels respondent as 1st and 2nd in order of imof job satisfaction, one would not expect any portance. The nonsalient need was defined
difference between extrinsic and intrinsic as the two job outcomes that were ranked
oriented employees in their job involvement, by the respondent as 14th and 15th in imOn the same basis, however, it would be dif- portance. The respondent's satisfaction scores
ficult to predict how the two groups would on the two salient job outcomes were added
react to work involvement measures. As to represent a single score for salient need
mentioned before, work involvement is not satisfaction. Likewise, the respondent's satdependent on present job satisfaction.
isfaction scores on the two nonsalient job
It may be recalled that the questionnaire outcomes were added to represent a single
used in this study included a 15-item job score for nonsalient need satisfaction. Each
satisfaction scale and a single-item overall of the six involvement scale scores were then
job satisfaction index. In order to examine correlated with the salient and nonsalient
the relationship of job satisfaction with job need satisfaction scores. The results are preand work involvement, the six involvement sented in Table 6. As expected, the job inscales were correlated with the two measures volvement measures correlated more strongly
of satisfaction (results are shown in Table with salient than with nonsalient need sat5). Tests of difference between the depen- isfaction. This pattern, however, was not obdent correlations (McNemar, 1969, p. 158) served for the work involvement measures.
clearly supported the first prediction that job
The third prediction regarding the job insatisfaction measures have a stronger rela- volvement of intrinsic and extrinsic oriented
tionship to job involvement than to work in- respondents was tested by following a provolvement (see Table 5 for t values).
cedure suggested by Gorn and Kanungo
The second prediction, that job involve- (1980). A group of intrinsic (n = 76) and

348

RABINDRA N. KANUNGO

Table 6

Correlation of Involvement Scores With Salient


and Nonsalient Need Satisfaction
Scale

Salient
need
satisfaction

Nonsalient
need
satisfaction

/(700)

JISD
JIQ
JIG
WISD
WIQ
WIG

-.24*
.49*
.56*
-.02
.10
.18*

-.12*
.31*
.32*
-.07
.10
.16*

3.06*
5.47*
7.21*
.05
.00
.51

Note. JISD = Job Involvement Semantic Differential


scale; JIQ = Job Involvement Questionnaire; JIG = Job
Involvement Graphic scale; WISD = Work Involvement
Semantic Differential scale; WIQ = Work Involvement
Questionnaire; WIG = Work Involvement Graphic
scale.
* p < .01.

a group of extrinsic ( = 42) respondents


were chosen on the basis of their perceived
importance of job outcome rankings. Each
member of the intrinsic group perceived the
two intrinsic outcomesinteresting nature
of work and responsibilityas being the two
most important job outcomes. Each member
of the extrinsic group, on the other hand,
perceived the two extrinsic outcomesmoney
and securityas the two most important job
outcomes. The choice of these outcomes to
represent intrinsic and extrinsic needs depended on two criteria. First, these outcomes
are clearly distinguishable as intrinsic and
extrinsic outcomes; second, within the intrinsic and extrinsic categories, these outcomes were cited most frequently in the
overall sample as the first or second ranking
job outcome.
Analysis of covariance was performed separately on scores obtained from each of the
six involvement scales, treating job satisfaction score (on the 15-item scale) as the covariate. The results revealed nonsignificant
F values (p > .05) in each case. This confirms the expectation that when controlling
for job satisfaction, job involvement of intrinsic and extrinsic individuals will not
differ.
Discussion
This study lends considerable support for
the conceptual distinction between job and

work involvement proposed by previous researchers (Gorn & Kanungo, 1980; Wollack
et al, 1971) and provides refinements in the
definition and measurement of involvement
in the two contexts. The results reveal that
all three job involvement scales and two of
the work involvement scales (WIQ and
WIG) have satisfactory psychometric properties. The scales have reasonable levels of
internal consistency and test-retest reliability. They seem to pass the tests of unidimensionality and of convergent and discriminant validity. The tests of criterion-related
concurrent validity of these measures also
add to their strength.
This study explored the use of three different formats for measuring job and
work involvement. Previous researchers have
mainly used the questionnaire format, but
for cross-cultural and comparative research
use of other formats such as graphic or pictorial techniques may be more useful. Results of this study show that the two graphic
scales (JIG and WIG) correlate highly with
their respective questionnaire scales (JIQ
and WIG), suggesting that the former can
easily act as substitute for the latter. For the
comprehension of the construct, graphic
scales (as opposed to questionnaires) demand very little linguistic competence of the
respondent. Hence, they might be more useful in cross-cultural and comparative research on involvement. They can also be
more effective when administered to less educated samples, or when time considerations
do not allow administration of longer questionnaires.
In contrast to the graphic and questionnaire formats, the use of the semantic differential format, particularly the WISD
scale, seems to have questionable validity.
Posttest interviews of some respondents revealed that they found it difficult to relate
to the abstract 7-point scales using words
such as fundamental-trivial and essentialnonessential. Evaluating their present jobs
in terms of these scales was relatively less
difficult than evaluating the generalized notion of work. This suggests that the semantic
differential format should be used with caution even when measuring involvement in the
present job context. Perhaps its usage should
be limited to only highly educated samples

