Affective, Continuance, and Normative Commitment To The Organization: An Examination of Construct Validity
Affective, Continuance, and Normative Commitment To The Organization: An Examination of Construct Validity
Affective, Continuance, and Normative Commitment To The Organization: An Examination of Construct Validity
Within the past few years, several studies have used the Affective, Continuance,
and Normative Commitment Scales (Allen & Meyer, 1990; Meyer & Allen, 1984,
1991) to assess organizational commitment. The purpose of this paper is to review
and evaluate the body of evidence relevant to the construct validity of these measures.
Although some empirical questions remain at issue, the overall results strongly support
the continued use of the scales in substantive research. q 1996 Academic Press, Inc.
This research was supported by research grants from the Social Sciences and Humanities
Research Council. We thank Meridith Black for assistance in data collection and analysis, Wendy
Bichard for preparing the tables, and two anonymous reviewers for their comments on an earlier
version of the article. Reprint requests should be addressed to Natalie J. Allen, Department of
Psychology, The University of Western Ontario, London, Ontario N6A 5C2, Canada. fax: (519)
661-3961. e-mail: ALLEN@sscl.uwo.ca.
252
0001-8791/96 $18.00
Copyright q 1996 by Academic Press, Inc.
All rights of reproduction in any form reserved.
1991; Morrow, 1993; O’Reilly & Chatman, 1986). Not surprisingly, some of
the conceptual changes in the study of organizational commitment have been
accompanied by efforts to refine the measurement of the commitment con-
struct.
One program of research in which conceptual and measurement work have
gone hand in hand has resulted in a three-component view of commitment
(Allen & Meyer, 1990; Meyer & Allen, 1991). According to this perspective,
the psychological linkage between employees and their organizations can take
three quite distinct forms, each of which is given a distinguishing label.
Affective commitment refers to identification with, involvement in, and emo-
tional attachment to the organization. Thus, employees with strong affective
commitment remain with the organization because they want to do so. Contin-
uance commitment refers to commitment based on the employee’s recognition
of the costs associated with leaving the organization. Employees with strong
continuance commitment, then, remain with the organization because they
have to do so. Finally, normative commitment refers to commitment based
on a sense of obligation to the organization. Employees with strong normative
commitment remain because they feel they ought to do so. As can be seen,
all three components of commitment have straightforward implications for
staying with (or leaving) an organization, but beyond that, are conceptually
quite different. Given these differences, commitment is most meaningfully
assessed using three separate measures. Two of the commitment measures
developed on the basis of this model were first used in published research
by Meyer and Allen (1984), and the third, by Allen and Meyer (1990). Since
then, the measures, referred to as the Affective, Continuance, and Normative
Commitment Scales, have been administered in several studies. Consequently,
there now exists a considerable body of evidence regarding the psychometric
properties of the measures and their relations with other organizational and
person variables.
In this paper, we examine the construct validity of the three commitment
scales by reviewing research in which they have been used. Following
Schwab’s (1980) recommendations concerning the validation of measures
used in organizational research, we begin by examining the reliability of the
three scales. This is followed by an examination of their factor structures.
Next we review the evidence concerning relations between the commitment
scales and other variables that are hypothesized, in the three-component model
(Meyer & Allen, 1991), to be antecedents and consequences of commitment.
These hypothesized relations provide a ‘‘nomological net’’ (Cronbach &
Meehl, 1955) to be used in the validation process. If the overall pattern of
relations obtained empirically is consistent with that proposed, it provides
evidence that the commitment scales are measuring the constructs as intended.
MEASURES OF AFFECTIVE, CONTINUANCE, AND NORMATIVE
COMMITMENT
The development of the Affective, Continuance, and Normative Commit-
ment Scales (ACS, CCS, and NCS, respectively) was based on the scale
Estimates of Reliability
Internal Consistency
Typically, the internal consistency of the measures has been estimated
using coefficient alpha. Reliabilities associated with each sample are shown
1
Shorter six-item versions of the measures have recently been developed (Meyer, Allen, &
Smith, 1993); these will be discussed later in the paper.
2
A description of all samples is available from the authors.
in Table 1. Median reliabilities, across both versions of the ACS, CCS, and
NCS are .85, .79, and .73, respectively, and with very few exceptions, all
reliability estimates exceed .70.
Test–Retest Reliability
Although most studies using the commitment measures have been cross-
sectional, some longitudinal data are available. Available test–retest reliabilities
are shown in Table 2. As can been seen, the duration between administrations
ranged from 7 weeks to 11 months and, in all but one study (Blau, Paul, & St.
John, 1993), longitudinal data were collected from organizational newcomers.
All the test–retest reliabilities are within an acceptable range and consistent
with those reported for comparable measures (e.g., the Organizational Com-
mitment Questionnaire; see Mowday, Steers, & Porter, 1979). It will be
noted, however, that the lowest reliabilities are those involving commitment
measures taken on employees’ first day at the organization (Vandenberg &
Self, 1993). Employees may find it difficult to respond meaningfully to com-
mitment items when they have almost no experience with the organization.
