Journal of Occupational Psychology, 1985,58,277-288.

Printed in Great Britain

0 1985 The British Psychological Society

The measurement and prediction of

career commitment
G A R Y J . BLAU*
Temple University
Using a sample of 119 registered nurses from a large urban hospital, this longitudinal
study investigated the nomological network of career commitment by: (a) determin-
ing if a distinct measure of career commitment could be operationalized, and
(b) examining whether such a measure showed a different relationship to withdrawal
cognition scales than measures of other work commitment concepts. The study also
tested the importance of situational and individual difference variables in predicting
career commitment.

The general topic of careers has been extensively studied for many years. Behavioural
scientists have generally focused their efforts on such career topics as: occupational/career
choice, career development, and career mobility (Hall, 1976). Related to these issues is the
concept of career commitment. According to Hall (1971), career commitment is:
. . . the strength of one’s motivation to work in a chosen career role. Commitment to
the entire career field or role is to be distinguished from commitment to the job (i.e.,
job involvement as described by Lodahl & Kejner, 1965), or to one’s organization
(i.e., organizational identification as described by Hall, Schneider & Nygren, 1970).
These three forms of commitment are often correlated, but they are theoretically dis-
tinct and may often have different causes and consequences’ (p. 59).
Beyond Hall’s (197 1) argument for distinguishing career commitment from other
work commitment constructs, Morrow (1983) has called for a moratorium on new work
commitment constructs until existing perspectives have been evaluated. One of her conclu-
sions is that empirical work testing for conceptual redundancy between work commitment
constructs, such as career commitment, organizational commitment and job involvement,
thus far has been inconclusive. Morrow (1983) has argued that more rigorous studies are
needed which establish the empirical validity of work commitment measures (including
career commitment), and use such measures as dependent variables. One purpose of this
study is to investigate the nomological network (Cronbach & Meehl, 1955) of career
commitment by: ( a ) testing the discriminant validity of a measure operationalizing career
commitment versus measures operationalizing job involvement and organizational com-
mitment, and (h) examining the relationships of career commitment, job involvement and
organizational commitment to different types of withdrawal cognitions. Withdrawal
cognitions (Mobley, 1977) are an important part of the individual turnover process, and
represent the extent to which an individual has thought about quitting his or her job and
searching for an alternative prior to actually quitting. Another purpose of this study is to
investigate the importance of specific situational and individual difference variables in
predicting career commitment.
Operationalizing career commitment
Morrow (1983) included Greenhaus’ (1971, 1973) measure of career salience in her

*Requests for reprints should be addressed to Gary J. Blau, Human Resource Administration Department,
Temple University, Philadelphia, PA 19 122, USA.
277
2 78 G A R Y J. B L A U

discussion of research operationalizing career commitment. Greenhaus (1 973) factor

analysed his 28-item measure into three subscales: ( I ) general attitude toward work (con-
taining value and job focus items), for example, ‘ Work is one of those necessary evils ’;
‘ It is difficult to find satisfaction in life unless you enjoy your job ’; (2) vocationalplanning
and thought, for example, ‘ I enjoy thinking about and making plans about my future
career ’, and (3) the relative importance of work (containing items that require the respon-
dent to express preferences between work and non-work activities), for example, ‘ I intend
to pursue the job of my choice, even it it allows only very little opportunity to enjoy my
friends ’. However, Morrow (1983) goes on to point out that the item content of several of
these subscales is partially redundant with other work commitment concepts such as: cen-
tral life interest (Dubin, 1956);job involvement (Lodahl & Kejner, 1965); and Protestant
work ethic (Blood, 1969). In their measure of career orientation, Marshall & Wijting
(1982) labelled one factor ‘career centredness’ and the second factor ‘ career commit-
ment ’. According to Marshall & Wijting (1982), career commitment focuses on a person’s
general commitment to work throughout all stages of life. A sample item from this
measure asks ‘ To what extent do work activities figure into life plans? ’ Such a definition
of career commitment seems to be redundant with the concept of work involvement (Gorn
& Kanungo, 1980; Kanungo, 1982), which is defined as the importance or centrality of
work throughout one’s life. Sample items for measuring work involvement (Kanungo,
1982) include, ‘ Work should be considered central to life’, and ‘ In my view, an
individual’s personal life goals should be work-oriented ’.
Given the conceptual and measurement overlap of Greenhaus’ (1971, 1973) and
Marshall & Wijting’s (1982) scales with other work commitment concepts, an alternative
conceptualization and operationalization of career commitment needs to be developed.
Research on concepts related to career commitment, such as professional commitment,
occupational commitment and career orientation, suggests both a conceptual definition of
career commitment and a way to operationalize the definition. Career commitment can be
defined as one’s attitude towards one’s profession or vocation. Although the referents
‘ profession ’ and ‘ vocation ’ are somewhat restrictive, they are necessary. It is important
to anchor the career commitment concept in more specific terminology than ‘work in
general ’, while also using broader referents than ‘job ’ and ‘ organization ’, so as not to
make career commitment redundant with other concepts (e.g. work involvement, job
involvement, organizational commitment). Past research operationalizing professional
commitment, occupational Commitment, and career orientation suggests a pool of items
from which to develop a measure of career commitment.
One approach to measuring professional commitment is to ask individuals the extent
to which they engage in certain professionally related activities such as reading journals,
attending meetings, or joining associations (e.g. Price & Mueller, 198I). Professional com-
mitment has also been operationalized in terms of the individual’s reluctance to leave
his/her professional role (e.g. Thornton, 1970). Typically, such measures of professional
commitment are either one-item or, if multi-item, they do not have supporting psycho-
metric information (e.g. reliability estimates). Occupational commitment (e.g. Downing
et al., 1978) has been operationalized in terms of the individual’s desire to remain in an
occupation following an assessment of feasible alternatives. Career orientation is a con-
cept which has been defined in different ways. One of the dimensions which Liden & Green
(1980) found to be part of career orientation was individual interest in a larger class of
work activities than those associated with a specificjob or organization. Other operation-
alizations of career orientation (e.g. Greenhaus, 1971; Jans, 1982) also seem to emphasize
this dimension.