MEASUREMENT OF JOB AND WORK INVOLVEMENT

and in very specific contexts (Edwards &


Waters, 1980).
The new scales for measuring job and
work involvement can be used in future research to achieve several objectives. First,
studies that aim at exploring the nature of
job and work involvement within organizations and at identifying the antecedent and
consequent conditions can use these instruments. Second, the instruments can be used
in studies that attempt to relate alienation
and involvement in different spheres of life,
such as work, family, and community. Third,
tests of theoretical predictions derived from
existing formulations on alienation and involvement (e.g., Kanungo, 1979) can be conducted more effectively with the use of these
scales. Finally, the use of these scales can
establish more meaningfully the cross-cultural validity and generalizability of findings
related to job and work involvement.
References
Blauner, R. Alienation and freedom: The factory
worker and his industry, Chicago: University of Chicago Press, 1964.
Blood, M. R. Work values and job satisfaction. Journal
of Applied Psychology, 1969, 53, 456-459.
Brislin, R. W., Lonner, W. J., & Thorndike, R. M.
Cross-cultural research methodology. New York:
Wiley, 1973.
Campbell, D. T., & Fiske, D. W. Convergent and discriminant validation by the multitrait-multtmethod
matrix. Psychological Bulletin, 1959, 16, 81-105.
Clark, J. P. Measuring alienation within a social system.
American Sociological Review, 1959, 24, 849-852.
Dubin, R. Industrial workers' worlds; A study of the
central life interests of industrial workers. Social
Problems, 1956,3, 131-142.

349

Edwards, J. E., & Waters, L. K. Academic job involvement: Multiple measures and their correlates. Psychological Reports, 1980, 47, 1263-1266.
Gorn, G. J., & Kanungo, R. N. Job involvement and
motivation: Are intrinsically motivated managers
more job involved? Organizational Behavior and
Human Performance, 1980, 26, 265-277.
Kanungo, R. N. The concept of alienation and involvement revisited. Psychological Bulletin, 1979,56", 119138.
Kanungo, R. N. Work alienation and involvement:
Problems and prospects. International Review of Applied Psychology, 1981, 30, 1-15.
Kanungo, R. N., Gorn, G. J., & Dauderis, J. J. Motivational orientation of Canadian anglophone and
francophone managers. Canadian Journal of Behavioral Science, 1916,8, 107-121.
Kavanagh, M. J., Mackinney, A. C., & Wolins, L, Issues in managerial performance: Multitrait-multimethod analysis of ratings. Psychological Bulletin,
1971, 75, 34-49.
Lodahl, T. M., & Kejner, M. The definition and measurement of job involvement. Journal of Applied Psychology, 1965, 49, 24-33.
McNemar, Q. Psychological statistics. New York:
Wiley, 1969.
Osgood, C. E., Suci, G. J., & Tannenbaum, P. H. The
measurement of meaning. Urbana: University of Illinois Press, 1957.
Porter, L. W., Steers, R. M., Mowday, R. T., & Boulian,
P. V. Organizational commitment, job satisfaction,
and turnover among psychiatric technicians. Journal
of Applied Psychology, 1974, 59, 603-609.
Rabinowitz, S., & Hall, D. T. Organizational research
on job involvement. Psychological Bulletin, 1977, 84,
265-288.
Saleh, S. D., & Hosek, J. Job involvement: Concepts
and measurements, Academy of Management Journal, 1976, 19, 213-224.
Wollack, S., Goodale, J. G., Wijting, J. P., & Smith,
P. C, Development of the survey of work values. Journal of Applied Psychology, 1971, 55, 331-338.
Received July 27, 1981