Also noteworthy is that the higher reliabilities reported are those based on
data obtained during the latter part of the newcomers’ first year in the organi-
zation. This pattern is consistent with that found with other measures of
organizational attitudes (Vandenberg & Self, 1993) and may reflect something
about the stabilization of attitude that occurs as newcomers gain experience
with their organizations (Mowday, Porter, & Steers, 1982).
Factor Analytic Evidence
The factor structure of the commitment measures has been examined in
several studies using both exploratory and confirmatory analyses. Some analy-
ses included items from all three measures, while others included only ACS
and/or CCS items. Some analyses also included other work-related measures
along with organizational commitment. Space limitations do not allow the
full reporting of each analysis. Instead, we outline the overall pattern of
findings and address three specific issues: (1) relations among the commitment
measures and related measures, (2) the factor structure of the continuance
commitment measure, and (3) the stability of the factor structures across time.
Relations among the ACS, CCS, NCS, and Related Measures
Results of the exploratory analyses indicate that the items making up the
three commitment measures load on separate factors. Reilly and Orsak (1991),
for example, conducted their analysis on responses to the 24 commitment
items and the 7 items from Blau’s (1988) career commitment measure. They
reported the presence of four clearly defined factors, each representing one
of the four measures included in the analysis. Other exploratory factor analy-
ses have shown that: (a) the ACS items are distinct from related measures
assessing career, job, and work value constructs (Blau et al., 1993), (b) ACS
items are distinct from CCS items (McGee & Ford, 1987), (c) ACS and NCS
TABLE 1
Internal Consistency Reliabilities for Affective, Continuance,
and Normative Commitment Scales
Commitment measurea
TABLE 1—Continued
Commitment measurea
Note. ACS, Affective Commitment Scale; CCS, Continuance Commitment Scales; NCS, Nor-
mative Commitment Scale.
a
The three studies indicated with an asterisk used the six-item versions of the commitment
scales. All others used the original eight-item scales.
b
Research conducted by Dunham et al. (1994) included nine separate samples; reprinted here
are the range of reliabilities across these samples.
items load on the same factor, but are distinct from CCS items (Cohen,
1993), and (d) ACS, CCS, and NCS items load appropriately on three factors
(Allen & Lee, 1993).
Consistent with the above, results of confirmatory factor analyses demon-
strated that the ACS and CCS (Meyer, Allen, & Gellatly, 1990; Moorman,
Niehoff, & Organ, 1993; Shore & Tetrick, 1991; Somers, 1993a), and the
ACS, CCS, and NCS (Dunham, Grube, & Castaneda, 1994; Hackett et al.,
1994; Lee, 1992; Meyer, Allen, & Smith, 1993) load on separate factors.
Further, Shore and Tetrick (1991) provided evidence for the distinction be-
tween the ACS, the CCS, job satisfaction, and perceived organizational sup-
port. As expected, the Organizational Commitment Questionnaire (OCQ) and
the ACS were highly correlated. A similar pattern was observed in those
samples to whom Dunham et al. (1994) administered the OCQ; the OCQ
converged with the ACS but was distinct from the CCS and NCS. Also,
consistent with prediction, Aven (1988) found that CCS items did not load
on the same factor as items from either the OCQ or the Hrebiniak and Alutto
(1972) commitment measure.3 Finally, in a recent study of organizational and
3
Although the Hrebiniak and Alutto (1972) measure has been described as a measure of cost-
based commitment, the evidence suggests that this is not the case (Meyer & Allen, 1984; Stebbins,
1970).
TABLE 2
Test–Retest Reliabilities for Commitment Measures
Reliabilities
occupational commitment, Meyer et al. (1993) reported that the best fit to the
data was provided by a six-factor solution involving the three organizational
commitment measures and three parallel occupational commitment measures.4
These results not only provide further evidence of the distinction among the
three forms of commitment but also that the measures are sensitive to the
particular foci in question.
Dimensionality of the Continuance Commitment Scale
The findings of the factor analytic studies are less clear with regard to
whether or not one of the measures, the CCS, represents a unidimensional
construct. The dimensionality of this measure was examined first by McGee
and Ford (1987) who reported results from two exploratory factor analyses
of the same data set. In the first analysis, two factors were specified and the
results supported an ACS/CCS distinction. In the second, the number of
factors to be extracted was not specified. Three of the four factors produced
in this analysis were interpretable and supported a distinction between the
ACS and two ‘‘dimensions’’ of the CCS, ‘‘the first based on perceptions that
few employment alternatives exist and the second on high personal sacrifice
associated with leaving the organization’’ (McGee & Ford, 1987, p. 640).