Career commitment and withdrawal cognitions

As part of testing the nomological network (Cronbach & Meehl, 1955) of career
 MEASUREMENT A N D PREDICTION OF CAREER COMMITMENT 279

commitment, one can investigate if career commitment has a different relationship to

withdrawal cognitions than job involvement or organizational commitment. Typically in
turnover research (Mobley, 1977; Mobley et al., 1978), withdrawal cognitions have been
operationalized with respect to the job, for example, asking: (1) how often the individual
thought about quitting his/her job; (2) how likely the individual would search for an
alternative job; and (3) the likelihood of an individual quitting his/her job. As such, job
involvement and organizational commitment (because a job is often associated with an
organization) have been found to be related to such withdrawal cognitions (e.g. Miller
et al., 1979; Mobley el al., 1979). However, career commitment would not necessarily be
related to such withdrawal cognitions since it is associated with a broader set of referents
(e.g. profession) than a specificjob. Yet an individual with a high level of career commit-
ment should be less likely to have career withdrawal cognitions.
The rationale for posing these differing relationships is drawn partly from
Gouldner's (1958) distinction between ' locals ' and ' cosmopolitans '. Individuals with a
high ' local ' orientation are high on loyalty to the employing organization and likely to
use an inner reference group orientation. Since such individuals are more likely to be com-
mitted to their organizations and involved in their jobs, they are less likely to think about
leaving the job. Meanwhile, individuals with a high ' cosmopolitan ' orientation are low on
loyalty to the employing organization and likely to use an outer reference group orien-
tation (e.g. professional association). Such individuals are more likely to be committed to
their professions and less likely to think about leaving the profession. Yet individuals with
high career commitment will more readily leave a job if they feel it will enhance their
career. Being able to empirically distinguish the relationship of career commitment with
different types of withdrawal cognitions from the relationship of job involvement and
organizational commitment to similar types of withdrawal cognitions would further jus-
tify the conceptualization and operationalization of career commitment.

Predicting career commitment

Research on career issues (e.g. choice) shows several different perspectives. One per-
spective, called the ' differentialist ' view (Van Maanen & Schein, 1977), emphasizes
individual differences (e.g. personality) as the primary predictors of subsequent individual
career behaviours (e.g. Holland, 1966). A second perspective, called the ' organizational '
view (Van Maanen & Schein, 1977), emphasizes situational characteristics (e.g. superior-
subordinate relationships) as the primary predictors of subsequent individual career
behaviours (e.g. Hall, 1971). Building upon these ideas, London (1983) has presented a
broad theoretical framework suggesting that both individual and situational variables are
important predictors of individual career behaviours.
Specific individual characteristics which London (1983) suggests in his model include:
primacy of work in one's life, internal control, development orientation, job involvement
and organizational commitment. One way of viewing the primacy of work in one's life is
by looking at the marital status of the individual. Work can have more primacy for
non-married individuals versus married individuals (Tinsley & Faunce, 1980). Internal
control refers to the belief that the individual can influence/control events affecting his/her
career. Development orientation suggests that individuals desire to grow on their jobs and
meet challenges. The degree to which the individual identifies with his or her job and
organization is also predicted to positively influence career commitment. In addition,
research on demographic predictors of other commitment-related concepts such as job
involvement and organizational commitment (Kanungo, 1982; Mowday et af., 1982) has
shown age and tenure to be positively related to job involvement and organizational com-
mitment. Such related research suggests testing whether age and tenure can help predict
career commitment.
280 G A R Y J. B L A U

Specific situational characteristics which London (1983) suggests can help predict
career commitment include: organizational ambiguity, and supervisor consideration and
control. Organizational ambiguity refers to the clarity, structure, and stability of work
goals and methods. Supervisor consideration and control deals with the degree to which
the supervisor develops friendships with subordinates as well as sets goals and monitors
work. As yet, very little empirical work has been done evaluating London's (1983) model.
By operationalizing the above-mentioned individual and situational characteristics and
seeing which ones are the most important for predicting career commitment, a partial test
of London's (1983) model can be made.