Since McGee and Ford’s (1987) impactful study, the dimensionality of the
CCS has received scrutiny using confirmatory factor analysis (Hackett et al.,
1994; Dunham et al., 1994; Meyer et al., 1990; Somers, 1993a). At issue
here is whether the CCS measures a unitary commitment construct or two
separable commitment constructs, one that develops from the employee’s
4
This study used the six-item versions of the commitment measures.
recognition that alternatives are few (CCS:LoAlt) and the other from a recog-
nition that the employee’s investments in the organization would be sacrificed
if he or she left (CCS:HiSac).
Meyer et al. (1990) used confirmatory factor analysis to compare several
models describing the CCS data taken from three independent samples. Al-
though the one-factor model provided a good fit to the data, the best fit was
provided by an oblique two-factor model (CCS:LoAlt and CCS:HiSac items
as separate factors). Interestingly, however, an orthogonal two-factor model
provided a poor fit, suggesting that the two continuance commitment facets
are not independent. (The correlation between the two CCS factors was .817.)
More recently, Somers (1993a) used both the ACS and CCS and compared
a two-factor model (ACS, CCS) and a three-factor model (ACS, CCS:LoAlt,
CCS:HiSac). Again, although the fit indices associated with the model treating
the CCS as unidimensional were acceptably high, fit improved when two
continuance commitment factors were specified. Using only the CCS items,
Dunham et al. (1994) conducted separate confirmatory factor analyses on data
from six samples (and a multigroup analysis). They found consistent evidence
favoring a two-factor oblique model over both a one-factor model and a two-
factor orthogonal model. Finally, Hackett et al. (1994) applied confirmatory
factor analyses to all three commitment measures and compared the fit of
one-, three-, and four-factor models. They found that a four-factor model
provided the best fit to the data, again supporting a two-dimensional CCS
structure.
Overall, therefore, a model hypothesizing a two-dimensional CCS structure
clearly provides a better fit to the data than does a unidimensional model.
Across all studies, however, this superiority is modest, and the two factors
are highly related.
Finally, as important as the issue of relative fit is the question of whether
the two ‘‘subscales’’ of the CCS correlate differently with variables of interest.
Apparently, they are differentially related to the ACS (McGee & Ford, 1987;
Meyer et al., 1990). This has also been examined with respect to ‘‘conse-
quence’’ variables including job performance (Meyer et al., 1989; Hackett et
al., 1994), absenteeism (Hackett et al., 1994), turnover intention (Allen &
Lee, 1993; Somers, 1993b), and quality of nonwork life (Allen & Lee, 1993).
In all cases, the correlations found between the variables in question and one
subscale paralleled those found with the other subscale and with the total
CCS. Thus, although there appears to be evidence for two strongly related
continuance commitment factors, the practical implications of treating the
two factors separately have yet to be demonstrated.
Stability of Factor Structures over Time
We turn now to confirmatory factor analytic evidence relevant to the inter-
pretation of change, over time, in the commitment measures. Change in the
mean scores of self-report attitude measures are often considered evidence
of true change in those attitudes (i.e., what Golembiewski, Billingsley, &
related factors. The practical implications of this, however, are not clear
and require further research. Finally, the conflicting evidence regarding the
temporal stability of the ACS and CCS suggests that researchers examining
change in commitment among newcomers should test and, if necessary, con-
trol for beta and gamma change. We turn now to an examination of links
between the commitment measures and measures of other constructs.
Commitment and Its Relations to Other Constructs: A Nomological Net
As noted earlier, the three-component model of organizational commitment
(Meyer & Allen, 1991) provides a nomological net (Cronbach & Meehl, 1955)
that can be used to evaluate further the construct validity of the commitment
measures. These measures have been used to address a variety of research
questions. Thus, in most cases, the number of studies examining a particular
relation is not yet sufficient for meta-analytic review. Our objective, therefore,
was to examine the pattern of findings across studies. Evidence for construct
validity is provided to the extent that correlations between the commitment
measures and other variables parallel those outlined in the model.
In this section of the paper, we outline the general predictions that are of
relevance to particular sets of variables and examine the pattern of correlations
between the commitment measures and: (a) other attitude measures, (b) work-
related characteristics, (c) turnover intentions and turnover, (d) other work-
related behavior, and (e) nonwork variables.
Relations with Other Attitude Measures
Although the various work attitude constructs that have been discussed in
the literature have different foci (e.g., the job, the organization, the occupa-
tion), most are affective in nature. Consequently, we would expect some
convergence between the measures used to assess these constructs and the
measure of affective commitment to the organization. Despite this, affective
measures with foci other than the organization should be empirically discern-
ible from the ACS, while those that focus on emotional attachment to the
organization, such as the Organizational Commitment Questionnaire, should
correlate strongly with the ACS. Though normative commitment is based on
obligation, not affect, it does appear to overlap somewhat with affective
commitment; thus, it is expected that NCS scores would be correlated mod-
estly with other affective measures (including the OCQ). Continuance com-
mitment, on the other hand, is considered to be affectively neutral: though
one may feel that the costs associated with leaving would be high, this, in
itself, does not necessarily generate positive or negative feelings toward the
domain in question. CCS scores, therefore, would be expected to share very
little variance with these other work attitude measures.
Shown in Table 3 are the correlations between the commitment measures
and other attitude measures. Several aspects of these data merit comment.
First, the strong relations between the OCQ and the ACS are entirely consis-
tent with expectation and provide evidence for convergent validity. Also
TABLE 3
Correlations between Commitment Measures and Other Attitude Measures
Commitment measure
ous research (e.g., Brooke, Russell, & Price, 1988; Mathieu, 1991), the magni-
tude of these correlations suggest that affective commitment to the organiza-
tion is related to, but distinguishable from, other work attitude variables.
Finally, recall the factor analytic results, described above; these provide clear
evidence that respondents distinguish between the ACS and commitment to
the career (Reilly & Orsak, 1991), career, job, and work values (Blau et al.,
1993), job satisfaction (Shore & Tetrick, 1991), and occupational commitment
(Meyer et al., 1993).
Third, as expected, the CCS and NCS correlate weakly with other attitude
measures; this provides further evidence of discriminant validity. Unfortu-
nately, however, there is nothing in these data to address the issue of conver-
gent validity. This is due, quite simply, to the fact that few comparable
measures exist. Although the Ritzer and Trice (1969) and Hrebiniak and
Alutto (1972) measures were purportedly developed to assess cost-based com-
mitment, their validity has been questioned (Meyer & Allen, 1984; Morrow,
1993; Stebbins, 1970). With respect to obligation-based (normative) commit-
ment, only one other such measure appears in the literature. To our knowledge,
this measure, developed by Wiener and Vardi (1980), has not been used in
conjunction with the NCS or the other two commitment measures.
Finally, two studies examined commitment and dispositional affect (Cro-
panzano, James & Konovsky, 1993; Reilly & Orsak, 1991). As would be
expected, the ACS was positively correlated with positive affect and nega-
tively correlated with negative affect. Of the four correlations involving the
CCS, however, only one was significant. This pattern is consistent with the
view of continuance commitment as an affectively neutral construct.
Relations with Work-Related Characteristics
Within the theoretical framework that guided the development of the com-
mitment measures, work experience variables are seen as ‘‘antecedents’’ of
commitment. Given that they describe such psychologically different orienta-
tions toward the organization, affective, continuance, and normative commit-
ment are expected to develop on the basis of quite different experiences
(Meyer & Allen, 1991). Specifically, affective commitment is expected to be
correlated with those work experiences in, and characteristics of, the organiza-
tion that make the employee feel ‘‘psychologically comfortable’’ (e.g., ap-
proachable managers, equitable treatment of employees) and that enhance his
or her sense of competence (e.g., challenging tasks, feedback). Continuance
commitment, on the other hand, purportedly develops on the basis of the
employee’s recognition of the investments he or she has made in the organiza-
tion (e.g., time and effort, pension contributions) and/or the lack of compara-
ble employment alternatives. In the case of normative commitment, experi-
ences within the specific organizational domain may be somewhat less influ-
ential than are other earlier experiences. Indeed, it is expected that normative
commitment develops on the basis of those socialization experiences in the
individual’s early life that encourage sustained commitment to one’s em-
Commitment measure
Organizational
support/leadership
Management receptiveness .48* 0.16* .20* Allen & Meyer (1990)
Organizational dependability .61* 0.04 .38* Allen & Meyer (1990)
Transformational leadership .39–.45* 0.03–.05.14*–.17* Bycio, Hackett, &
Allen (1995)a
Transactional leadership 0.22–.36* 0.06–.13* 0.07–.20*
Organizational dependability .56* 0.13* .47* Lee (1992)
Organizational support .64* 0.08 — Shore & Wayne
(1993)
Support from supervisors .43* 0.08 — Withey (1988)
Competence-related variables
Feedback .36* 0.18* .21* Allen & Meyer (1990)
Goal difficulty .56* 0.17* .25* Allen & Meyer (1990)
Job challenge .63* 0.14* .29* Allen & Meyer (1990)
Challenge .47* 0.13* .29* Lee (1992)
Feedback .38* 0.03 .29* Lee (1992)
Challenge .59* 0.16 .19* Meyer, Irving, & Allen
(1993)
Justice variables
Interactional .51* 0.10 — Gellatly (1995)
Procedural (lay-off) .08 — — Kelloway & Barling
(1992)
Distributive (lay-off) .20* — — Kelloway & Barling
(1992)
Procedural (drug-testing) .44* 0.09 — Konovsky &
Cropanzano (1991)
Distributive (pay) .39* 0.07 .00 Lynn (1992)
Procedural (pay) .52* 0.14* .25* Lynn (1992)
Procedural .50* .09* — Moorman et al. (1993)
Role-related variables
Role clarity .53* 0.11 .39* Allen & Meyer (1990)
Role ambiguity 0.48* .09 0.31* Lee (1992)
Employment alternatives
Perceived alternatives 0.13* 0.43* 0.08 Allen & Meyer (1990)
Perceived alternatives .09 0.28* 0.01 Lee (1992)
Availability of alternatives .03 .01 — Whitener & Walz
(unemployment rate) (1993)
Specificity/transferability
variables
Skills transferability .25* 0.20* .19* Allen & Meyer (1990)
Education transferability 0.02 0.12* .03 Lee (1992)
Skills transferability .13* 0.35* .08 Lee (1992)
Specificity of skills 0.07 .28* — Withey (1988)
a
Bycio et al. (1995) administered three measures of transformational leadership and two
measures of transactional leadership. Reported here are the ranges of correlations involving these
measures.