METHOD

Subjects andprocedure
A questionnaire was constructed and administered twice on a voluntary basis to
registered staff nurses working at a large hospital located in a midwestern city. Registered
nurses were unionized and union approval for the study was given. Administrative level
nursing personnel within the hospital helped coordinate the distribution of the question-
naire to staff nurses on a group or individual basis. Prior to the first administration, the
questionnaire was pre-tested on a small sample of supervisory level nursing personnel.
While working on their normal shift, 221 staff nurses filled out the questionnaire at Time 1
(TI) with the understanding that their individual responses would be kept confidential.
The participating sample of 22 1 registered staff nurses constituted approximately 40 per
cent of the total registered nursing staff at the hospital. To check if such a sample was
biased, demographic characteristics of the TI sample were compared to those of the total
registered nursing staff hospital population (made available by the recruitment office).
Results of the comparison indicated no significant differences in terms of age, sex, edu-
cation, race, tenure (nurse and hospital), shiftwork, and part-time vs. full-time status. A
second administration of the same questionnaire was conducted approximately seven
months later (TJ. Two hundred and twenty-eight staff nurses filled out this questionnaire,
including 1 19 repeat-respondents from T I and 109 study newcomers. A comparison of the
repeat-respondent group to the T, newcomer group and TI dropout group indicated no
significant differences on the above-mentioned demographic characteristics. The results in
this paper are based only on the 119 repeat-respondents. A demographic breakdown of
the 119 nurses revealed that: (1) 98 per cent were female; (2) 39 per cent had a
baccalaureate degree in nursing; (3) 91 per cent were Caucasian; (4) 50 per cent worked
on a rotating shift; and ( 5 ) 82 per cent worked full time (40 or more hours/week).
Measures
Career commitment. Career commitment is defined as one's attitude towards one's pro-
fession or vocation. Eight items were used to measure career commitment on a five-point
scale (1 = strongly disagree, 5 =strongly agree), and items were linearly summed to create
a scale score. The items were drawn from a pool of items based on previously mentioned
studies which measured the following concepts: professional commitment (Price &
Mueller, 1981); occupational commitment (Downing et a f . , 1978); and career orientation
(Liden & Green, 1980). The eight items consisted of the following: (1) ' If I could get
another job different from being a nurse and paying the same amount, I would probably
take it '; (2) ' I definitely want a career for myself in nursing '; (3) ' If I could do it all over
again, I would not choose to work in the nursing profession '; (4) ' If 1 had all the money I
needed without working, I would probably still continue to work in the nursing pro-
fession '; (5) ' I like this vocation too well to give it up '; (6) ' This is the ideal vocation for a
life work '; (7) ' I am disappointed that I ever entered the nursing profession' ; (8) ' I spend
a significant amount of personal time reading nursing-related journals or books '. The
 MEASUREMENT A N D PREDICTION O F CAREER COMMITMENT 28 1

basic content for the first four items is taken from Liden & Green (1980), the next three
items are drawn from Downing et al. (1 978), and the last item comes from Price & Mueller
(198 I). Since this is an unproven measure and career commitment is the focus of the study,
descriptive and psychometric information on this scale are reported in the Results section.
(For reverse-scored items in this and other measures see listing in Table 1.)

Job involvement was measured using a nine-item scale based on Kanungo’s ( 1 982) study.
Job involvement is defined as the degree to which the individual identifies with a job,
i.e. the importance of a job in one’s life (Dubin, 1956). Much of the item content of
Kanungo’s (1982) measure is based upon Lodahl & Kejner’s (1965) measure (e.g. ‘ 1 live,
eat, and breathe my job ’). However, the Lodahl & Kejner (1965) measure itself has been
found to have factor stability problems (e.g. Schwyart & Smith, 1972; Cummings &
Bigelow, 1976). Answers are recorded on a five-point response scale ( 1 =strongly disagree,
5=strongly agree), and items are linearly summed to create a scale score. Relevant
descriptive and psychometric information about this and the other measures described
below are given in Table 2.

Organizational commitment was measured using Porter & Smith’s (1970) nine-item
measure, where organizational commitment is defined in terms of one’s identification with
the organization. Evidence for the construct validity of this scale is provided by Mowday
et al. (1 979). A sample item is ‘ I really care about the fate of this organization ’. Answers
are recorded on a five-point scale (1 =strongly disagree, 5 =strongly agree), and items are
linearly summed to create a scale score.

Job withdrawal cognitions were measured using a three-item scale based upon the previous
work of Mobley (e.g. Mobley et al., 1978; Mobley et al., 1979), who has shown that;
(1) thoughts of quitting the job, (2) intention to search for another job, and (3) intention to
quit the job, are important withdrawal cognitions associated with turnover. Answers are
recorded on a seven-point scale (1 =very unlikely, 7 = very likely), and items are linearly
summed to create a scale score (Michaels & Specter, 1982).

Career withdrawal cognitions were measured using a three-item scale parallel to the job
withdrawal cognitions scale, but using the word ‘ profession ’ as a referent instead of
‘job ’. The three items involved: (1) thinking about leaving the nursing profession;
(2) intention to look for a different profession; and (3) intention to stay in the nursing
profession for some time (reverse-scored), Items are measured on a five-point scale
( I =strongly disagree, 5 = strongly agree), and item responses are linearly summed to
create a scale score.

Age was measured using a one-item self-report measure which consisted of 10 categories
(1 = less than 2 1 years, 2 = 22-26 years, 3 = 27-3 1 years, 4 = 32- 36 years, 5 = 37-4 1 years,
6 =42-46 years, 7 = 47-5 1 years, 8 = 52-56 years, 9 = 57-61 years, 10 = 62 years or more).

Nurse tenure and hospital tenure were each measured using one-item self-report measures
in the form of 10 categories. Differences between these two types of tenure are possible,
since nurse tenure asks the individual how long he or she has worked as a registered nurse
overall, while hospital tenure asks how long the individual has worked as a registered
nurse at the particular hospital. The 10 categories for both measures are: 1 =&6 months,
2=7-12 months, 3 = 13 months-2 years, 4 = 3 4 years, 5=5-6 years, 6=7-8 years,
7=9-10years,8= 11-12years, 9 = 13-14years, 10=more than 14years.
282 GARY J. BLAU

Marital stutus was measured using a one-item self-report measure. There were two
categories: 0 = not married ( n = 48), 1 = married (n = 71).

Growth needstrength was measured using Hackman & Oldham’s (l974)l I-item Job-choice
Scale, where individuals are asked to choose between two different job situations on a
five-point scale (1 =strongly prefer Job A, 5=strongly prefer Job B). The growth need
strength scale operationalizes the degree to which an individual seeks Maslow’s (1954)
higher-order needs (e.g. esteem, achievement) versus lower order needs (e.g. security). A
sample item is Job A =’ where the pay is very good ’ versus Job B = ‘ where one has a
chance to use his/her skills and abilities ’, Item responses were linearly summed to create a
scale score.

Locus qf control was measured using Levenson’s (1973) eight-item Chance Scale. Much of
the item content of Levenson’s (1973) measure is based upon Rotter’s (1966) measure (e.g.
’ T o a great extent my life is controlled by accidental happenings ’). However, the Rotter
(1966) measure itself has been found to have factor stability problems (e.g. Collins, 1974;
Blau, 1984). Answers are recorded on a six-point scale ( I = strongly disagree, 6 =strongly
agree), and items are linearly summed to create a scale score: the lower one’s score, the
more internal one’s locus of control orientation.

Role amhiguity was measured using Rizzo et al.’s (1970) six-item measure. Answers are
recorded on a five-point scale (1 = strongly disagree, 5 = strongly agree), and items are
linearly summed to create a scale score. A sample item is ‘ 1 know exactly what is expected
of me ’.

Considerution was measured using the 10-item scale from Form 12 of the Leader Behav-
ior Description Questionnaire (Hemphill & Coons, 1957). Subordinates are asked how
they perceive their supervisor. A sample item is ‘ She/he is friendly and approachable ’.
Answers are recorded on a five-point scale ( 1 = never, 5 =always), and items are linearly
summed to create a scale score.

hiticiting structure was measured using the 10-item scale from Form 12 of the Leader
Behavior Description Questionnaire (Hemphill & Coons, 1957). Subordinates are asked
how they perceive their supervisor. A sample item is ‘ She/he lets me know what is expec-
ted of me ’. Answers are recorded on a five-point scale (1 =never, 5 =always), and items
are linearly summed to create a scale score.