* p õ .05.
TABLE 5
Correlations between Commitment Measures and Turnover Variables
Commitment Measures
a
Somers (1993b) correlated commitment scales with intentions to remain for 1, 2, and 5 years.
All nine correlations were significant (p õ .05). The mean correlations across the three intention
measures are reported here.
* p õ .05.
sures to a large sample of nurses and collected voluntary turnover data approx-
imately 1 year later. Turnover was significantly related to the ACS and the
NCS, but not the CCS. Whitener and Walz (1993) administered the ACS and
CCS to a sample of bank tellers 1 year prior to obtaining voluntary turnover
data and reported significant negative correlations between turnover and both
commitment measures.
TABLE 6
Correlations between Commitment Measures and Performance Indicators
Commitment measure
Work performance
Self-report measures Allen & Smith (1987)
.25* 0.14* .07 Resourcefulness/
innovation
.12* 0.07 .16* Consideration for others
Lee (1992)
.46* 0.12* .32* Spontaneity
.27* 0.08 .20* Helping behavior
.29* .09 .23* Job performance
McDonald (1993)
.14* .00 0.12* Organizational citizenship
Meyer & Allen (1986)
.33* 0.13 .17 Extra-role behavior
Meyer, Allen, & Smith
(1993)
.23 0.04 .17* Voice
.41* 0.03 .40* Loyalty
0.38* .25* 0.23* Neglect
.10 0.06 .00 Helping others
.13* .08 .11* Use of time
.07 0.15* .14* Professional activity
Morrison (1994)a
.21*–.30* — .09–.22* Job breadth
.22*–.27* — .08–.24* Work behavior
Wahn (1993)
— .14* — Unethical behavior
Independent measures Hackett et al. (1994)
0.01 0.21* 0.16 Commendations
0.17 0.01 0.03 Complaints
0.15 0.00 0.18 Accidents/year
0.03 0.07 0.12 Incognito performance
rating
Konovsky & Cropanzano
(1991)
.28* 0.20* — Performance
Meyer et al. (1989)
.15 0.25* — Composite performance
23* 025* — Overall performance
23* 0.46* — Promotability
Moorman et al. (1993)a
.16*—.28* .05–.14* — Citizenship behavior
.16* .06 — In-role behavior
Shim & Steers (1994)
Sample 1
.31* 0.15* — Composite performance
.26* 0.17* — Overall performance
TABLE 6—Continued
Commitment measure
Sample 2
.02 0.12 — Composite performance
.06 0.10 — Overall performance
Shore & Wayne (1993)
.22* 0.20* — Altruism
.14* 0.20* — Compliance
Absenteeism
Self-report measures MacDonald (1993)
.00 .10 .08 Frequency (No. incidents)
0.01 .07 .04 Total days absent
Meyer et al. (1993)
0.13* 0.05 0.15* Voluntary absences
.03 .03 .08 Total days absent
Independent measures Gellatly (1995)
0.18* .09 — Frequency (No. incidents)
0.14* .04 — Total days absent
Hackett et al. (1994)
0.22* .04 0.09 Culpable absences
0.07 .13 .14 Nonculpable absences
Somers (1993b)
0.07 .01 .02 Frequency (No. incidents)
0.15* .05 0.07 Annexed absences
a
Research conducted by Morrison (1994) and Moorman et al. (1993) involved several compa-
rable measures; reported here are the ranges of correlations across measures.
* p õ .05.
This finding—that ACS scores are positively, and CCS scores negatively,
correlated with supervisory ratings—is interesting in light of the Shim and
Steers (1994) study. Only one of the two organizations they studied showed
this pattern. Follow-up interviews revealed that the organization in which
commitment was related to performance placed much more emphasis on
satisfying customers and continuous performance improvement. Shim and
Steers suggested that this emphasis may have provided, to employees, clear
directives about how best to express their commitment, something lacking in
the other organization. Although reasonable, this finding obviously requires
replication. Further, it is not clear why the observed management differences
between the organizations would moderate the links between continuance
commitment and performance.