Analysis
Factor analysis was used to determine whether career commitment could be opera-
tionalized using a measure which demonstrated discriminant validity from measures of
job involvement and organizational commitment. Factor analysis has been previously
used to test the discriminant validity of measures involving job characteristics and job
satisfaction (e.g. Ferratt et ul., 198 I). To demonstrate discriminant validity, career commit-
ment items should load on a different factor than the job involvement and organizational
commitment items. A principal components analysis was initially carried out on the 26
career commitment, job involvement, and organizational commitment items at Time 1
(T,) and Time 2 (T,) to determine the number of dimensions.
To test which measured individual and situational characteristics are the most
important for predicting career commitment, multiple regression analysis (Nie, 1983) was
used. Several multiple regression analyses were made possible by the collection of two data
waves. Following the procedure used by Bateman & Strasser (1984), a static (within-time)
regression analysis of the Time 1 data was followed by a second static regression analysis
 M E A S U R E M E N T A N D P R E D I C T I O N OF CAREER COMMITMENT 283

Table 1. Factor loadings of career commitment,job involvement, und organizational commitment

items over time
~~

Time 1 Time 2

Items 1 2 3 1 2 3

1. Would take different job paying

same" 0.18 0.15 0.59 0.22
2. Want career in nursing -0.03 0.30 0.13 0.64 0.10
3. If could do it all over, would not
choose nursing. 0.28 0.20 0.74 0.1 5
4. If had all the money needed, still
work in nursing -
0.31 0.23 0.76 0.22
5. Ideal vocation too well to give it
UP 0.24 0.07 0.17 0.67 0.30
6. Ideal vocation for a life work 0.27 0.10 0.04 0.55 0.24
7. Disappointed ever entered nursing
profession8 -0.01 -0.01 0.73 0.1 2
8. Spend time reading nursing-
related material 0.24 0.27 0.1 8 0.31 0.1 7

9. Most important things involve job 0.03 0.11 0.51 0.23 0.08 0.38
10. Job small part of self* 0.29 0.09 0.65 0.1 1 0.27 0.51
11. Very personally involved in job 0.22 0.20 0.45 - 0.03 0.04 0.65
12. Live, eat and breathe job 0.08 0.17 0.59 0.04 0.1 9 0.69
13. Most interests centred around job 0.15 0.09 0.49 0.18 0.30 0.49
14. Very strong ties to job 0.25 0.22 0.36 0.16 0.23 0.41
15. Most of life goals are job-oriented 0.01 0.10 0.50 0.05 0.1 3 0.58
16. Consider job central to existence -0.02 0.08 0.49 0.15 0.11 0.57
17. Like to be absorbed in job 0.27 0.04 0.37 0.1 7 0.16 0.33

18. Willing to put in extra effort for

organization 0.73 0.14 -0.01 0.76 0.22 0.06
19. Talk up organization as great place
to work 0.45 -0.03 0.10 0.55 0.1 0 0.1 4
20. Accept almost any job to stay in
organization 0.68 0.18 0.15 0.62 0.24 0.1 9
21. Find that values are similar to
organization 0.81 0.16 0.07 0.80 0.32
__ 0.13
22. Proud to tell others that part of
organization 0.66 0.19 0.10 0.63 0.25 0.16
23. Organization inspires one's best
job performance 0.67 0.23 0.14 0.81 0.22 0.1 5
24. Glad chose this organization over
others 0.61 0.15 0.20 0.60 0.07 0.1 6
25. Really care about fate of organiza-
tion 0.50 0.24 0.05 0.58 0.30 0.09
26. This is best organization to work -
for 0.46 0.30
- 0.10 0.46 0.1 8 0.1 1

Eigenvalues 8.9 2.6 2.0 9.0 2.9 2.1

Note. n = 119. Factor loadings>0.30 are significant ( P < 0 , 0 5 ) .

#Reverse-scoreditem.
284 GARY J . BLAU

applied to the Time 2 data. This second static regression analysis served as a replication
test for the predictor variables. Finally, a multiple regression was performed using the
Time 1 predictor variables as ' independent ' variables in a regression equation predicting
Time 2 career commitment. The impact of the predictors was assessed hierarchically, i.e.
after the influence of Time 1 career commitment on Time 2 career commitment was first
controlled for. Such a time-lagged regression analysis has the effect of removing some
method variance inherent in a cross-sectional self-report methodology. In addition,
because of the temporal ordering of the predictor (Time 1) variables and the criterion
(Time 2 career commitment), a multivariate test which is more representative of the
presumed causal order is created.

RESULTS
A scree test (Cattell, 1966) performed on the T , principal components analysis indi-
cated that a three-factor solution was the most appropriate and accounted for 54 per cent
of the total variance. Subsequent factor analysis with varimax rotation showed the three-
factor solution to be the clearest and most interpretable. Similar factor-analytic results
were found using the equivalent T, data. A three-factor solution was found to be the most
appropriate and accounted for 54 per cent of the total variance. Table 1 shows the results
of these factor analyses. A minimum factor loading of 0.30 (Nunnally, 1978) was used as a
guideline for considering an item to be part of a factor. These results show career
commitment to be operationally distinguishable from job involvement and organizational
commitment. In addition, a factor congruency coefficient (Harman, 1967) of 0.80 was
found, further indicating that the factor structure of the career commitment, job involve-
ment, and organizational commitment items remained stable over time. Based upon these
factor analyses, an eight-item measure of career commitment was constructed and used in
subsequent data analysis.