Finally, Hackett et al. (1994) used a ‘‘trained incognito rater’’ to assess
the performance of bus drivers on a single occasion. This measure was not
correlated with any commitment measure, nor was the number of performance
complaints bus drivers had received since being hired. The number of com-
mendations drivers received over their tenure, however, was negatively related
to their CCS scores and, when the effects of control variables (age and tenure)
were removed, ACS scores were negatively correlated with the number of
accidents in which drivers had been involved.
Absenteeism: self-report measures. Meyer et al. (1993) reported significant,
albeit modest, negative correlations between self-reports of voluntary (avoid-
able) absenteeism and both affective and normative commitment. In neither
this study nor another (McDonald, 1993) was continuance commitment corre-
lated with self-reported absenteeism.
Absenteeism: independent measures. Absenteeism measures obtained from
actual personnel records have been examined in three recent commitment
studies. Significant negative correlations were reported between affective
commitment and total days absent over 12 months (Gellatly, 1995), frequency
of culpable absences over 5 years (Hackett et al., 1994), and frequency of
absences over 12 months in one study (Gellatly, 1995) but not another (Som-
ers, 1993b). Somers found that affective commitment correlated negatively
with the frequency, over 12 months, of annexed absences (absences on days
connected to a weekend or holiday). Neither continuance nor normative com-
mitment were significantly correlated with absenteeism in these studies.
some negative consequences, for how people perceive aspects of their lives
beyond those at work.
Finally, in a sample of working parents of young children, Jayne (1994)
found that employees who strongly identified with the ‘‘provider role’’ had
stronger continuance commitment than did those with weak identification
with this role. Affective commitment was unrelated to identification with the
provider role. This is consistent with the idea that continuance, but not af-
fective, commitment is based on perceived cost. Presumably, the perceived
costs associated with leaving ones’ organization would be greater for those
parents who felt they were the ‘‘provider’’ for their children.
Summary
In this above section, we presented wide-ranging evidence relevant to the
construct validity of the three commitment measures. In examining evidence
of this sort, one must be mindful of Schwab’s (1980) caution against overinter-
preting a single correlation and instead focus on patterns, across studies and
measures, that make theoretical sense. Fortunately, the breadth of samples
and variables examined allowed us to do this. The overall pattern of findings
produced by the research described here suggests that the three commitment
measures are distinguishable from other commonly used work attitude mea-
sures and relate to measures of ‘‘antecedent’’ and ‘‘consequence’’ variables
largely in accordance with theoretical prediction.
CONCLUSIONS
Within the study of work attitudes and behavior, the organizational
commitment construct has come to play an increasingly important role
(Morrow, 1993). Thus, it is critical that the construct validity of commonly
used measures of this construct are subjected to considerable scrutiny. The
purpose of this paper was to examine the substantial body of evidence
relevant to the construct validation of the Affective, Continuance, and
Normative Commitment Scales. These measures were developed on the
basis of a theoretical framework that integrated existing views of attitudinal
commitment and have been used in a wide variety of samples and settings.
Further, they have been critically examined, both empirically and concep-
tually, by a number of researchers (Cohen, 1993; Dunham et al., 1994;
Hackett et al., 1994; McGee & Ford, 1987; Vandenberg & Self, 1993).
Certainly, construct validation efforts have gone far beyond the ‘‘one-
time or limited term project’’ against which Schwab (1980) cautioned
organizational researchers and, instead, are represented by a large and
varied set of studies that provide much relevant data.
Taken together, the data reported here strongly suggest that the continued
use of the commitment measures in substantive research is justified. This
does not mean, of course, that all issues regarding the ACS, CCS, and NCS
are completely resolved or that continued scrutiny of the measures is not
warranted. Indeed, there are several issues that merit research attention.
One such issue involves the possibility that the Continuance Commitment
Scale actually consists of two subscales. Although, as indicated by Hackett
et al. (1994), ‘‘the preponderance of evidence is in support of a three-compo-
nent model of commitment’’ (p. 21), the factor analytic results cannot be
ignored. To be prudent, it is suggested that researchers first determine whether
the two sets of CCS items are differentially correlated with other variables
before reporting results based on the full CCS (Meyer et al., 1990). While
the evidence thus far suggests otherwise, there may be some antecedent or
consequence variables for which a different pattern exists.
The normative commitment construct also deserves considerably more at-
tention. Although affective and normative commitment have several common
correlates, factor analytic research has shown clearly that the latter is distin-
guishable from both affective and continuance commitment. Unfortunately,
few of the variables considered to be unique correlates of normative commit-
ment have been examined. This can be explained, in part, by the challenges
involved in collecting the relevant data. Hypothesized antecedents, for exam-
ple, include variables not typically examined in organizational research such
as one’s early familial and cultural socialization experiences. With respect to
consequences, it has been suggested (Meyer & Allen, 1991) that the unique
influence of normative commitment on work behavior may not be on the
quantity or quality of work performance, but on the ‘‘tone’’ with which work
is carried out. To our knowledge, no satisfactory measure of such a subtle
aspect of work behavior has been developed.
Another issue with respect to normative commitment involves its fairly
substantial correlation with affective commitment. In an effort to more clearly
tap into respondents’ own feelings of obligation to the organization, modifica-
tions were made recently to the NCS items. This work was also motivated
by a desire to shorten each of the three scales and to reduce the number of
negatively keyed items (cf. Magazine, Williams, & Williams, in press). The
resulting six-item versions of the measures were reported by Meyer et al.
(1993) and have been used in two subsequent studies. Interestingly, Meyer
and colleagues showed that self-reports of the ‘‘professional activity’’ (e.g.,
courses taken, involvement in professional associations) carried out by nurses
were positively related to the shorter NCS, negatively related to the CCS,
and unrelated to the ACS (see Table 6). While normative commitment to
one’s organization may carry with it the obligation to be involved, and remain
current, in one’s profession, it is unclear at this point whether this finding is
due to the use of a particularly appropriate NCS correlate or to the scale
revisions. In any case, the correlations between the revised versions of the
ACS and NCS parallel those found with the earlier versions, suggesting
that, although affective and normative commitment are clearly distinguishable
constructs (as demonstrated by numerous factor analyses), they may have
inherent psychological overlap. It simply may not be possible to feel a strong
obligation to an organization without also having (or developing) positive
emotional feelings for it. Beyond this, the revised scales appear to have
Carson, K. D., & Bedeian, A. G. (1994). Career commitment: Construction of a measure and
examination of its psychometric properties. Journal of Vocational Behavior, 44, 237–262.
Cohen, A. (1993). On the discriminant validity of the Meyer and Allen (1984) measure of
organizational commitment: How does it fit with the work commitment construct? In N. S.
Bruning (Ed.), Proceedings of the Annual Meeting of the Administrative Science Association
of Canada: Organizational Behaviour, 14, 82–91.
Cronbach, L. J., & Meehl, P. E. (1955). Construct validity in psychological tests. Psychological
Bulletin, 52, 281–302.
Cropanzano, R., James, K., & Konovsky, M. A. (1993). Dispositional affectivity as a predictor
of work attitudes and job performance. Journal of Organizational Behavior, 14, 595–606.
Dalton, D. R., Krackhardt, D. M., & Porter, L. W. (1981). Functional turnover: An empirical
assessment. Journal of Applied Psychology, 66, 716–721.
Dunham, R. B., Grube, J. A., & Castaneda, M. B. (1994). Organizational commitment: The
utility of an integrative definition. Journal of Applied Psychology, 79, 370–380.
Farrell, D. (1983). Exit, voice, loyalty, and neglect as responses to job dissatisfaction: A multidi-
mensional scaling study. Academy of Management Journal, 26, 596–607.
Finegan, J. E. (1994). [Organizational commitment in the petrochemical industry]. Unpublished
data.
Gellatly, I. R. (1995). Individual and group determinants of employee absenteeism: Test of a
causal model. Journal of Organizational Behavior, 16, 469–485.
Golembiewski, R. T., Billingsley, K., & Yeager, S. (1976). Measuring change persistency in
human affairs: Types of change generated by OD designs. Journal of Applied Behavioral
Science, 12, 133–157.
Greenberg, J. (1994). Using socially fair treatment to promote acceptance of a work site smoking
ban. Journal of Applied Psychology, 79, 288–297.
Hackett, R. D., Bycio, P., & Hausdorf, P. A. (1994). Further assessments of Meyer and Allen’s
(1991) three-component model of organizational commitment. Journal of Applied Psychol-
ogy, 79, 15–23.
Hirschman, A. O. (1970). Exit, voice, and loyalty: Responses to decline in firms, organizations,
and states. Cambridge, MA: Harvard University Press.
Hrebiniak, L. G., & Alutto, J. A. (1972). Personal and role-related factors in the development
of organizational commitment. Administrative Science Quarterly, 18, 555–573.
Jackson, D. N. (1970). A sequential system for personality scale development. In C.D. Spielberger
(Ed.), Current topics in clinical and community psychology, Volume 2, pp 61–96. New
York: Academic Press.
Jayne, M. (1994). Family role identification as a source of gender differences in the relationship
between parenthood and organizational commitment. Unpublished doctoral dissertation,
Tulane University, New Orleans.
Jenkins, J. M. (1993). Self-monitoring and turnover: The impact of personality on intent to leave.
Journal of Organizational Behavior, 14, 83–91.
Katz, D. (1964). The motivational basis of organizational behavior. Behavioral Science, 9, 131–
146.
Kelloway, E. K., & Barling, J. (1992, June). Perceived justice, job insecurity and organizational
commitment: The case of lay off survivors. In G. Johns (Chair), Organizational commitment.