Table 2. Means. standard deviutions. internal consistencies, and test-retest reliabilities of study
variables

Time 1 Time 2
Test-retest
Variable Mean SD a' Mean SD a reliability

1 . Age 3.9 1.3 4.0 1.3 n.a.b 0.90

2. Nurse tenure 5.2 2.8 n.a.b 5.5 2.9 n.a.b 0.85
3. Hospital tenure 4.1 2.3 n.a.b 4.5 2.2 0.87
4. Marital status 0.6 0.3 0.6 0.3 n.a.b 0.86
5. Growth need strength 35.7 5.4 0.63 37.3 5.0 0.61 0.63
6. Locus of control 21.8 5.7 0.75 21.4 5.8 0.76 0.67
7. Role ambiguity 14.2 3.5 0.76 13.9 3.3 0.73 0.61
8. Consideration 32.0 4.7 0.77 32.1 4.6 0.74 0.55
9. Initiating structure 35.8 3.9 0.73 35.1 4.9 0.80 0.51
10. Job involvement 33.3 7.5 0.74 34.1 7.9 0.79 0.63
1 1 . Organizational commitment 31.7 5.8 0.89 31.8 5.6 0.88 0.74
12. Career commitment 28.5 6.5 0.87 29.0 6.2 0.85 0.67
13. Job withdrawal cognitions 9.8 4.3 0.70 9.6 4.2 0.73 0.60
14. Career withdrawal cognitions 8.1 2.4 0.67 8.0 2.4 0.71 0.57

Note. n = 1 1 9.
.Internal consistency estimate (Cronbach's alpha).
bAge, nurse tenure, hospital tenure, and marital status are one-item measures.
 Table 3 . Intercorrelations among the major study variables"

Variable 1 2 3 4 5 6 7 8 9 10 11 12 13 14

1. Age 0.08 -0.04 -0.06

2. Nurse tenure 0.26" 0.24" -0.28"
3. Hospital tenure 0.05 -0.10 -0.08
4. Marital status -0.1 8' -0.10 -0.02
5. Growth need strength 0.1 6' 0.04 0.06
6. Locus of control -0.1 9' -0.08 -0.1 1
7. Roleambiguity -0.38" 0.1 6' 0.1 4
8. Consideration 0.17' -0.07 -0.03
9. Initiating structure 0.25" 0.01 0.01
10. Job involvement 0.21 * -0.20' -0.09
11. Organizational commitment 0.1 8 0.23' -0.1 1
12. Career commitment - -0.07 -0.38"
13. Job withdrawal cognitions -0.08 - 0.40"
14. Career withdrawal cognitions -0.41 '. -0-45" -

Note n = 119
'Correlations above the diagonal (-) are for Time 1 and below the diagonal (-) are for Time 2
'P<O 05, "P<O 01 (twotailed)
286 G A R Y J. BLAU

The descriptive results in Table 2 show no significant changes in study variables over
time. Table 3 displays the static zero-order correlations among all variables at Time 1 and
Time 2. As can be seen, many of the correlations which are significant at Time 1, are also
significant at Time 2. In particular, the results in Table 3 show that career commitment has
a different relationship to withdrawal cognitions than job involvement and organizational
commitment. Career commitment shows significant negative relationships to Career with-
drawal cognitions, whereas job involvement and organizational commitment do not show
significant negative relationships to career withdrawal cognitions. However, job involve-
ment and organizational commitment show significant negative relationships to job
withdrawal cognitions, whereas career commitment does not. These results add to the
nomological network differentiating career commitment from other commitment scales
(e.g. job and organization), and also call attention to the type of referent (i.e. job vs.
career) used by withdrawal cognition scales in turnover research.
In terms of predicting career commitment, the correlational results in Table 3 con-
sistently indicate that individuals who: have been nurses longer, are not married, have a
higher growth need strength, have an internal locus of control orientation, perceive less
role ambiguity and more supervisor initiating structure, and identify with their job and
organization are more career committed. These correlational results are consistent with
London's (1983) model.
The results for all of the regression analyses-Time 1 static, Time 2 static, and
time-lagged-are quite consistent with one another. Together, they indicate that nurse ten-
ure, marital status and perceived supervisor initiating structure and role ambiguity are
significant (P<0.05)predictors of career commitment. For the Time 1 static regression
analysis, the standardized beta weights for nurse tenure, marital status, initiating structure
and role ambiguity are 0.23, -0.19, 0.20 and -0.38. For the Time 2 static regression
analysis, the standardized beta weights for these same variables are 0.18,-4.16, 0.18 and
-4.21. Finally, the standardized beta weights of nurse tenure, marital status, initiating
structure and role ambiguity for the time-lagged regression are 0.16, - 0 . 1 5,0.17 and -0.27.
Together these four significant predictors account for the following percentage variance
in career commitment: Time 1 static regression -27 per cent, Time 2 static regression
- 14 per cent, and time-lagged regression - 15 per cent. As a partial test of London's
(1 983) model, these regression results consistently indicate that only certain individual
and situational characteristics are important predictors of career commitment.

DISCUSSION
One purpose of this study was to investigate the nomological network of career
commitment by determining: (a) if a distinct measure of career commitment could be
operationalized, and (b) if such a measure showed a different relationship to withdrawal
cognitions scales than measures of other work commitment concepts. Despite the
encouraging results in answering these questions, it is important to realize that the opera-
tionalization of career commitment in this study, because it emphasizes a ' profession '
referent, is somewhat restrictive. Nurses have a higher professional orientation than many
other occupations due to their education, training, professional organizations, etc. Such
an orientation could enhance nurses seeing their ' profession ' as a career (as might aca-
demics, scientists, doctors, lawyers, engineers, etc.). This is partly because membership in
such a profession goes beyond a specific job or the organization where one works. How-
ever, efforts to generalize this career commitment scale to workers in less ' professionally
oriented ' occupations (e.g. mechanics, custodians, factory workers, waitresses/waiters)
may prove to be difficult. Certainly the application of this measure to additional samples
of a diverse nature is imperative to see if corroborating reliability and validity evidence can
be found. In addition. if one views career commitment as a more stable orientation,
 MEASUREMENT A N D P R E D I C T I O N O F CAREER COMMITMENT 287

lengthening the time period beyond seven months to test the measure’s stability would be
useful in future research.
Turning to the second study purpose, a partial empirical test of London’s (1983)
model was successfully made. Two individual and two situational characteristics were
found to be the best predictors of career commitment. Being unmarried and having more
work experience, as well as perceiving a structured work situation (low role ambiguity and
high supervisor initiating structure) help lead to stronger career commitmert. However
such results must be tempered by noting several limitations. First, the results are based on
measures of a very limited number of individual and situational characteristics. London’s
(1983) model is broad and encompasses over 60 different individual and situational predic-
tors of individual career behaviours. Second, the generalizability of results is limited due
to the nature of the sample (predominantly female registered nurses). Yet because of its
close relationship to other career concepts such as career motivation, and its apparent dis-
tinction from job involvement and organizational commitment, it is important for future
research to further explore and understand the career commitment construct.

REFERENCES
BATEMAN, T. & STKASSER, S. (1984). A longitudinal analysis of the antecedents of organizational commit-
ment. Acadrmy qfManagement Journal, 17,95-112.
BLAU.G. (1984). Brief note comparing the Rotter and Levenson measures of locus of control. Percrptual
and Motor Skil1.y. 58, 173-1 74.
BLOOD,M. (1969). Work values and job satisfaction. Journal of Applied Psychology, 53,456 459.
CATTELL, R. (1966). Handbook of Multivariate Experimental Psychology. Chicago: Rand McNally.
COLLINS.B. (1974). Four components of the Rotter internal-external scale. Journal of Personality and
Social Psychology. 29,38 1-39 I .
CRONBACH, L. & MEEHL,P. (1955). Construct validity in psychological tests. Psychological Bulletin, 21,
281-302.
CUMMINGS, T. & BIGELOW, J. (1976). Satisfaction, job involvement, and intrinsic motivation: An extension
of Lawler and Hall’s factor analysis. Journal qfdpplied Psychology, 61,523-525.
DOWNING, P., DUNLAP, M., HADLEY, B. & FERRELL, L. (1978). Work Attitudes and Performance Question-
nuirr. San Francisco: School of Nursing, University of California.
DURIN,R. (1956). Industrial workers’ worlds: A study of the central life interests of industrial workers.
Social Prohlems, 3, 131-142.
FERKATT, T., DUNHAM. R. & PIERCEJ. (1981). Self-report measures of job characteristics and affective
responses: An examination of discriminant validity. Academy of Management Journal. 24,78&794.
GORN,G. & KANUNGO, R. (1980). Job involvement and motivation: Are intrinsically motivated managers
more job involved? Organizational Behavior and Human Performancr, 26,265-277.
GOIJLDNER, A. (1958). Cosmopolitans and socials: Towards an analysis of latent social roles. Administra-
tive Science QuarzerIy, 2,281-316.
GREENHAUS, J. (1971). An investigation of the role of career salience in vocational behavior. Journul of
Vr,cational Behavior, 1,209-216.
GREENHAUS, J. (1973). A factorial investigation of career salience. Journal of Vocational Behavior, 3.95 98.
HACKMAN, R. & OLDHAM, G. (1974). The job diagnostic survey: An instrument for the diagnosis of jobs
and evaluation of job redesign projects. Technical report. Yale University.
HALL,D. (1971). A theoretical model of career subidentity development in organizational settings.
Organizational Behavior and Human Performance, 6,5&76.
HALL,D. (1976). Careers in Organizations. Santa Monica, CA: Goodyear.
HALL, D., SCHNEIDER, B. & NYGREN,H. (1970). Personal factors in organizational identification.
Administrative Science Quarterly, 15, 176-190.
HEMPHILL, J . & COONS,A. (1957). Development of the leader behavior description questionnaire. In
R. Stogdill & A. Coons (eds), Leadership Behavior: Its Description and Measuremrnt. Columbus: Ohio
State University, Research Monograph No. 88.
HOLLAND, J. (1966). The Psychology of Vocational Choice, Waltham, MA: Blaisdell.
288 GARY J. BLAU

JANS,N. (1982). The nature and measurement of work involvement, Journal of Occupational Psychology,
55.57-67.
HARMAN, H. (1967). Modern Factor Analysis. Chicago: University of Chicago Press.
KANUNGO, R. (1982). Work Alienation: An Integrative Approach. New York: Praeger.
LEVENSON, H. (1973). Multidimensional locus of control in psychiatric patients. Journal of Consulting and
Clinical Psychology, 41,397-404.
LIDEN, R. & GREEN,S . (1980). On the measurement of career orientation. Paper presented at the Midwest
Academy of Management meeting. Cincinnati.
LODAHL, T. & KEJNER, M. (1965). The definition and measurement of job involvement. Journal of Applied
Psychology, 49,2433.
LONI)ON. M. (1983). Toward a theory of career motivation. Academy ofhfanagement Review, 8,620430.
MARSHALL, S. & WIJTING, J. (1982). Dimensionality of woman's career orientation. Sex Roles, 8,620446.
MASLOW, A. (1954). Motivation and Personality. New York: Harper & Row.
MICHAELS, C. & SPECTER,P. (1982). Causes of employee turnover: A test of the Mobley, Griffeth, Hand,
and Meglino model. Journal of Applied Psychology, 67,53--59.
MILLER, H., KATERBERG, R. & HULIN,C. (1979). Evaluation of the Mobley, Horner and Hollingsworth
model of turnover. Journal of Applied Psychology, 64,509-5 17.
MOBLEY,W. (1977). Intermediate linkages in the relationship between job satisfaction and employee
turnover. Journal of Applied Psychology. 63,237-240.
MOBLEY, W . , HORNER, S. & HOLLINGSWORTH, A. (1978). An evaluation of precursors of hospital employee
turnover. Journul qf Applied Psychology, 63,408414.
MOBLEY, W . , GRIFFETH,R., HAND,H. & MEOLINO, R. (1979). Review and conceptual analysis of the
employee turnover process. Psychological Bulletin, 86,493-522.
MORROW,P. (1983). Concept redundancy in organizational research: The case of work commitment.
Acudemy ofManagemen! Review, 8,486500.
MOWDAY, R., STEERS, R. &PORTER, L. (1979). The measurement of organizational commitment. Journulof
Vocational Behavior. 14,224247.
MOWDAY, R., PORTER, L. & STEERS,R. (1982). Employee-Organization Linkages: The Psychology of Com-
mitment, Absenteeism, and Turnover. New York: Academic Press.
"1.: N. (1983).SPPS": Statistical Puckagefor the Social Sciences. New York: McGraw-Hill.
NUNNALLY, J. (1978). Ps.ychometric Theory. New York: McGraw-Hill.
PORTER,L. & SMITH,F. (1970). The etiology of organizational commitment. Unpublished manuscript,
University of California at Irvine.
PRICE, J. & MUELLER, C. (1981). A causal model of turnover for nurses. Academy c$Management Journal,
24,543 565.
Rizzo, J., HOUSE,R. & LIRTZMAN, S . (1970). Role conflict and ambiguity in complex organizations.
Administrative Science Quurterly, 15, 150-163.
ROTTER,J. (1966). Generalized expectancies for internal versus external control of reinforcement.
P.vychological Monographs. 80( I ) , Whole no. 609.
SCHWYART, W. & SMITH, P. (1972). Factors in the job involvement of middle managers. Journal of Applied
P.~,vchology,56,227-233.
THORNTON.R. (1970). Organizational involvement and commitment to organization profession.
Administrative Science Quarterly, 15,417426.
TINSLEY, 0.& FAUNCE, P. (1980). Enabling, facilitating, and precipitating factors associated with women's
career orientation. Journal q/ Vocational Behavior, 17,183-194.
VAN MAANEN, J. & ScmiN, E. (1977). Career development. In R. Hackman & L. Suttle (eds), Improving
Lijie a/ Work. Santa Monica, CA: Goodyear.

Received I December 1984: revised version received 23 April 1985