Symposium conducted at the meeting of the Canadian Psychological Association, Quebec
City.
Konovsky, M. A., & Cropanzano, R. (1991). Perceived fairness of employee drug testing as a
predictor of employee attitudes and job performance. Journal of Applied Psychology, 76,
698–707.
Lee, K. B. (1992). A study of affective, continuance, and normative commitment to the organiza-
tion. Unpublished master’s thesis, Sung Kyun Kwan University, Seoul, Korea.
Lynn, D. F. (1992). Incorporating the procedural/distributive dichotomy into the measurement
of pay satisfaction: A study of recently graduated engineers and full-time faculty members
of Ontario universities. Dissertation Abstracts International, 54, 4371A.
Magazine, S. L., Williams, L. J., & Williams, M. L. (in press). A confirmatory factor analysis
examination of reverse coding effects in Meyer & Allen’s Affective and Continuance Com-
mitment Scales. Educational and Psychological Measurement.
Mathieu, J. E. (1991). A cross-level nonrecursive model of the antecedents of organizational
commitment and satisfaction. Journal of Applied Psychology, 76, 607–618.
Mathieu, J. E., & Zajac, D. (1990). A review and meta-analysis of the antecedents, correlates,
and consequences of organizational commitment. Psychological Bulletin, 108, 171–194.
Mayer, R. C., & Schoorman, F. D. (1992). Predicting participation and production outcomes
through a two-dimensional model of organizational commitment. Academy of Management
Journal, 35, 671–684.
McDonald, P. R. (1993). Individual-organizational value congruence: Operationalization and
consequents. Unpublished doctoral dissertation, University of Western Ontario, London,
Canada.
McGee, G. W., & Ford, R. C. (1987). Two (or more?) dimensions of organizational commitment:
Reexamination of the Affective and Continuance Commitment Scales. Journal of Applied
Psychology, 72, 638–642.
Meyer, J. P., & Allen, N. J. (1984). Testing the ‘‘side-bet theory’’ of organizational commitment:
Some methodological considerations. Journal of Applied Psychology, 69, 372–378.
Meyer, J. P., & Allen, N. J. (1986, June). Development and consequences of three components
of organizational commitment. Paper presented at the annual meeting of the Administrative
Sciences Association of Canada, Whistler, British Columbia.
Meyer, J. P., & Allen, N. J. (1991). A three-component conceptualization of organizational
commitment. Human Resource Management Review, 1, 61–89.
Meyer, J. P., Allen, N. J., & Gellatly, I. R. (1990). Affective and continuance commitment to
the organization: Evaluation of measures and analysis of concurrent and time-lagged rela-
tions. Journal of Applied Psychology, 75, 710–720.
Meyer, J. P., Allen, N. J., & Smith, C. A. (1993). Commitment to organizations and occupations:
Extension and test of a three-component conceptualization. Journal of Applied Psychology,
78, 538–551.
Meyer, J. P., Bobocel, D. R., & Allen, N. J. (1991). Development of organizational commitment
during the first year of employment: A longitudinal study of pre- and post-entry influences.
Journal of Management, 17, 717–733.
Meyer, J. P., & Gardner, R. C. (1994, March). Assessment of change in organizational commit-
ment during the first year of employment: An application of confirmatory factor analysis.
Paper presented at the Academy of Management, Research Methods Division, Conference
on Causal Modelling, West Lafayette, Indiana.
Meyer, J. P., Irving, P. G., & Allen, N. J. (1993, August). Person x environment interaction in
the development of organizational commitment. Presented at the annual meeting of the
American Psychological Association, Toronto, Ontario.
Meyer, J. P., Paunonen, S. V., Gellatly, I. H., Goffin, R. D., & Jackson, D. N. (1989). Organiza-
tional commitment and job performance: It’s the nature of the commitment that counts.
Journal of Applied Psychology, 74, 152–156.
Moorman, R. H., Niehoff, B. P., & Organ, D. W. (1993). Treating employees fairly and organiza-
tional citizenship behavior: Sorting the effects of job satisfaction, organizational commit-
ment, and procedural justice. Employee Responsibilities and Rights Journal, 6, 209–225.
Morrison, E. W. (1994). Role definitions and organizational citizenship behavior: The importance
of the employee’s perspective. Academy of Management Journal, 37, 1543–1567.
Morrow, P. C. (1993). The theory and measurement of work commitment. Greenwich: JAI Press.
Mowday, R. T., Porter, L. W., & Steers, R. M. (1982). Employee–organization linkages: The
psychology of commitment, absenteeism, and turnover. New York: Academic Press.
Mowday, R. T., Steers, R. M., & Porter, L. W. (1979). The measurement of organizational
commitment. Journal of Vocational Behavior, 14, 224–247.
O’Reilly, C. A., & Chatman, J. (1986). Organizational commitment and psychological attachment: