: Eff-79‘5“»
. 'I, "‘- 1 , '
533 5.? ‘ 31“ *-'-‘-""1"'-""1‘

 

em."
U'.‘ INK" 'm

Ian' \f-I

1 “(E :~I'I7I . '-I' II HA» 4;
' Ni 3;}??? :i 'in
H S
W ~

I .52:
all“ ' " IQ 233'
"7“; . I I ydfihyn'.

In I :i: ‘ {'1'iIII
«:st
“Vlhffi‘ "0'1

r

 

"'1'”.
“\‘l‘l' q
I‘M-“iv! ' IuI'

 

rr‘a 'L “ ‘ . V .‘ 'fhl" 45 '
II , I I III I. I , , I" .. 57. .1.“ -,'.!';I‘.| j . '.- fl] ‘ I\.‘v.5."h\"7
I I v 5 'I ‘ I ‘ 'I ' Y I
A U...

II'15I’H 9.8“,"
WIN?" “EJO‘i'l'
Ir I 'I . «la " . N‘ ' ‘ ' " ’ | 7." 'v "tat:

) I5 .' III .. ".u“ , 2 ‘. 5Ilu‘ . “H. I/.::.;.'5° 4-,
I . 5 5 .5 .. a :3 -.-:.' 5 :.~ri:£n" ..c «5 5. '....'

 

III.“ J’k' “u". ‘
y W;
mew “HI.

. ‘p!.’
I I I I _ I _ I 35'“: ,,
. j, y u I ' ‘5 tap "In .:‘ " 5- ‘ ‘I-yI‘ .." ' ' ; '0. HAW».

W51 95?.” ‘

V L I~IHI l1” 1

.IIIIIIIIII - I “It“

 

h.

IRESZ!

 

This is to certify that the

thesis entitled

SELF-ESTEEM, LOCUS OF CONTROL
AND THE JOB EXPERIENCE:
A CAUSAL ANALYSIS

presented by

MICHAEL P. FITZGERALD

has been accepted towards fulfillment
of the requirements for

M-A- degree mm

 

‘ ) I .- . ,
/ ,. / K/ /
\"' ':/ /14/ .V 61/1/2449“:

Major professor

Date ////az A4722
/ /

0—7639

 

2" lymll55l555nlu5yu55n5n55555555"' L

w:
25¢ per day per item

RETURNING LIBRARY MATERIALS:

Place in book return to remove
charge from circulation records

..mm4 '58 5 5,9," 5
5 "AR 1
. ‘WS‘ Q ~ Ogim;
Wino A 297 *
Wm , mifgfi‘i' ‘ w
I - {Jum H992
wflfdfififi @7135- l 169

0.0

H 3’36fiflpowm 32:3, in 21331

5 051941?» noao
$22.37 {$le

 

 

 

 

 

 

 

 

SELF-ESTEEM, LOCUS OF CONTROL
AND THE JOB EXPERIENCE:

A CAUSAL ANALYSIS

BY

Michael P. Fitzgerald

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

MASTER OF ARTS

Department of Psychology

1980

5’70

67/6

ABSTRACT

SELF-ESTEEM, LOCUS OF CONTROL
AND THE JOB EXPERIENCE:
A CAUSAL ANALYSIS

BY

Michael P. Fitzgerald

Previous research and practice concerning the relation-
ship between job experience and personality are reviewed.
The variety of theoretical frameworks and study designs used
to empirically study this relationship have not resolved the
issue of causality. The present study was_designed to
explore this issue further. Specifically, causal relation-
ships between two widely accepted, well-researched person-
ality constructs (locus of control and self-esteem) and job
characteristics, as measured by the Job Diagnostic Survey
(Hackman & Oldham, 1975) were examined. HA cross-lag corre-
lation design, corrected cross-lag coefficient analysis,
dynamic correlations, and frequency-of-change-in-product-

moment techniques were used to analyze data. Inconsistent

w‘ —‘—u
—-

and nonsignificant results were noted and possible expla-

nations are discussed.

TABLE OF

LIST OF TABLES . . . . . . .

LIST OF FIGURES. . . . . . .

REVIEW OF THE LITERATURE . .

Locus of Control and Job Experience

Self-Esteem and Job Experience

Focus of Study . . . . . .
Job Tenure as a Moderator.

METHOD 0 O O O O O O O O O 0

Subjects . . . . . .
Measures . . . . . . .
Procedure. . . . . .
Data Analyses. . .
Combination of Subsamp

les.

RESULTS AND DISCUSSION . . .

Reliability and Dimensionality

Locus of Control . . . . .
Self-Esteem. . . . . . . .

CONCLUSIONS. . . . . . . . .
REFERENCES . . . . . . . . .
APPENDICES
Appendix
A. Job Diagnostic Survey.
B. Locus of Control . . .

C. Self-Esteem. . . . . .

ii

CONTENTS

of Scales

Page

17
21
24
26
28
28
29
33
33
41
42
42
45
50
54

59

66
70

71

Appendix

D.

Causal Analysis Between Locus of Control and
New Hire

Perceived Job
Sample. . . .

Causal Analysis Between Locus of Control and
Total LCC

Perceived Job

Characteristics:

Characteristics:

Vocational Education Sample .

Causal Analysis
Perceived Job
LCC Sample. .

Causal Analysis
Perceived Job
LCC Sample. .

Causal Analysis
Perceived Job

Between Locus of
Characteristics:

Between Locus of Control and
High Tenure

Characteristics:

Between Locus of Control and
New Hire-

Characteristics:

Control and
Low Tenure

Low Tenure (LCC) Combined Sample.

Causal Analysis Between Self-Esteem and
New Hire

Perceived Job
Sample. . . .

Causal Analysis Between Self-Esteem and
Total LCC

Perceived Job

Characteristics:

Characteristics:

Vocational Education Sample .

Causal Analysis
Perceived Job
(LCC) Sample.

Causal Analysis
Perceived Job
(LCC) Sample.

Causal Analysis
Perceived Job

Between Self-Esteem and
Low Tenure

Characteristics:

Between Self-Esteem and
High Tenure

Characteristics:

Between Self-Esteem and
New Hire-

Characteristics:

Low Tenure Combined Sample. .

iii

Page

72

73

74

75

76

77

78

79

80

81

LIST OF TABLES

Table Page
1. Reliability Coefficients . . . . . . . . . . . . 43

2. Significantly Different Cross-Lags Between
Locus of Control and Perceived Job
Characteristics 0 O O O C O O O O I I O O O O O 46

3. Significantly Different Cross-Lags Between

Self-Esteem and Perceived Job Character-
istics O I O O O O O O O O O O O O O O O O O O 52

iv

LIST OF FIGURES

Figure Page

1. Cross-Lag Analysis of Overall Job Diagnostic
Survey Score (JDS) and LocuschControl (LOC). . 34

REVIEW OF THE LITERATURE

The relationship between personality and job experience
is one which has been of concern to researchers who study
individuals at work for many years. In this review, I will
begin with a discussion of how early pioneers dealt with the
causality issue in the work experience-personality relation-
ship and how it has affected the field of Industrial Psy-
chology. Next, the theories and research of those that have
questioned the traditional View of the causal process between
work and personality will be reviewed. Finally, I will dis-
cuss the personality traits which are of concern to this
study to determine what sort of relationship they might have
to an individual's job experience.

In 1887, Karl Marx noted the negative effects of the
bureaucratic social environment on the "characteristics" of
the individual. This thesis was elaborated by Durkheim
(1902), who maintained that the characteristics of individ-
uals could be examined as a function of the labor class to
which they belonged. These early sociological theorists
felt that occupational experience had an impact upon a
worker's psychological functioning.

Early pioneers in the field of industrial psychology,

however, seemed to follow another direction. One of the

principles of scientific management, as developed by
Frederick Taylor (1911), stated, "employees selected for the
work should be as perfectly matched to the physical and men-
tal demands of the job as possible." The logic of this
principle underlies much of the work in personnel testing,
then and still today. Individuals are viewed as static with
a stable set of characteristics which can be matched against
those of successful job incumbents and/or specified job
requirements to predict future performance on the job. The
assumption is that certain personal qualities are best suited
for particular occupations and that the process of choosing
and succeeding in an occupation often involves matching one's
personal qualities with those required in a given line of
work. Any correspondence between an individual's occupation
and personality are viewed as a result of selection and/or
modification of the job to meet incumbents' needs and values
(Kohn & Schooler, 1973).

This same "selection" thesis underlies much of the logic
of the vocational preference literature. For example,
Rosenberg (1957) presented correlational evidence that self-
confidence and "manner of relating to people" were related
to occupational choice. He claimed his results suggest that
personality will influence the type of career an individual
accepts. More recently, the theory of vocational choice
proposed by Holland (1973, 1976) assumes most peOple can be
categorized as resembling one of six personality types and

that people tend to prefer and search for environments that

are consistent with their dominant personal characteristics.
Many studies have reported positive relationships between
individual differences and preferences for type of work
(e.g., Robey, 1974). Morse and Young (1973) showed person-
ality related to task preferences. These works, however,
were cross-sectional in design and the direction of causality
was all too often assumed, rather than empirically tested.

The selection hypothesis of the work experience/person-
ality relationship can be placed in a "self-consistency"
theoretical framework as suggested by Korman (1970). He
offers the following hypothesis as to the nature of work
behavior, "All other things being equal, individuals will
engage in and find satisfying those behavioral roles which
will maximize their sense of cognitive balance or consis-
tency." This implies, he notes, that individuals will tend
to choose and find most satisfying those job and task roles
which are consistent with their self-cognitions. Korman
emphasized the role of self-esteem in determining work out-
comes. His theory and research adds credence to the hypoth-
esis that individual personality will effect both occupa-
tional choice and reactions to the job experience.

Much of the research dealing with task design in the
past decades has emphasized the moderating influence of
individual differences on the relationship between job char-
acteristics and behavioral and affective worker response.
For example, Hackman and Oldham (1975), Wanous (1974), and

Brief and Aldag (1975) have consistently demonstrated that

 

 

job
posi

for

 

rese
inco
the

task

ual'

56h
itlj

VdI

do
dif
dif

mcv

hel
SOH
CEI
Pro;
One
his
liVe
inCc

outc

job satisfaction is related to task characteristics more
positively for employees with high-growth need strength than
for workers with low-growth needs. Although some of this
research treating growth needs as a moderator has led to
inconclusive results (e.g., Stone, 1976), no one has denied
the importance of individual characteristics in the study of
task design and the importance of "fit" between an individ-
ual's personality and the characteristics of his or her job.

Research and practice in the fields of personnel
selection, vocational preference and task design have implic—
itly treated individual characteristics as "causal"
variables which influence differences in affective and
behavioral response to work. What these bodies of research
do not indicate, however, is whether individual personality
differences, themselves, result in part from qualitative
differences in the kinds of work people experience as they
move along their career paths.

A variety of theoreticians in the social sciences have
held the position that given a certain social milieu, per-
sons playing various culturally defined roles will take on
certain personality attributes (Mead, 1944; Reisman, 1950;
Fromm, 1941). E. C. Hughes (1958) claimed, "A man's work is
one of the more important parts of his social identity, of
his self; indeed, of his fate in the one life he has to
live." An understanding of occupational life, then, is
incomplete unless one understands the social-psychological

outcomes of occupational role performance (Nosow & Form, 1962) .

The "job experience affecting personality" hypothesis,
suggested in the theoretical work above, is one which has
been neglected in much of Industrial Psychology literature.
If valid, however, it would not only provide a better under-
standing of occupational life, but it would be of practical
use to career counselors. Individuals who have developed a
career plan at the initial stage of their careers might mod-
ify such a plan if individual perceptions, values and think-
ing processes are affected by the job relationship.
Brousseau (1978) noted "a job designed to mesh well with an
individual's current personality may differ from the kind of
job for which he would be best suited in the future." Kohn
and Schooler (1978) noted that this line of research would
provide a critical test of a theoretical question central to
the entire field of social structure and personality--whether
social structure affects personality only through its influ-
ence on childhood socialization processes or also through a
continuing influence during the entire life span.

This "personality consistency" issue was examined in
the comprehensive longitudinal study conducted by Block
(1971). One hundred and seventy-one men and women were
studied and followed quite closely from early childhood to
adulthood. Though not the main focus of the study, Block
concluded, "The unity or consistency of personality is com—
pellingly apparent in these data and is manifest in so many
and so diverse ways as, perhaps, to establish the unity

principle empirically once and for all."

This question of how personality is developed and influ-
enced during individual's lives has been, and continues to
be, a major concern of psychology. Research and theory in
this area has been guided by four major psychological models:
trait psychology, psychodynamics, situationism, and inter-
actionism. An excellent overview of these approaches can be
found in Engler and Magnusson's (1976) review of personality
theory and research.

One of the most dominant forces in psychology research
has been the trait perspective. Although various trait
theorists disagree to the specific structure and content of
traits, they agree that traits are the prime determinants of
behavior and serve as a predispositional basis for apparent
response-response (correlational) consistencies of behavior
in different situations. As trait theory does, psychody-
namic theories assume a basic personality core which serves
as a predispositional basis for behavior in various situ-
ations. This approach is concerned with the dynamics of the
elements of the personality structure. The work of Freud
and other neo-Freudian's (e.g., Jung, Erickson) characterizes
the psychodynamic model.

Focusing on environmental factors, as opposed to person
factors, situationism regards the stimuli in the situation
as the basic determinants of individual behavior. Although
some working in this vein infer internal motives, the clas-
sical Situationist, Skinner (1953), denies the legitimacy of

these motives and is concerned only with the empirical

analysis of the stimulus and the reinforcement contingencies
that shape behavior. The interactional model goes beyond

the situational approach, stressing the importance of person-
situation interactions in personality. This approach assumes
that the individual's behavior is influenced by the situation,
but the person also selects the situations in which he or

she performs.

The majority of research in Industrial and Organiza-
tional Psychology stems from the trait perspective. This is
not a surprising observation, since the field "grew out of
psychology's early success in describing and measuring dif-
ferences between people (Dunnette, 1976)." The areas of
Industrial Psychology referred to earlier could definitely
be classified as trait-oriented. Research dealing with per-
sonality and the work situation, however, seems to stem from
either the trait or situationist frameworks.

Many authors have suggested ways in which job experience
(the situation) may affect personality. Although none were
specified exactly as such, many of these theories seem to
stem from a situationist perspective. One of the main themes
of Argyris' Personality and Organizational Theory (1957, 1964,
1973) is the study of individual differences. He suggests
many ways of creating work worlds in which these individual
differences could flourish. He feels that individuals have
predispositions (or needs) which are highly influenced by
the situation, yet also highly potent to the individual.

The individual seeks to fulfill these predispositions, yet

their exact nature, potency, and degree to which they have
to be fulfilled are influenced by the organizational context,
for instance, the job content or context (Argyris, 1973).

He suggests that the more the organization approximates the
properties of a formal organization (for example, work which
prohibits independent behavior), the more individuals will
be required to seek expression of needs which approximate
"infant" (as opposed to "adult") personality characteristics.
By blocking the expression of certain adult needs (for exam-
ple, the need for competence), employees are forced to adapt
to this kind of environment by adjusting internal personal
forces so that these needs are no longer felt or no longer
create tension.

Indeed, research evidence exists which shows that
experience on a task or job may affect the need and goal
states of employees. For example, Breer and Locke (1965),
in a laboratory study varied the degree to which collective
and interdependent task behavior was required of research
subjects. An effect of working on tasks requiring inter-
dependent behavior was a fairly substantial change in the
subjects' measured attitude regarding the value of collec—
tive endeavors in a wide variety of situations.

Several empirical researchers have presented evidence
from the work sphere that supports the "job experience
causes personality" hypothesis. For example, Kornhauser
(1965) presented findings which indicated that task expe-

rience could in fact alter the personal orientations of the

participants on an enduring basis. Based on interviews with
655 blue-collar, auto industry workers, his most outstanding
finding was that mental health varied consistently with the
level of jobs the men held. When he compared factory
workers by occupational categories, the higher the occupation
(with respect to skill and associated attributes of variety,
responsibility, and pay), the better the average mental
health. Those in lower skilled jobs reported low self-
esteem, high anxiety, and an absence of an active or goal-
orientation. To determine if these occupational differences
in mental health were effects of job conditions or were due
to differential selection of the kinds of persons who enter
and remain in the several types of work, Kornhauser con-
trolled for education and a variety of other pre-job char-
acteristics (for example, father's socioeconomic status,
school success). He found that observed differences in
mental health could not be accounted for by these factors
and concluded that his analyses underscored the existence

of significant occupational effects apart from other
determinants.

Kornhauser felt that work performed essential psycho-
logical functions, stating, "It operates as a great stabi-
lizing, integrating, ego-satisfying central influence in
the patterns of each person's life." Consequently, if the
job fails to fulfill the needs of the personality (or, at
least, movement toward satisfaction of these needs), symp-

toms of impaired mental health are likely to appear. He

10

felt it unlikely that individuals could find adequate sub-
stitutes to the job to provide a sense of significance and
achievement, purpose and justification for their lives.

Kohn and Schooler's research (1969, 1973) reported
similar findings based on a sample of 3,101 men employed in
the United States. They found that occupational conditions
conducive to the exercise of self-direction in one's work-—
namely, freedom from close supervision, substantively com-
plex work, and a non-routinized flow of work are empirically
tied to valuing self-direction and to having an orientation
to oneself and to the outside world consonant with this
value. Using a two-stage least squares technique, they
found that "substantive complexity of the job" significantly
affected occupational commitment, job satisfaction, parental
valuation of self-direction, anxiety, self-esteem, stance
toward change and intellectual flexibility, much more than
these facets of psychological functioning affected sub-
stantive complexity. They concluded, "In all cases, job
affects man more than man affects job."

Kohn and Schooler's (1969) work, seemed to be cast in
a situational framework. They suggested that job experience
influences personality by shaping perception of reality.
Individuals' jobs affects their perceptions, values and
thinking processes because it shapes the everyday realities
and demands they must cope with. The researchers thought a

generalization model best explained their findings, claiming

11

that workers' ways of coping with the realities of their
jobs is generalized to non-occupational realities.

The conclusions reached in both the Kornhauser and the
Kohn and Schooler studies, however, seem unwarranted, given
the cross-sectional data employed. Although both attempted
to "work around" this problem, causal interpretations are
not as credible as those reached from a longitudinal inves-
tigation, as Kohn and Schooler (1978) conducted later.

Using a follow-up sample of 687 male subjects from
itheir original study, the researchers employed a maximum-
1ikelihood confirmatory factor analysis by which they were
able to separate measurement error from real change in the
parameters of interest. A reciprocal causal model was
developed for these measures of the two variables of inter-
est in their study, "substantive complexity" of the job and
the "intellectual flexibility" of the respondents. By sub-
stantive complexity, the researchers meant the degree to
which the work in its very substance requires thought and
independent judgment. Their index of intellectual flexi-
bility was measured by an individual's intellectual per-
formance in an interview situation.

Despite the fact that these variables showed high sta-
bility over a ten-year time span, a structural equation
causal analysis demonstrated that the effect of the sub-
stantive complexity of an individual's job on intellectual
flexibility was strong--almost one-fourth as great as the

men's earlier levels of intellectual flexibility on their

12

later intellectual flexibility. This effect, however, was
contemporaneous. The lagged effect of intellectual flexi-
bility on substantive complexity was found to be even more
pronounced.

Given the high stability of intellectual flexibility
scores over the ten year period, Kohn and Schooler inter-
preted these findings as impressive evidence of the imme-
diate effects of substantive complexity. The causal effects
of intellectual flexibility on substantive complexity of the
job, however, proved even more pronounced and occurred more
gradually over time than the opposite effect. They con-
cluded, "Current intellectual flexibility has scant effect
on current job demands, but it will have a sizeable effect
on the further course of one's career."

The authors claimed their findings came down solidly
in support of those who see occupational conditions as
affecting personality. They didn't deny the fact that indi-
vidual personality is a major determinant in job selection,
but this, they claimed, was not seriously at issue. They
suggested their results offer clear evidence that one
dimension of social structure, substantive complexity of
work, affects personality not only during childhood social-
ization, but also throughout adulthood.

These conclusions, however, might be a bit overstated.
The researchers found that the influence the job had on

their personality variable was only contemporaneous and

 

given the constructs they dealt with the results seem to be

l3

quite logical. Surely, a complex job (as opposed to one
which is simple) would seem to have immediate effect on an
individual's flexibility in attempting to cope with the
intellectual and verbal demands of an interview situation.
More convincing evidence of a "job effects personality"
hypothesis would have been indicated by a strong lagged
effect of substantive complexity on intellectual flexibility.

Such an effect was noted in the work of Brousseau (1978),
who also conducted a longitudinal study. His research was
grounded in a situational theoretical framework, much like
that of Kohn and Schooler's (1973), based on the proposition
that "individuals' life orientation and levels of emotional
well-being are influenced by the stimulus complexity of
their job experiences." The job is viewed as a source of
stimuli which has the potential to affect the development of
individuals' capacities for abstract or complex cognitive
processes and, thereby, their perceptual and emotional ori-
entations and goals. Citing the work of Schroder, Driver
and Streufert (1967), the author claimed a high degree of
cognitive complexity should allow the development of a
"proactive orientation" (striving to attain more things)
toward life. This, in turn, would contribute positively to
an individual's self-confidence, self-esteem, and general
emotional well-being.

Brousseau suggests that jobs which entail performing
complex or risk types of work should, over time, enhance

the development of one's cognitive capacities, and contribute

14

to a high level of emotional well-being. This relationship,
however, may be moderated by certain factors. An example

the author offers is the "extent a worker is cognitively
complex prior to taking a job." Brousseau feels that cog-
nitive complex individuals, because of a propensity to shape
their circumstances and perceive numerous alternative courses
of action, would involve themselves in "richer" experiences
in the non-work sector of their lives.

Brousseau collected longitudinal data from a sample of
116 scientists and managers working at a large petroleum
products company, with a five year lag, in an attempt to
demonstrate support for these theoretical relationships.
Five characteristics of employees' jobs were gathered as a
measure of job complexity, using the Job Diagnostic Survey
(Hackman & Oldham, 1975). Both individual worker response
and mean response for particular manpower categories were
analyzed. Four personality scales derived from a short
version of the Guilford-Zimmerman Temperament Survey were
collected to measure, (1) active orientation to life,

(2) philosophical orientation, (3) freedom from depression,
and (4) self-confidence.

Results from partial correlation analyses provided
moderate support for his hypotheses. The use of longitu-
dinal data and partial correlations analyses ruled out
"selection effects" as an explanation of the relationship
between job characteristics and personality. Individual

JDS score data and mean manpower category JDS responses

15

indicated that changes in two personality subscales--"active
orientation" and "freedom from depression" were positively
associated with the "Task Significance" and "Feedback from
Job" JDS subscales.

Brousseau hypothesized that job tenure acts as a mod-
erator of the job experience-personality relationship,
and thus associations between job characteristics and
personality change would increase in magnitude as function
of time. His findings, however, provided only weak (at best)
evidence of such an effect. Partial correlation coefficients
between Job Diagnostic Survey scores and post-test scores on
"Active Orientation" and "Freedom from Depression" scales
were more highly positive for the high tenure group than for
the low tenure group.

Although Brousseau used a better design than the pre-
vious research in this area, some aspects of his study should
be carefully scrutinized. The use of such a limited, high
status sample rather limits the generalizability of his
results and the variability of JDS responses. The use of
four factor analyzed dimensions of the Guilford-Zimmerman
Survey with no mention of construct validity should be ques-
tioned, especially since the author seemed to have difficulty
defining his factors.

Another difficulty with the Brousseau study was the
application of partial correlation analyses to his data. To
be able to totally control for a variable, perfect relia-

bility must be assumed. Given that his personalitqrconstructs

16

at time one were measured with measurement error (clearly
not an unreasonable assumption) his partial correlation
coefficients are subject to bias. Another, more pressing
problem, with partial correlation analysis is the inability
of the procedure to test for spuriousness, by which is meant
that the relationship between two variables is not due to
the causal effects of either but to the effects of a third
variable. The possibility of both personality and job expe-
rience reflecting causal variables outside of the work
sphere is a tenable hypothesis. For example, socio-economic
status has been reported to be related to both personality
and occupational choice (Kohn & Schooler, 1969; Kohn, 1969).
It is quite possible that socio-economic status or some com—
bination of other variables are the causal elements which
influence both personality and perceptions of job experience.
Despite the problems associated with the "job affecting
personality" research, these studies do suggest that the
type of work an individual experiences has the potential of
influencing certain aspects of personality. Studies reviewed
to this point have investigated the influence work has on an
overall index of mental health, intellectual functioning as
evidenced by a measure of intellectual flexibility, and four
factor analyzed dimensions of the Guilford-Zimmerman Tem-
perament Survey. Surprisingly absent from this list, how-
ever, are more widely accepted, more thoroughly understood

aspects of personality.

17

The author feels that processes by which the job expe-
rience might effect or reflect personality will be more
fully understood in a study that investigates certain aspects
of personality, rather than employing the global measures of
emotional well-being that have been used in the past. Of
the many personality theories which abound in psychology
today, a large number make use of the concept of traits. In
many theories, the organization of those traits constitute
the personality. Rather than searching for causal processes
between job experience and abstract definitions of person-
ality, it is important to first explore some of these trait
dimensions of personality.

A wide variety of personality dimensions have been
employed to enrich theories of behavior in organizations.
Two dimensions, locus of control and self-esteem, have been
theorized and empirically shown to be particularly relevant
to the work situation. For this reason I have chosen to
examine these personality constructs in relation to job
experience. Following is a discussion of these constructs,
reviews of how these variables have been related to individ-
ual behavior in the work setting and discussions of how
these variables might be causally related to the work

experience.

Locus of Control and Job Experience

 

The internal-external locus of control construct, as

derived from Rotter's (1954) social learning theory, refers

18

to the extent to which an individual perceives that success
or failure is dependent upon his own initiative as opposed
to being the result of fate, luck, chance or powers beyond
one's personal control. High "internals" perceives a rein—
forcing event as dependent upon their own behavior. On the
opposite end of the continuum, high "externals" perceive
what happens to them as being determined by forces over
which they have no control. Literally hundreds of studies
investigating this personality variable are in the psycho-
logical literature (see reviews by Phares, 1976; Rotter,
1966; Strickland, 1975).

The literature does provide strong evidence that inter-
nals do exhibit more initiative and competence in attempts
to control their environment (e.g., Joe, 1971; Lefcourt,
1972; Phares, Ritchie & Davis, 1968; Rotter, 1966; Seeman &
Evans, 1962). This initiative is evident in Valecha's (1972)
finding that internals tend to be better informed about their
occupations. In a study reported that same year, Pines and
Julian found internals in problem-solving situations to be
particularly oriented toward gathering and processing infor-
mation while externals seemed more concerned with social
requirements and doing what was expected of them. Organ and
Green's (1974) results suggest that internals are indeed
successful at controlling their environments. Internals
experienced significantly less ambiguity about their work

roles than did externals.

l9

Hamner and Organ (1977) note that this phenomenon is
logical. If individuals believe that their rewards are con-
tingent on their behavior, they will place higher values on
and actively search for strategic kinds of information and
knowledge. This would seem to indicate that internals would
be most attracted to job experiences in which they were able
to receive information that would indicate how well they
were performing. Such jobs, for example, could be charac-
terized by much performance feedback and, perhaps, by com-
pletion of a "whole" and identifiable piece of work (because
a "whole" piece of work provides much more information as
to how well an employee performed, as opposed to simply
completing a part of that job). In addition, one could
hypothesize that internals, in an effort to control their
environment, would seek out highly autonomous jobs, in which
they could determine work schedules and procedures. One
could also hypothesize, based on Pine and Julian's (1972)
findings that externals will seek out positions in which
requirements are clearly elaborated. For example, work
experiences found to be high in autonomy, and low in feedback
might prove threatening to the high external and would be
avoided by them.

On the other hand, one should not conclude that a large
correlation between locus of control and job characteristics
means support for the selection hypothesis. As opposed to
"selecting into" jobs, it is also quite possible that expe-

rience in a job with certain characteristics could alter

20

locus of control. For example, Rotter (1975) attributed an
increase in external scores in college populations during
the late sixties to increased feelings of alienation due to
societal events. This same process could be happening in
the work environment, those employed in highly authoritative
working conditions which provide little or no feedback as to
performance could experience an increase in externality. In
a more positive light, those employed in highly enriched jobs,
which provide feedback and autonomy, could become internals,
because they have a means to receive information necessary
to control their environment.

Keeping within the boundaries of Brousseau's theoretical
framework, one could argue that a highly complex job could
affect locus of control orientation because the high source
of stimuli would enhance one's cognitive capabilities.

Citing Schroder, Drive and Streufert(l967), Brousseau argued
"a high degree of cognitive complexity provides the individ-
ual with the capacity to generate more elaborate plans and
goals and a wider array of options for dealing with circum-
stances." Those with highly developed cognitive abilities
are able to perceive a greater variety of courses for coping
with situations (Lefcourt, 1975). Wolk and Ducette (1974)‘
have presented evidence which suggest that internals show
marked superiority over externals in amount of both inten-
tional and incidental learning. They regard this incidental
learning ability as a cause, rather than effect of internality

despite the fact their research was of a correlational nature.

21

Within this framework, one could reason that the increased
cognitive complexity of the employee in the enriched job
experience would result in a higher internal locus.

A study which did concern this causality issue was con—
ducted by Andrisani and Nestel (1976). They found that locus
of control systematically influenced success in the world of
work, independent of individual differences in skills, abil-
ities, and demographic distribution. Also, they found evi-
dence that advancement in occupational status, advancement
in annual earnings and reentry into the labor force is sys-
tematically related to increasing internal control over a
two-year span. Despite the fact the researchers dealt only
with measures of success in the job experience, their find-
ings suggest that locus of control is responsive to employ-
ment experience. The authors' note the consistency of this
finding with their hypothesis that "unfavorable work expe-
riences are thought to increase tendencies toward external

control."

Self-Esteem and Job Experience
The construct of self-esteem has been the subject of
much attention in psychological literature. In fact, reviews
of the literature (i.e., Wylie, 1974), suggest that this
trait has been related to almost every variable imaginable.
Care should be taken in comparing much of this research,
however, as Gergen (1971) warns that the term has been used

in a variety of ways by different authors.

22

Much of the work done with this construct in Industrial/
Organizational literature has employed Korman's theory of
work behavior which postulates three sources of self-esteem.
Chronic self-esteem is seen as "a relatively persistent per-
sonality trait that occurs relatively consistently across
various situations" (Korman, 1970, p. 32). Task-specific
self-esteem is described as a function of experience with
identical or similar tasks. Socially influenced self-esteem
is a function of others' expectations of one's behavior.
Although this trichotomized conceptualization of self-esteem
has proved invaluable in motivation and satisfaction research,
the present study is concerned only with chronic self-esteem.

The approach to this construct that I take is influenced
by Coopersmith's (1967) and Rosenberg's (1965) work. Like
these authors, I refer to self-esteem as the individuals
judgment of their self-worth. I will treat it as a global
dimension that is resistent to change in the short run.

Brousseau's (1978) theoretical framework for his 1:99”
.effects’pgrsgnality: hypothesis predicted higher levels of
self-confidence and self-esteem for those who have expe-
rienced high stimulus jobs. As noted earlier, the more
highly cognitive capabilities which they felt would result
from such experience would have a positive influence on
ability to perceive a greater variety of courses of action
for coping with difficult situations, which Lefcourt (1973)
found to lessen anxiety and emotional distress. Brousseau

believes this lessening in anxiety and emotional distress

23

would contribute positively to an individual's self-esteem.
This reasoning seems to stem from the fact that so many
researchers have found self-acceptance to be negatively
associated with anxiety measures (e.g., Cowen, Heilizer,
Axelrod, & Alexander, 1957; Hanlon, Hofstaetter & O'Conner,
1958).

Support for this hypothesis can be found in the work of
Bachman and O'Malley (1977). Employing path analysis to
longitudinal data, these researchers found that occupational
attainment had a modest, but direct positive impact on self-
esteem. On the other hand, they concluded that high school
self-esteem had little or no direct causal impact on later
educational or occupational attainment. Given the fact that
occupational status has been shown to be related to more
complex jobs (Fitzgerald & Schmitt, Note 1), Bachman and
O'Malley's data would seem to support the hypothesis that
job experience affects personality.

Another perspective on how job experience might relate
to self-esteem stems from Korman's (1966, 1967, 1969)
research, which supports the hypothesis that self-esteem is
a determinant of occupational choice. In these studies,
support was found for the prediction that chronic (as opposed
to task-specific and socially influenced) self-esteem is pos-
itively related to seeking out and choosing occupations which
are seen as satisfying one's self-perceived needs.

Reasoning for this phenomenon is grounded in self-

consistency theory. High self-esteem individuals, who have

24

had needs satisfied in the past will seek out situations in
which these needs can be satisfied in the future. On the
other hand, low self-esteem workers who have not been able
to satisfy needs in the past are more likely to become more
familiar with non-need satisfying situations (Korman, 1969).
Hence, it is predicted that self-esteem will influence per-

ceptions of job experience.

Focus of Study

 

The present study is designed to explore causal rela-
tionships between the two personality traits, locus of con-
trol and self-esteem, and perceived job characteristics.
Previous research and practice concerning this relationship
have provided support for a "personality effects job"
hypothesis, a "job effects personality" hypothesis and
another hypothesis which proposed that the two affect one
another. In lieu of formulating specific hypotheses in
accord with one of the above, this researcher has conducted
an exploratory study which has tested for the possibility
of any causal relationships (including reciprocal) between
the variables of interest.

Given this open framework, the present study is dif-
ficult to classify in one of the theoretical personality
models discussed earlier. In fact, the major assumptions
of all three models were investigated: (1) do traits deter-
mine the situation (job experience), (2) does the situation

determine the traits, or (3) is there evidence that not only

25

do events affect the behavior of individuals, but the indi-
vidual is also an active agent in influencing environmental
events (reciprocal causation)?

In addition, the proposed research was designed to
avoid the interpretational difficulties of the cross-
sectional nature of the early studies and the inadequate
analyses of later longitudinal studies concerning causality
between personality and job experience. Accordingly, a
longitudinal design was used and the resulting panel data
analyzed. To eliminate the problems which Brousseau's par-
tial correlation analytic procedure imposed on his con-
clusions, a causal analytic technique "better adapted for
panel data analysis" (Kenny, 1975), which assumes measure-
ment error, was utilized.

This technique can also evaluate evidence of reciprocal
causation, as suggested by Kohn and Schooler's (1978)
research. As a further test to determine which variable has
the greatest influence and whether any causal relation-
ships are negative or positive, the frequency-of—change-in-
product-moment (FCP) were used. To determine the possibil-
ity that additional variables are causing the variables of
interest to cdyary, a possibility which remained untested
in both the Brousseau and Kohn and Schooler analyses, the
present study also employed Vroom's dynamic correlation

analysis.

26

Job Tenure as a Moderator

Brousseau (1978) hypothesized that the causal relation-
ship between change in personality and job characteristics
will become more pronounced as job tenure increases. Sup-
port for this hypothesis in his research, however, was very
weak. In their study of career stages, Hall and Nougaim
(1972), found the new organizational member to be more con-
cerned with "defining an identity" than longer-tenured
workers. This finding would seem to indicate that the
greatest amount of interplay between personality and work
would occur during the early stages of job experience.
Indeed, Katz's (1968) findings suggest that the strongest
attitudinal responses to job characteristics occur between
the first and third years of job tenure. The possibility
of stronger causal relationships between the job and person-
ality during the early stages of an individual's career was
investigated in this research effort. An appropriate sample
was split into low tenure and high tenure subgroups and dif-
ferences in the direction and strength of causal relation-
ships between the subsamples was evaluated.

In summary, the present research is concerned with the
direction of causality between two personality constructs,
locus of control and self-esteem, and job experience.
Exploratory in nature, this study was not confined to any
one of the many theoretical frameworks that surrounds the
job experience-personality relationship. Results which

indicate that self-esteem and locus of control "cause" job

27

characteristics would lend support to much of the work done
in personnel selection and other areas of industrial psy-
chology concerned with personality and work. On the other
hand, findings which point to the reverse hypothesis—-"job
experience effects personality" would add credibility to

some of the theoretical positions discussed in the literature
review.

Results indicating that job experience would effect
personality or, perhaps, a reciprocal causation between the
variables of interest would suggest a dynamic interplay
between job and personality. The fit between an individual
and his or her job would then have to be reviewed as a
dynamic rather than as a static relationship. This would be
of considerable interest to those involved in personnel test-
ing, job design and career counseling. The type of job peo-
ple are best suited for at the beginning of the job expe-
rience, may not be the type for which they would be best

suited in the future (Brousseau, 1978).

METHOD

Subjects
As suggested by Kenny (1975), longitudinal designs

should include replications across different groups of
subjects. Hence, two samples were employed in this study.
The first consisted of 120 full-time employees, all newly
hired into a variety of State Civil Service jobs and a large

service industry. This sample, referred to as the new
hires" hereinafter, was chosen because of the diverse vari-
ety of jobs and the short length of job tenure of these
individuals, all being with their respective organizations

for one year or less at the time of the first data col-
lection. The average age of the 71 females in this sample

was just under 25, while the mean age of the remaining 49
males was just under 22. The job level of the group was
generally low, averaging the equivalent of a clerical job,
though some low-level professionals were included. Sur-
prisingly, however, the mean education level of the group
indicated some college education for a majority. In addition,
the workers seemed to be adequately paid, with a mean hourly
wage of six dollars.

The second group analyzed included 331 full-time

employed individuals attending a vocationally related program

28

29

of study at Lansing Community College (LCC). This group
contained an even wider variety of jobs than the first sam-
ple. The 118 females and 112 males in this sample had a
mean age of 33. This group was surprisingly similar to the
first in both mean level of education (one or two years of
college study) and mean hourly wage (approximately six dol-
lars an hour).

The LCC sample was divided into two subgroups on the
basis of job tenure. As suggested by Katz's (1978) work,
those with two years or less job tenure at the time of the
first data collection were placed in the "low" job tenure
group. The LCC vocational education sample was divided into
130 low tenure subjects and 200 high tenure. When dichot-
omized as such, the groups differed very little on a number
of demographic variables, such as the size of community in
which they reside, amount of schooling, sex, and race. On
the other hand, expected differences in age and hourly wage
were found; the high tenure group having a mean age of 34,
two years higher than that of the low tenures and a mean
hourly wage of seven dollars an hour, fifty cents higher

than that of the low tenures.

Measures

Job Experience

 

One of the problems in comparing research in this area
is the variety of measures of job experience used by the

different researchers./ This is understandable given the

30

different theoretical frameworks these authors have used.
For example, substantive complexity of the job, as developed
and utilized by Kohn and Schooler (1973, 1978) was based on
ratings of the degree to which thought and independent judg-
ment were used, judged by sociologists. On the other hand,
Kornhauser's (1965) work was based simply on experience in
jobs of different skill levels.

The same problem would seem to hamper current efforts
to test differing causal relationships between the person-
ality traits of interest and job experience. As noted in
the sections dealing with these constructs, type and amount
of feedback, level of autonomy, level of need satisfying
ability and stimulus complexity are all dimensions of the
job experience which hypothetically could reflect or affect
personality.

The Job Diagnostic Survey (Hackman & Oldham, 1975),
however would seem to provide a solution to this dilemma.

A version of the short form of this instrument, which was
used to assess workers' perceptions of the characteristics
of their job, provides much of the information needed to
test the various theories discussed earlier. Most impor—
tantly, Brousseau (1978) argues that the characteristics
measured by the instrument are closely related to those he
feels would contribute to the stimulus complexity of indi-
viduals' job experiences. The JDS has been used success-
fully as a measure of complexity (vs. simplicity) by

Ivancevich (1978), also.

31

The short version of the JDS measures five "core" job

dimensions:

Skill Variety. The degree to which a job
requires a variety of different activities
in carrying out the work, which involve the
use of a number of different skills and

talents of the employee.

Task Identity. The degree to which the job
requires completion of a "whole" and iden-
tifiable piece of work--i.e., doing a job

from beginning to end with a visible outcome.

Task Significance. The degree to which the
job has a substantial impact on the lives

or work of other people--whether in the imme-
diate organization or in the external

environment.

Autonomy. The degree to which the job pro-

vides substantial freedom, independence, and
discretion of the employee in scheduling the
work and in determining the procedures to be

used in carrying it out.

Feedback from the Job Itself. The degree to
which carrying out the work activities required

by the job results in the employee obtaining

32

direct and clear information about the

effectiveness of his or her performance.

In the analyses, each dimension of these core job char-
acteristics was explored. In addition, an overall index of
job complexity was derived from these scores. This score
provides a summary measure of job experience. As suggested
by Brief, Wallace, and Aldag (1975) and Dunham (1976), a

simple linear combination was employed.

Locus of Control

 

This construct was measured by 11 items, which were
responded to using four-point Likert-type scale of agree-
ment, as developed by Herbert S. Parnes (Andrisani & Nestel,
1976). The scale constitutes an abbreviated version of
Rotter's (1966) Internal-External Locus of Control Scale.
Parnes selected items from the original scale on the basis
of their appearance to be more general, adult-oriented and

work-related (see Appendix B).

Self-Esteem

 

This trait was measured by the ten-item scale developed
by Rosenberg (1965), to measure the self-acceptance aspect
of self-esteem. Rosenberg treated self-esteem as a global
concept rather than as a number of specific ones and this
approach is apparent in the content of his items. None of
the questions deal with task-specific situations or the

reactions to the expectations of others (socially—influenced

33

self esteem). For this reason, this scale provides a mea-
sure of Korman's generalized ("chronic") self-esteem (see

Appendix C).

Procedure

 

All scales of interest to the present study were a part
of a larger questionnaire mailed to the subjects on two
occasions, approximately one year apart. On both occasions,
subjects were paid three dollars to return the surveys.
Return rates were approximately 40% on both occasions for

both samples.

Data Analyses

 

Cross-LagiAnalysis

 

The cross-lag panel correlation analytic technique
initially discussed by Simon (1954) was utilized, using
corrections for changes in reliability in the variables
over time and the Pearson-Filon test of differences
between correlations described by Kenny (1975). This tech-
nique has been refined and utilized successfully by a num-
ber of researchers, including Greene (1979), Ivancevich
(1978), Vroom (1966) and Lawler (1968). Feldman (1975) and
Kenny (1975) have carefully specified the limitations of
cross-lagged analysis; their suggestions were incorporated
in this analysis. Most importantly, the technique does not
establish causality in the way an experimental study would,

but it does establish the most likely direction of causality,

34

particularly when competing hypothesis concerning causality

are available (Feldman, 1975), as is the case in the present

 

 

   
  
  
 
 
 

 

 

study.
Cross-Lag Analysis of Overall Job Diagnostic
Survey Score (JDS) and Locus of Control (LOC)
r
JDSl JDSlJDS2 JDS2
, (autocorrelation) ///////’
r
JDSzLOCl
(cross-lag correlation)
r r
JDSlLOCl JDSZLOC2
(synchronous (synchronous
correlation) correlation)

r
JDSlLOC2

(cross-lag correlation)

\

 

 

r
LOCl LOClLOC2

(autocorrelation)

LOC

 

 

Figure l

The cross-lagged panel technique requires measures of
the variables be taken at two points and the computations
of six correlations for each pair of the variables of inter-
est (see Figure 1). For example, correlations between over-
all (JDS) score and locus of control (LOC) at the same point
are called synchronous correlations, the correlation between
JDS at time one and JDS at time two and LOC at time one and

LOC at time two are autocorrelations, and correlations

35

between JDS and LOC across time are the cross-lagged cor-
relations. If JDS causes LOC then the magnitude of the

> (rJDS LOC =

correlations should be such that: rJDS LOC 1

1 2 1

ZJDSZ) > rJDszLocl. It is important to test for the

equality of the synchronous correlations. The cross-lag

rJDS

model demands this equality before a valid interpretation
of cross-lag differences can be made. If LOC causes the
JDS score then the positions of rJDleoc2 and rJuszLoc1
reversed. In addition, evidence of a reciprocal relation-

are

ship between the variables of interest would be found if
both cross lagged coefficients were approximately equal and
larger in magnitude than the synchronous correlations. For
example, rJDleoc2 = rJDSZLOCI > rJDleoc1 = rJDSZLOCZ.

The hypotheses tested in cross-lagged analysis are,
first, the equality of synchronous correlations to test for
stationarity and, second, the equality of cross lags to
test for spuriousness. One cannot use Fisher's Z trans-
formation to test for the significance of the differences
between these correlations since the correlations are cor-
related. Instead, a transformation cited by Peters and
Van Voorhis (1940) and attributed to Pearson and Filon may
be used to test for differences. A statistical test of
reciprocal relations in the cross lag technique, to test for
the feasibility of the reciprocal hypothesis, has not yet
been offered in the literature. Therefore, if this situation

should arise in a cross lag analysis, a subjective judgment

must be made.

36

Theoretically, if all the assumptions of the cross-lag
analysis procedure have been met and equal cross-lags result,
a null hypothesis of spuriousness might be accepted. That
is, the variables of interest do not cause each other,
but are co-symptoms of some set of common causes. Although
this is the null hypothesis of the test of equality of
cross-lagged coefficients, alternate explanations for equal
cross-lags are possible. For example, causal relationships
might be too small to be detected, the measured cross-lag
might not correspond to the actual cross-lag and, as men-
tioned, there is the possibility that the variables could
cause each other.

In many cases, synchronous correlations have been
shown to be unequal due to attenuation by measurement error.
Kenny (1975) refers to this situation as quasi-stationarity
and claims that the variable with the higher autocorrelation
will appear to Operate as an effect and the variable with
the lower autocorrelation will appear to be the causal
variable. He has presented a quasi-stationarity correction
procedure which is applied to the cross-lagged correlations
prior to the calculation of the test of the difference of
these two correlations. This correction method involves
the calculation of a reliability ratio. A reliability ratio
greater than one indicates an increase in reliability over
time, while a value less than one suggests a decrease in

reliability. The calculation of the reliability ratio

37

requires at least three variables; in this study--JDS scores,

self-esteem, and

Reliability

JDS

LOC

locus of control.

ratios were calculated as follows:

r r r
JDSZSE2 JDSZLOC2 SElLOC1

r r r
JDSlsE1 JDSlLOC1 SEZLOC2

 

r I" r
SEZJDS2 LOCZSE2 JDSlLOC1

r
1 1 1 1 JDSzLocz

 

 

r r r
LOCZJDS2 LOCZSE2 JDSlSEl

r r r
LOClJDS1 LOClSEl JDSZLOC2

 

The reliability ratios were then used to correct the

observed cross-lagged correlation coefficients as follows:

rJDS LOC

1

rJDS LOC

2

7 (rJDS

(K LOCZ)

2 JDS/KLOC) 1

k (rJDS

Locl)

1 ’ (KLOC/KJDS) 2

Dynamic Correlation Analysis

 

As an indication that additional variables are causing

the two variables of interest, Vroom's (1966) dynamic cor-

relation coefficient were examined. This coefficient is

calculated by correlating the change in X from time one to

time two with the change in Y over the same time interval.

According to Vroom, the stronger this correlation is, the

lower is the probability that the covariance in X and Y can

be attributed to

the effects of a third variable.

38

Researchers have suggested that dynamic correlations
will be biased whenever the scores on the measurements of
the variables of interest at time two regress toward the
mean of the time one scores. When this form of regression
toward the time one mean does occur, the time one scores
on a variable will be negatively correlated with the change
scores on the same variable. To correct for this problem,
partial correlations were used to compute the dynamic cor-
relations, holding the time one scores of each variable
involved constant.

As an example in the present study, the dynamic cor-
relation coefficient is the correlation between the dif-
ferences between JDS scores over time and the differences
between locus of control scores over time. It is assumed
that the stronger the correlation, the less likelihood
there is that changes in the two variables of interest are
both caused by a third variable. To correct for the neg-
ative correlation in dynamic correlations which results from
regression of time two scores toward the mean of time one
scores, partial correlations were computed. For example,
the dynamic correlation between JDS score and LOC were com-
puted by holding initial values of these two variables con-
stant. Given significant differences in cross-lagged cor-
relations, a significant, large dynamic correlation would
suggest the existence of a causal relationship between the

variables of interest.

39

Frequency in Change of Product Moment Analysis

 

Feldman (1975) noted that another problem encountered
with cross-lag analysis is the limited number of causal
inferences which the analysis makes possible. For example,
it is not possible to distinguish between the source and
direction of influence of two correlated variables. Of con-
cern to this study, in which reciprocal relationships are
possible, cross-lags are not able to determine which variable
had the greatest influence. The frequency-in-change-of—
product-moment (FCP) technique was developed by Yee (1968;
Yee & Gage, 1968) to overcome these problems and it was
employed in this study.

The FCP technique requires that the data collected for
each employee to be placed into one of four categories. For
example, with JDS score and locus of control score, the data
were placed into a JDS+, JDS-, LOC+, LOC- category based on
the following steps:

1. The time one and time two raw scores for LOC and

JDS were converted to standard scores. Thus,

(£2)
5

2. The direction of influence, positive or negative,

Z = was computed for each score.

was identified for each case by determining and
evaluating the four cross-product Z-scores for

each subject:

Z
(a) JDSlLOCl

Z
(b) JDSZLOC2

40

Z

(c) JDSlLOC2
(d) zJDS LOC
2 1
If it was found that zJDSZLOC2 > ZJDSlLOC1 and
zJDSlLOC2 > ZJDSZLOCl, then the direction of change

is considered positive (+) and job characteristics
are identified as the primary positive influence

(JDS+). If, however, ZLOClJDS2 > ZLOCZJDS1 was

found, then the direction of the change would still
be positive, but locus of control is considered the

source of positive influence (LOC+). Similarly, if

Z Z Z Z
JDSlLOCl > JDSZLOC2 and JDSlLOCZ > LOClJDS2 was

found, then the direction of influence is negative
and job characteristics are identified as the pri-

mary negative influence (JDS-). Finally, if

Z Z
LOClJDS2 > JDS1

of change is negative and locus of control is

LOC2 was found, then the direction

identified as the negative source of influence
(LOC-).

After each of the cases is classified as JDS+, JDS-,
LOC+, or LOC- three chi-square significance tests
were used to determine the source and direction of
causality. If {(JDS+) + (JDS-)} > {(LOC+) +
(LOC-)}, then job characteristics was identified
as the primary source of influence (Test One).
Secondly, if (JDS+) > (LOC+), then the direction

of causal influence from JDS was to increase the

41

correlation between JDS and LOC (Test Two). Finally,
if (JDS-) > (LOC-), then the direction of the causal
influence from JDS was to decrease the correlation

between JDS and LOC (Test Three).

Combination of Subsamples

The significance tests associated with the cross-lag
analyses require large sample sizes for adequate power. For
this reason, the low tenure (LCC) sample and the new hire
sample were combined. Box's (1949) test of equality of the
variance/covariance matrices was used to determine the com-
parability of the samples before combination. In a test
involving the JDS subscales locus of control and self-esteem,
the observed variances/covariances were shown to be not
different x2(28) = 2.2489, g > .05. In a test which
included the summed JDS score, locus of control and self—
esteem, no significant difference was found x2(6) = 2.8801,

p > .05. Thus, the samples were combined.

RESULTS AND DISCUSSION

Reliability and Dimensionality of Scales

 

Kenny (1979) has noted the fact that the "more valid,
reliable, and unidimensional the measure, the more straight-
forward is the interpretation (of the cross-lag panel cor-
relations)." In Table l, coefficient alpha and test-retest
correlations for all scales in each sample are displayed.
Overall, measures of internal consistence appear reasonable.
Of special interest, however, are those scales in which a
marked change in reliability is displayed. For example, in
the total LCC group, the locus of control soale's coefficient
alpha changes from .536 at time one to .736 at the time of
the second data collection. As mentioned in the data anal-
ysis section, these differential reliabilities of the time
one and time two measures can greatly bias the comparison
of cross-lagged correlations. Campbell (1963) first pointed
out that variables which increase in reliability will appear
to be effects and variables that decrease in reliability will
appear to be causes. Given the psychological nature of the
scales dealt with in this study, a "quasi-stationarity"
model is proposed. This model assumes measurement error
has attenuated synchronous correlations between the JDS sub-

scales and the psychological scales. This assumption allows

42

43

TABLE 1

Reliability Coefficientsa

 

 

. a a Autocorrelation
Time I Time II (Test-Retest)
New Hires
Self-esteem .855 . .877 .630
Locus of control .746 .711 .603
Autonomy .760 .839 .501
Skill variety .788 .778 .616
Feedback from job .681 .764 .377
Task identity .764 .775 .369
Task significance .643 .666 .596
Summed JDS .868 .900 .618
Feedback from others .798 .847 .484
LCC (total)
Self-esteem .765 .848 .502
Locus of control .536 .736 .584
Autonomy .744 .777 .465
Skill variety .833 .797 .644
Feedback from job .613 .704 .435
Task identity .749 .707 .478
Task significance .613 .642 .453
Summed JDS .870 .875 .588
Feedback from others .791 .826 .468
LCC (low tenure)
Self-esteem .776 .866 .571
Locus of control .589 .548 .607
Autonomy .766 .827 .447
Skill variety .852 .819 .651

Feedback from job .648 .778 .446

44

TABLE 1 (cont'd.)

 

 

a a Autocorrelation
Time I Time II (Test-Retest)
LCC (low tenure) (cont'd.)
Task identity .747 .713 .553
Task significance .641 .636 .379
Summed JDS .887 .889 .586
Feedback from others .829 .854 .411
LCC (high tenure)
Self-esteem .760 .834 .454
Locus of control .523 .746 .572
Autonomy .728 .734 .481
Skill variety .817 .783 .651
Feedback from job .588 .657 .424
Task identity .749 .703 .419
Task significance .590 .652 .511
Summed JDS .858 .866 .591
Feedback from others .762 .805 .495
Combined
Self-esteem .818 .872 .597
Locus of control .671 .717 .605
Autonomy .765 .834 .479
Skill variety .821 .806 .631
Feedback from job .666 .784 .408
Task identity .755 .749 .459
Task significance .642 .647 .483
Summed JDS .878 .899 .603
Feedback from others .813 .851 .405

 

45

for the correction of the cross-lagged correlations for
changes in reliability over time.

Principal components analyses of the locus of control
and self-esteem scales were performed in each sample. Only
the locus of control scale in the new hire sample proved
more than unidimensional. Examination of the rotated fac-
tor matrix, however, showed items loaded perfectly on two
factors differentiating internal and external items. For
this reason, I have interpreted this principal components
finding as a result of a response bias, rather than as evi-

dence of multidimensionality in the construct.

Locus of Control

 

Before any valid comparison of the corrected cross-lag
coefficients can be made, one must first test for the plau-
sibility of the quasi-stationarity assumption. This was
accomplished by testing for differences between the cor-
rected synchronous correlations within each relationship to
be examined. Appendix D through H contain results of the
cross-lagged analysis, corrected dynamic correlations and
FCP analyses for the samples explored in this study. One
can see, by comparing corrected correlations between each
synchronous correlation at the Same point in time, that the
quasi-stationarity assumption has been met in all relation-
ships involving locus of control in all samples. Signif-

icant cross-lag results are summarized in Table 2.

46

Ammm.v ovm. Avmm.v«vmm. “may

 

 

who. AmHN.V Ham. Ammm.v Hum. AHBV
Amaoom meEdmv moo
. . . onEmm
Aqua v mma AmHN vsmmm ANBV cofiumoscm
vvo. Aboa.v Hmo. Amma.v mmH. AHBV Hmcoflumoo>
cosmoflmwcmwm xmme UUQ Hmuoa .m
Amoa.v «ma. AHNO.V 0H0. ANBV
«moa. Amwa.v«mam. Ammo.v hmo. AHBV
suaucmoH xmoe
Amom.v «ma. Roma.v who. Away
mao.l Amha.vamhm. Aqmo.v oma. AHBL
sooaum> Haaxm
Amom.v mmm. Amaa.v mmo. “may
ova. Anem.vamem. AmoH.o oom. “flee oamsom
>Eocous< muw: 362 .H
Amev “Hey
mcoflumHmuuou Houucoo mo moooq mamom moo onEmm

oflEmc>Q cwuomuuou

 

Amfimwnucwumm cfl Umuomuuoocsv
mcoHumHouuou msocoucoc>m cam
mcommmqnmmouu Uwuomuuoo

v1.1.1.1.- ’-1-.-!-i1ll 1‘4111- I. III-

mofiumflumuomumnu poo co>fimouom can
Houucou mo msooa comzuom moonlmmouo acoummmao >HuchAMAcmHm

N mam¢9

47

wouoouuoo cooSumn Amy

mo. v m..."

. .mucwwoflmwooo moanmmouo
pcm mcofiumamuuoo msocousocxm oouoouuoo coo3umn AHV mo cocooflmacmfim mo unmao

 

 

 

Amoa.v mna. Ammm.vsmvm. ANBV
omo. AmmH.v moa. Amom.v mmH. .Ha.
cosmoflmwcmfim xmme
Aom~.v hmH. Amha.v wma. ANBV
Hoo. Amvm.vswom. Amma.. mmm. “HBO onEmm Auuqv
suauomoH shoe assume sea: .4
AHmH.V mam. th~.vsham. “may
mao.| Ahma.v moH. Aoam.v oom. AHBV mamamm 004V
Spoauo> HHflxm ounces 30A .m
Amev Aaeo
mCOADMHouuoo Houucoo mo msooq onom mob onEmm

oHEmcxo couomuuou

 

Amwmmnucmumm cw wouomuuoocov
mcoflumHmuuou mnocounoc>m can
mcoommqnmmouu wouomuuou

 

..o.usoov m mamas

48

In the new hire sample, marked differences in the cross-
lags show evidence of perceptions of autonomy, skill variety
and task identity causing locus of control in a nonspurious
manner. These are clearly tenable findings, supporting
Brousseau's (1978) theoretical framework in which a job with
more complex activities that would require individuals to
deal with high levels of stimuli would affect their locus
of control orientation. The cross-lags of the new hire sam-
ple would appear to confirm this hypothesis, given the sig-
nificant differences found between cross-lags involving the
summed JDS scale.

Examining the corrected dynamic correlations in the
Appendices, spuriousness is evident. Despite the fact that
two of these correlations are statistically significant,
all are very low, which would indicate that other variables
are causing both the perceptions of the job and locus of
control.

James, Hornick and Demaree (1978) have noted that such
discrepancies between cross-lag results and dynamic cor-
relations are not uncommon. In fact, these authors have
concluded that "given the usual condition in which the cross-
1ags and synchronous correlations have the same signs, low
to moderate dynamic correlations may occur in a number of
situations, including conditions of spuriousness and con-
ditions where causality is a strong possibility." The
authors recommend that, in general, the dynamic correlation

should not be employed as a test of spuriousness.

49

Assuming the corrected dynamics are highly fallible
indicators of spuriousness and we reject the possibility of
spuriousness in these relationships, the interpretation of
the source and direction of the causal relationships are
still impossible given the nonsignificant results of the
LOC >

1 2

rAUTOZLOCI, it is still unclear if perceived autonomy is

causing an increase in an individual's internal locus of

FCP analysis. For example, given the fact that rAUTO

control orientation or that the locus of control orientation
is causing a decrease in the perceived autonomy. Ideally,
the second and third Chi-square tests would give us an indi-
cation as to the positive or negative nature of the causal
relationship found in the cross-lags. Unfortunately, the
FCP analysis, which utilizes a nominal interpretation of

the data, indicates that no causal relationships are to be
found in the cross-lags.

Hopefully, some of these problems of interpretation
would have been alleviated by comparing results in the other
sample. A consistent pattern of cross-lag differences in
the LCC Vocational Education sample would have suggested
actual causal relations. There was, however, no consistency
in the pattern of significantly different cross-lag coef-
ficients was found. Even in the most comparable samples,
the new hires and the low tenure LCC Vocational Education
sample, no consistent pattern of results were found.

In the total vocational education sample, corrected

cross-lags indicate a causal direction from the locus of

50

control orientation to the summed JDS scale and the Task
Significance subscale. When divided into subsamples based
on tenure, the low tenure employee cross-lags indicate that
the locus of control variable causes perceptions of skill
variety. In the high tenure sample, experience with a
position high on task identity is shown to cause a change

in locus of control and locus of control orientation causes
task significance. Once again, the corrected dynamic cor—
relations and FCP analyses do not support these causal find-
ings. Examination of the combined new hire and low tenure
LCC group results reveal no significantly different cross-
1ag coefficients. Given the inconsistency of results among
the different analytic techniques and the samples, the inter-
pretation of the statistically significant cross-lag dif-
ferences is questionable. In fact, these significant dif-

ferences could simply be the result of chance.

Self-Esteem

 

Analysis of relationships concerning the self-esteem
variable proved even more disappointing than those involving
locus of control. Appendices I through M contain results
from all samples and subgroups involving the self-esteem
variable. They show that in all samples, corrected syn-
chronous correlations are not significantly different from
one another, which is a necessary precondition for a valid
interpretation of cross-lag differences. However, signif—

icant differences of cross-lagged coefficients were found

51

only in the low tenure sample. These differences are sum-
marized in Table 3.

These differences indicate that self-esteem causes the
perceptions of skill variety, task significance, and overall
complexity of the job as measured by the summed JDS scale.
Once again, however, the corrected dynamic correlations are
consistently low, indicating spuriousness. In addition, the
FCP analysis revealed no causal relationships between the
JDS subscales and self-esteem. This inconsistency among the
analyses and among samples renders interpretation of the
significant cross-lagged differences questionable.

Strictly speaking, the lack of cross-lag differences
should not lead one to accept the null hypothesis of spuri-
ousness of this method. One alternate explanation for a
lack of consistent cross—lag differences might include the
fact that the variables studied might indeed be causally
related, but the magnitude of the effect is too small to be
detected. The problem of the measured lag not corresponding
to the causal lag, referred to earlier, might also serve to
attenuate any true cross-lag differences. Finally, this
lack of cross-lag differences could also be evidence that
the variables cause each other in a positive, reciprocal
manner. Although this explanation is feasible, in terms of
an interactionist perspective, the low magnitude of the cross-
lag coefficients when compared with the synchronous cor-

relations in all samples rule out this possibility. The

52

mo. v.m¥
.mucowowmmooo @mHnmmouo
owuomuuoo cmmzuwn Amy pom mcoflumamuuoo msocounoc>m couomuuoo cmospmn Adv mo cosmoflmwcmwm mo unmem

 

 

“mom.v can. Aoom.v«~o¢. “may
HvH. Abam.v boa. Ammm.v mum. AHB.
Awamom pofiEflmv mac
Amma.v OFH. Awqm.vsmmm. Away
omo.l A¢HH.V moo. Ammm.v HNN. AHBV
cosmowmacmwm xmme
Amm~.v ham. Aoom.vsvmv. Amav
mmo. Amam.v mud. Avam.v ohm. AHBV oHQEmm
3638, 33m 8Q: amazon. 33
“may .Hev

 

mcowumHouHou Emoummnuawm onom mob onEmm
owsac>o couoouuou

 

Anamonucoumm :« omuoouuoocov
mcowumaouuoo moocounoc>m can
nowmmqnmmouu oouomuuou

moflumflumuomumnu non cw>fiooumm
can Sooummnmamm consumm manqummouo ucmuoumao SHHGMOHMAcmHm

m mqmdh

53

lack of consistent cross-lag differences within and across
samples and the low corrected dynamic correlations adds

credence to a "spuriousness" explanation.

CONCLUSIONS

In this study, an attempt was made to determine the
direction of causality between two aspects of personality,
locus of control and self-esteem, and job experience,as
measured by worker responses to the Job Diagnostic Survey.
Exploratory in nature, the research was designed to examine
possible causal relationships suggested by previous theory
and empirical investigations. Cross-lag, corrected dynamic
correlation and frequency-in-change-in-product-moment cor-
relation analysis were performed to uncover these possible
causal relations. Although significant cross—lag differ-
ences were found results of the other analyses did not cor-
respond with these differences. More seriously, the cross-
lag differences were inconsistent across samples, suggesting
that the few significant differences might have been simply
the result of chance.

Given the inconsistency of the cross-lag differences
and the nonsignificant results of the corrected dynamic and
FCP analyses, it appears that the study suffered from a very
serious problem of spuriousness. Background variables which
could influence both the psychological variables and

responses to the JDS hampered the interpretability of

54

55

possible causal relations in the study and must be included
in future research.

A variety of background variables might have served as
the cause of both the perceived job characteristics and the
psychological variables. For example, socio-economic status
has been highly correlated with both locus of control and
responses to the Job Diagnostic Survey (see Andrisani &
Nestel, 1976; Fitzgerald & Schmitt, Note 1) and could be
cause of both. A strategy to control for such background
variables in a cross-lag analysis is to simply subtract out
the effects of these sources of spuriousness by computing
partial correlations between the relevant variables control-
ling for the background variables.

Although the partialling of background variables might
reduce spuriousness within the relationships in this study,
this problem might remain given the perceptual nature of the
job characteristics measure. Although Hackman and Lawler
(1971) claim the "major determinant of such (JDS) perceptions
is the objective nature of the job," one should not deny the
existence of a perceptual process by which objective task
characteristics are "transformed" into the reported per-
ceived task characteristics. In his discussion of "enact-
ment" processes Weick (1977) states that individuals "enact"
their environments by actively producing reality from their
objective environments and this reality is influenced by
various personal characteristics of the individual. In terms

of this study, the data suggest that the same psychological

56

or sociological processes which influence an individual's
self-esteem or locus of control orientation also alter the
individual's perceptions of the work reality.

The spuriousness evident in this study should lead to
a reevaluation of the past research in this area that has
neglected to investigate this possibility. Brousseau's
(1978) findings, based on partial correlations, did not
address the possibility of other causal variables. A more
dramatic omission of the possibility that an important
variable is unspecified is represented by Kohn and Schooler's
(1978) causal model of substantive complexity and intellec—
tual flexibility, in which error in prediction terms in the
hypothesized model were fixed at zero. This specification,
which presumes that no relevant variables have been omitted
from the model, is unwarranted.

Beyond this obvious problem of spuriousness, actual
causal relationships between personality and job perceptions
might not have been tapped as a result of problems in the
research design. The measurement instruments, all of which
were questionnaire scales utilizing Likert-type formats,
might not have been sensitive enough to adequately evaluate
the concepts dealt with. In retrospect, the use of JDS
scores as a measure of work experience may be inadequate.
To fully appreciate and understand the worker's experience,
other methods of data collection are needed. For example,
the use of time diaries, records of activities and personal

interviews may be more appropriate.

57

The results of these methods could be compared with JDS
responses to determine if the worker's perceptions and behav-
ior actually do correspond. These techniques would also tap
other dramatic changes, outside the work sphere, such as
marital, family and health problems, that might serve as
causal influences on both personality and perceptions of
work experience.

Another design flaw which could have hampered this
research is the use of a one year lag. This lag might have
been simply too short a period to detect any job experience
effects on self-esteem and locus of control. On the other
hand, this same year lag might have been too long a period
to uncover any selection effects. The question of an appro-
priate time Ihx; could be dealt with in future research with
the use of shorter lags and multiple measurement points.
Such a design would allow an investigation of possible
causal effects within different time lags and a more thorough,
systematic analysis of the work experience-personality rela-
tionship during a person's career.

Although the analyses presented in the study prohibit
the interpretation of either selection or job experience
effects, the author feels that investigations in this area
must be pursued to fully appreciate the developmental
processes involved in the careers of individuals. A more
thorough understanding of these causal processes within the
job experience and personality relationship necessitates a

move away from an intermediary causal design such as

58

cross-lagged analyses and a movement toward the formation of
a more holistic theoretical system which incorporates many
of the sources of spuriousness which were not addressed in
this study. These causal systems would best be explored
with use of structural equation and path analytic models
(of. Duncan, 1975; Heise, 1975).

The use of structural equation modeling would permit
the researcher to develop more fully the role of social-
ization and background factors in personality trait formation
and the role of job experience in the changes in these per-
sonality traits. Causal paths could be compared from these
two sources of influence to determine the most powerful
causal agent. More importantly, a nonrecursive causal model
could be developed in which possible reciprocal relation-
ships could be examined. If indeed, these causal relations
are found, the fit between an individual and his or her job,
often treated as a static one in personnel testing and job

design, must be reexamined.

REFERENCES

REFERENCE NOTE

lM. Fitzgerald, & N. Schmitt, Organizational size,
work group size, socioeconomic status and perceived job
characteristics, Paper presented at Midwestern Psychological
Association, Chicago, May, 1979.

 

 

 

59

REFERENCES

Andrisani,P. J., & Nestel, G. Internal-external control as
contributor to and outcome of work experience. Journal
of Applied PsycholOQY: 1976, 61, 156-165.

Argyris, C. Personality and organization. New York:
Harper and Row, 1957.

Argyris, C. Integrating the individual and the organization.
New York: Wiley, 1964.

Argyris, C. Personality and organization theory revisited.
Administrative Science Quarterly, 1973, 18, 141—167.

Bachman, J. G., & O'Malley, P. M. Self-esteem in young men:
A longitudinal analysis of the impact of educational and
occupational attainment. Journal of Personality and
Social Psychology, 1979, 35, 363-380.

 

Box, G. E. P. A general distribution theory for a class of
likelihood criteria. Biometrika, 1949, 36, 317-346.

 

Breer, P. E., & Locke, E. A. Task experience as a source of
attitudes. Homewood, IL: The Dorsey Press, 1965.

Brief, A. P., & Aldag, R. J. Employee reactions to job
characteristics: A constructive replication. Journal
of Applied Psychology, 1975, 69, 182-186.

Brief, A. P., Wallace, M. J., & Aldag, R. J. Linear vs.
nonlinear models of the formation of affective reactions:

The case of job enlargement. Decision Sciences, 1976,
7' 1-90

 

Brousseau, K. R. Personality and job experience. Organi-
zational Behavior and Human Performance, 1978, 22,
235—252.

Campbell, D. T. From description to experimentation: Inter-
trends as quasi-experiments in C. W. Harris (Ed.),
Problems in measuring change. Madison, WI: University
of Wisconsin Press, 1963.

60

61

Coopersmith, S. The antecedents of self-esteem. San
Francisco: Freemont,l967.

 

Cowen, E. L., Heilizer, F., Axelrod, H. S. & Alexander, S.
The correlates of manifest anxiety in perceptual
reactivity, rigidity, and self-concept. Journal of
Consulting PsycholOQY: 1957, 21, 405-411.

 

 

Duncan, 0. D. Introduction to structural equation models.
New York: Academic Press, 1975.

 

Durkheim, E. Division of labor in society. New York:
Macmillan, 1947.

 

Feldman, J. Considerations in the use of causal-
correlational techniques in applied psychology. Journal
of Applied Psychology, 1975, fig, 663-670.

 

Fromm, E. Escape from freedom. New York: Rinehart and
Company, 1941.

 

Gergen, K. J. The concept of self. New York: Holt,
Rinehart and Winston, 1971.

 

Greene, C. N. Questions of causation in the path-goal the-
ory of leadership. Academy of Management Journal,
1979, 22, 22-41.

 

Hackman, J. R., & Oldham, G. R. Development of the Job
Diagnostic Survey. Journal of Applied Psychology,

Hall, D. T., & Nougaim, K. Correlates of organizational
identification as a function of career pattern and
organizational type. Administrative Science Quarterly,
1972, 21, 340-350.

Hamner, W. C., & Organ, D. W. Organizational behavior: An
applied psychological approach. Dallas: Business
Publications, 1978.

 

 

Hanlon, T. E., Hofstaetter, P., & O'Connor, J. Congruence
of self and ideal self in relation to personality
adjustment. Journal of Consulting Psychology, 1958,
22, 215-218.

Heise, D. R. Causal analysis. New York: Wiley, 1975.

 

Holland, J. L. Making vocational choices: A theory of
careers. Englewood Cliffs, NJ: Prentice Hall, 1973.

 

62

Holland, J. L. Vocational preferences, in M. D. Dunette
(Ed.), Handbook of industrial organizational psychology.
Chicago: Rand McNally, 1976.

Ivancevich, J. M. The performance to satisfaction relation-
ship: A causal analysis of stimulating and nonstimu-
1ating jobs. Organizational Behavior and Human
Performance, 1978, 22, 350-365.

 

 

James, L. R., Hornick, C. W., & Demaree, R. G. A note on
the dynamic correlation coefficient. Journal of
Applied PSYChOlOQXI 1978, 22, 329-337.

 

Joe, V. C. Review of the internal-external control construct
as a personality variable. Psychological Report, 1971,
gg. 619-640.

 

Katz, R. The influence of job longevity on employee
reactions to task characteristics. Human Relations,
1978, 22, 703-725.

 

Kenny, D. A. Cross-lagged correlation: A test for spuri-
ousness. Psychological Bulletin, 1975, 22, 887-903.

 

Kohn, M. L. Class and conformity: A study in values.
Homewood, IL: Dorsey Press, 1969.

Kohn, M. L., & Schooler, C. Class, occupation, and orien-
tation. American Sociological Review, 1973, 22,
659-678.

Kohn, M. L., & Schooler, C. The reciprocal effects of the
substantive complexity of work and intellectual
flexibility: A longitudinal assessment. American
Sociological Journal, 1978, 22, 24-52.

 

ijorman, A. Self-esteem as a moderator of the relationship
between self-perceived abilities and vocational choice.
Journal of Applied Psychology, 1966, 22, 479-486.

Korman, A. K. Self-esteem as a moderator of the relationship
between self-perceived abilities and vocational choice.
Journal of Applied Psychology, 1967, 22, 65-67.

Korman, A. K. Self-esteem as a moderator in vocational
choice: Replications and extensions. Journal of
Applied Psychologyp 1969, 22, 188-192.

 

 

f Korman, A. K. Toward an hypothesis of work behavior.
Journal of Applied Psychology, 1970, 22, 31-41.

 

Kornhauser, A. Mental health of the industrial worker. New
York: Wiley, 1965.

 

63

Lawler, E. E. A correlational-causal analysis of the rela-
tionship between expectancy attitudes and job perfor-
mance. Journal of Applied Psychology, 1968, 22,
462-468.

 

Lefcourt, H. M. Internal versus external control of rein-
forcement: A review. Psychological Bulletin, 1966,
22, 206-220.

 

Lefcourt, H. M. Recent developments in the study of locus
of control. In B. Maher (Ed.), Progress in experi-
mental personality research (Vol. 6). New York:
Academic Press, 1972.

 

 

Mead, G. H. Mind, self and society. Chicago: University
of Chicago Press, 1944.

 

Morse, T., & Young, D. Personality development and task
choices: A systems view. Human Relations, 1973,
22, 307-324.

 

Nosow, S., & Form, W. H. Man, work and society. New York:
Basic Books, 1962.

 

Organ, D. W., & Greene, C. N. Role ambiguity, locus of
control and work satisfaction. Journal of Applied
ngchology, 1974, 22, 101-102.

 

 

Peters, C. C., & Van Voorhis, W. R. Statisticalpprocedures
and the mathematical bases. New York: McGraw-Hill,
1940.

 

 

Phares, E. J. Locus of control ingpersonality. Morristown,
NJ: General Learning Press, 1976.

 

Phares, E. J., Ritchie, D. E., & Davis, W. L. Internal-
external control and reaction to threat. Journal of
Personality and Social PsycholOgY: 1968, 22, 402-405.

 

 

Pines, H. A., & Julian, J. W. Effects of task and social
demands on locus of control differences in information
processing. Journal of Personality, 1972, 22, 407-416.

 

Reisman, D. The lonely crowd. New Haven: Yale University
Press, 1950.

 

Roby, D. Task design, work values and worker responses: An
experimental test. Organizational Behavior and Human
Performance, 1974, 223 264-273.

 

 

Rosenberg, M. Occupations and values. Glencoe, IL: The
Free Press, 1957.

 

64

Rosenberg, M. Sociepy and the adolescent self-imagg.
Princeton, NJ: Princeton University Press, 1965.

 

Rotter, J. B. Social learning and clinical psychology.
New York: Prentice-Hall, 1954.

 

Rotter, J. B. Generalized expectancies for internal versus
external control of reinforcement. Psychological
Monographs, 1966, 22,(1 whole no. 609).

 

 

Rotter, J. B. Some problems and misconceptions related to
the construct of internal versus external control of
reinforcement. Journal of Consulting and Clinical
Psychology, 1975, 22, 56-67.

 

 

Schroder, H. M., Driver, M. J., & Streufert, 8. Human
information processing. New York: Holt, Rinehart,
and Winston, 1967.

 

Seeman, M., & Evans, J. W. Alienation and learning in a
hospital setting. American Sociological Review, 1962,
21, 772-783.

 

Skinner, B. F. Science and human behavior. New York:
Macmillan, 1953.

 

Stone, E. F. The moderating effect of work-related values
in the job scope - job satisfaction relationship.
Organizational Behavior and Human Performance, 1976,
22, 147-167.

 

Strickland, B. R. The prediction of social action from a
dimension of internal-external control. Journal of
Social Psychology, 1965, 22, 353-358.

 

 

Taylor, F. W. The princ2ples of scientific management. New
York: Harper, 1911.

 

Valecha, G. K. Construct validation of internal-external
locus of control of reinforcement related to work-
related variables. Proceedings of the 80th Annual Con-
vention of the American Psychological Association,
1972, 2, 455-456.

 

 

Vroom, V. H. A comparison of static and dynamic correlation
methods in the study of organizations. Organizational
Behavior and Human Performance, 1966, 2, 55-70.

 

 

Walk, 5., & Ducette, J. Intentional performance and inci-
dental learning as a function of personality and task
dimensions. Journal of Personality and Social Psy-
chology, 1974, 22, 90-101.

 

65

Wanous, J. P. Individual differences and reactions to job
characteristics. Journal of Applied Psychology, 1974,
22, 616-622.

 

Weick, K. E. Enactment processes in organizations. In B.
Staw & G. R. Salancik (Eds.), New directions in orga-
nizational behavior. Chicago: St. Clair Press, 1977,
267-300.

 

 

White, J. K. Individual differences and the job quality-
worker response relationship: Review, integration and
comments. Academy of Management Review, 1978, 2,
267-280.

 

Wylie, R. C. The self-concept. (Rev. ed., Vol. 1) Lincoln:
University of Nebraska Press, 1974.

 

Yee, A. H. The source and direction of causal influence in
teacher-pupil relationships. Educational ngchology,
1968, 2, 275-282.

 

Yee, A. H., & Gage, N. L. Techniques for estimating the
source and direction of causal influence in panel data.
ngchological Bulletin, 1968, 2, 115-126.

 

APPENDICES

APPENDIX A

JOB DIAGNOSTIC SURVEY

APPENDIX A

JOB DIAGNOSTIC SURVEY

Respondents were asked to answer each of the following questions

concerning their job. Except where noted, the subjects responded to the

following scale:

How accurate is the statement in describing your job?

 

1. Very Inaccurate

2. Mostly Inaccurate
3. Slightly Innacurate
4. Uncertain

5. Slightly Accurate
6. Mostly Accurate

7. Very Accurate

Autonomy
1. How much autonomy is there in your job? That is, to what extent does

your job permit you to decide on your own how to go about doing your
work?

 

l --------- 2 --------- 3 --------- 4 --------- 5 --------- 6 --------- 7
Very little; the job Moderate autonomy; Very much; the job‘
gives me almost no many things are gives almost com-
personal "say" about standardized and not plete responsi-
how and when the under my control, but bility for deciding
work is done. I can make some deci- how and when the
sions about the work. work is done.

The job denies me any chance to use my personal initiative or judg-
ment in carrying out the work.

The job gives me considerable opportunity for independence and
freedom in how I do the work.

66

67

Task Identipy

 

 

 

4. To what extent does your job involve doing a "whole" and identifiable
piece of work? That is, is the job a complete piece of work that has
an obvious beginning and end? Or is it only a small part of the
piece of work, which is finished by other people or by automatic
machines.

1 --------- 2 --------- 3 --------- 4 --------- 5 --------- 6 --------- 7

My job is only a tiny My job is a moderate- My job involves

part of the overall sized "chunk" of the doing the whole

piece of work; the overall piece of work; piece of work, from
results of my activ- my own contribution start to finish;
ities cannot be seen can be seen in the the results of my
in the final product final outcome. activities are

or service. easily seen in the

final product or
service.

5. The job is arranged so that I do ppE_have the chance to do an entire
piece of work from beginning to end.

6. The job provides me the chance to completely finish the pieces of

work I begin.

Skill Variety

 

7. How much variety is there in your job? That is, to what extent does
the job require you to do many different things at work, using a
variety of your skills and talents?

l --------- 2 --------- 3 --------- 4 --------- 5 --------- 6 --------- 7

Very little; the job Moderate variety. Very much; the job

requires me to do requires me to do

the same routine many different

things over and over things, using a

again. number of dif-

ferent skills and
talents.

8. The job requires me to use a number of complex or high-level skills.

9. The job is quite simple and repetitive.

Feedback from the Job

 

10. To what extent does doing the job itself provide you with informa-

tion about your work performance? That is, does the actual work

68

itself provide clues about how well you are doing--aside from any
"feedback" coworkers or supervisors may provide?

1 --------- 2 --------- 3 --------- 4 --------- 5 --------- 6 --------- 7

Very little; the job Moderately; sometimes Very much; the job

itself is set up so I doing the job provides is set up so that

could work forever "feedback" to me; I get almost con-

without finding out sometimes it does not. stant "feedback"

how well I am doing. as I work about
how well I am
doing.

11. Just doing the work required by the job provides many chances for
me to figure out how well I am doing.

12. The job itself provides very few clues about whether or not I am
performing well.

Task Significance

 

13. In general, how significant or important is your job? That is, are
the results of your work likely to significantly affect the lives
or well-being of other people?

 

l --------- 2 --------- 3 --------- 4 --------- 5 --------- 6 --------- 7
Not very significant; Moderately significant. Highly significant;
the outcomes of my the outcomes of my
work are 292_1ikely work can affect
to have important other people in
effects on other very important
people. ways.

14. The job is one where a lot of other people can be affected by how
well the work gets done.

15. The job itself is not very significant or important in the broader
scheme of things.

 

Feedback from Others

 

16. The supervisors and coworkers on this job almost never give me any
feedback about how well I am doing this job.

17. Supervisors often let me know how well they think I am performing
the job.

69

18. To what extent do managers or coworkers let you know how well you
are doing on your job.

1 --------- 2 --------- 3 --------- 4 --------- 5 --------- 6 --------- 7

Very little; people
almostneverlet me
know how well I am
doing.

Moderately; sometimes
people may give me
"feedback", other
times they may not.

Very much; man-
agers or coworkers
provide me with
almost constant
"feedback" about
how well I am
doing.

APPENDIX B

LOCUS OF CONTROL

APPENDIX B

LOCUS OF CONTROL

The following items assess locus of control on a 4-point
scale of agreement.

1.

10.

11.

Many of the unhappy things in people's lives
are partly due to bad luck.

In the long run, people get the respect they
deserve in this world.

Without the right breaks, one cannot be a good
leader.

What happens to me is of my own doing.

Becoming a success is a matter of hard work;
luck has little or nothing to do with it.

When I make plans, I am almost certain that I
can make them work.

In my case, getting what I want has little or
nothing to do with luck.

Who gets to be boss often depends on who was
lucky enough to be in the right place first.

Most people don't realize the extent to which
their lives are controlled by accidental hap-
penings.

Many times I feel that I have little influence
over the things that happen to me.

In the long run, the bad things that happen
to us are balanced by the good ones.

70

APPENDIX C

SELF-ESTEEM

APPENDIX C

SELF-ESTEEM M‘, 1

The following items assess self—esteem on a 4-point scale
of agreement.

1.

I feel that I'm a person of work, at least on
an equal basis with others.

I feel that I have a number of good qualities.
All in all, I tend to feel that I am a failure.

I am able to do things as well as most other
people.

I feel I do not have much to be proud of.

I take a positive attitude toward myself.

On the whole I'm satisfied with myself.

I wish I could have more respect for myself.
I certainly feel useless at times.

At times I think I'm no good at all.

71

APPENDIX D

CAUSAL ANALYSIS BETWEEN LOCUS OF CONTROL

AND PERCEIVED JOB CHARACTERISTICS:

NEW HIRE SAMPLE

72

{Cu \ 1‘. I

.nmoc. : ¢J¥15 Am. “n+mccv a rifle: Am. “AAImcn. + A+mcovv \ *A1x2_. + A+ucgvv A~vn
.mucoqo_uwcoo gamummOLU ccuoctuoo accrues .m. can mcowucfiouuoo maocoL£oc>m couoouuoo coozuwn AH. cocoo_uwcc.m Co amok:

 

 

 

 

 

 

 

.om~.. mmm. .mo~.. mam. Ame.

eve. evv. moo. an mm mm an cma. .oon.. ova. .ov~.v mom. Ade.
muozuo EoLu xvmnpmom

.~o~.. can. Ace—.. «fid. .Nb.

moo.~ woo. mmm. em mm m~ mm are. .oc~...~vn. Ao-.. -~. Arr.
Aodmom seesaw. moo

.om~.. amu. .~0~.. mma. .mb.

mmo. moo. mmo. ma mm ma om «emu. .«o~.. NAN. A~o~.. so". Ruse
ouccoaudcvdm xmmk

ano~.. cam. Anno.. ego. “we.

Ono.m cam. oov. mm 0“ LA mm «mou. .moa.vca-. Awho.. 500. Age.
>oaotooc ence

.mm~.. noa. Ammm.. Nag. “we.

"no. Avm.n Hom.~ ma ma ha on coo. .o-.. ova. Amon.. cod. .uh.
non Eben xomncomm

Amo~.. tau. Roma.» 050. .Nb.

one. ~90.~ mmw. «a mu NH cm ago.u Anon..cmhm. Avmo.. ONH. .ah.
>uoaao> -axm

Amo~.. «mm. An~A.. moo. .Nb.

ammo.v mmo. ooo.~ um mm m cw ov~. Abba..cmbn. .m0n.~ oo~. Adv.
>50c0usd

Am. Amy Am. .:.moo .+.moo .svooa .+.ooq Ana. Ada.
nmumwh ouosgwuuzu nuanceavoum fixx_v mcowuaaouuou mauucou no have; macaw moo
Dusnc>o tenuouuou
.quucou uo waved. .aanezucvuun :« vouuouuooca.
humane: nnOIL acomuonouuoo u:0:0uzoc>m vco

GCUVOIAInaOMU vouoouuou

kh.LEMFLL:L.FI.:flUt1hulDIUnLllan."AF|..-..I..tn Lut.“lhhhvd"l..nl.l .nhkh.hfl.urrflhba.u'ri.hu ’C£3“5tt

 

“III! ’1' L. .

 

DNA I 2
NJLZ<m mm“: ID: “wUHPmnszU<¢<:U non Dm>~m9¢mn G21 AOKFZOU LO mDUOQ 2Mfl3kmm m~mr4<2¢ AdMD‘U

D Xuazmmn<

.‘dhuthLLV4h "r vulvfiuwll

APPENDIX E

CAUSAL ANALYSIS BETWEEN LOCUS OF CONTROL AND

PERCEIVED JOB CHARACTERISTICS: TOTAL LCC

VOCATIONAL EDUCATION SAMPLE

mo. v m.

..umoo. A .-ooav .m. a.+moo. A .+ooa. Am. n~.1mow. + A+woovw A AAuooq. + .+ooqov .d.
.mucowofluueou ocnnmmOLU wouuwuuoo cuozuwn AN. use macauoacuuoo maoconsoc>w cvuuoquu :003uon AH. oucmufiuHCOAm uo umwhm
. . o

 

73

 

 

 

 

 

 

 

“hv~.. mmm. .moH.. Hmn. Ame.

0mm. coo. mmn. on $5 mo mp coo. .m-.. «mm. Ammu.. Om“. Ade.
muozuo Sena xucncoom

Ammm.. OVN. Avmm..¢vm~. Ame.

m~¢.~ omm. mma. mm mm me we who. Ao-._ gum. .mm~.. Hh~. Age.
.335 pee-asw. moo

.vv~.. on“. .mmn.vcnm~. Ana.

ooo. 0mg. who. hm no on we owe. Aho~.. goo. .msa.. omH. Ade.
occupauucumm xmme

Avou.. hon. Amha.v 00H. Away

vmv. Nov. ooo. 0v pm on we avo. “00“.. -~. .bhH.. wow. .Hh.
>uwucopn xmub

AmH~.. no”. Aoo~.. mow. Auk.

ooo. nmo. moo. mm oo on on .mHH. .v-.. ~om. 1-~.. nn~. .He.
now ibuu xocncoom

.mau.~ «om. .Nvu.. mv~. .~b.

mam.~ omo. ovo. mm mo m9 hm who. .mnu.. on“. Amen.» cow. Ade.
>uowua> Huaxm

.mhu.. 0mg. .~N~.. naw. Away

avo. mom. can. om om wv no Amo. .@HH.. Own. Amh~.. mad. Auev
>§Oc0ua<

In. .~. A“. .-omoo 1+.woo .-eoos .+.ooa .~e. .He.
nmumeh ohaskuanu moaocosvoum AOOq. mcoauadouuou «ouucou no wooed oucuw moo
ofifiuc>o vauoouuoo
“~0uucco we moved. .mdwozucuuua nu vouuouuoucov
mu—amom Auuum mcowum«ouuou unoccusuc>m vac
avooouqnnnouu vouuouuoo
anm I z

mdaxdm ZO~P<02DM A420~F¢UO> UUJ 4(909
"woubmnmmfib<m<:o mow Dw>~m0¢ma D2¢ AOGPZOU ho MDUQA zmmzhmm m~m>4<z¢ A<WD<U

m XHDZMQA<

APPENDIX F

CAUSAL ANALYSIS BETWEEN LOCUS OF CONTROL

AND PERCEIVED JOB CHARACTERISTICS:

LOW TENURE LCC SAMPLE

74

no.

\u.

..-moo. A TARx: Am. “1+mooo A A+ooq. .m. "A.-moo. + ..maovv A ..-ooJo + c:Xx:1 .~.n
.mucofiUCchU Godlmmouu wouUOLLOU coo3uon .3 can mcoquoawuucU mzoc0u£0c>m vuuowuuou cowluon A: oucoofiuficvgm CO amok...

 

 

 

.mwm.v mom. “ovm.. mv~. Awe.

vac. ooo. mOM. mm mm hm mm moo. Amm~.. haw. A~m~.. omm. Aaev
muonuo Soho xumnooom

.mm~.. mum. .vhm.. cow. .mk.

ovv. 000. we". mw om Hm mw mmo. Amo~.. own. .mm~.. #hm. Auk.
Awaoum poEEsmv moo

Ah-.. Nna. Amm~.. mmn. ANS“

mm“. poo. mma. vH AM ha MN mmo.| Avoo.. woo. .Nv~.. mmu. Ade.
ouccowuwcmam xmcfi

.mmd.. oma. .m0m.. mma. ANP.

ooo. mac. heo. ed mN hm hm omo. Amo~.. poo. A~a~.v oma. AME.
>u1ucoo~ xnae

Anh~.. Hmm. Aomm.. nmm. Ame.

Hmo. one. mac. ma hm hm mm moo. Rama.. and. Ammw.. vv~. Aaav
now scum xocncoom

A~m~.. mum. thw...h~n. Amh.

~o~.~ cow.“ hvo. NH mm mu mm m~o.; .hc~.~ non. .o~n.. cem. Ads.
>uoaao> «Haxm

.~«~.. mod. A~n~.. nma. “we.

mmv. va. 000. ed on ma mm nco. .umo.. Nmo. Aooa.. cow. Ade.
>50c0u3<

Am. AN. .A. .u_wow .+.moo Au.ooq .+?HH_ Ame. A~FV
mummk ouozvmnwzu mowocoavoua “Rx—v mcoqucuouuou deuucoo we have; oqaum mac

n

L511L.r’,§1r£rlrx LET‘f‘E n. . n .

 

 

.~0uucou no $2004.
madamom AIUnL

vxu die... n.

omeoc>o ovuumuuou

I .hfl H .l u ~Il r k. V“.N.II.NII.I ~uu‘),’"l 4...‘

and n z
mqafim .UUJv "5:..sz I:

H.

 

 

.mfimozucouea :w pouuouuoocov
mcoguagouuou moo:0uzuc>m 6:6
upooaaqummOLU vauuouuoo

win. a tin. u

”WU—Pm~xmfib<x<:0 ECO ou>~mUmxz 02¢ AOZFZOU LO mDUQJ ZMMZme m~w>4<z< J<mD<U

L x~Dmem<

a.dfl1fl¢~.lnfl:ul ‘4 «HI Sq: flihxt wflflnwli Ink". .

APPENDIX G

CAUSAL ANALYSIS BETWEEN LOCUS OF CONTROL

AND PERCEIVED JOB CHARACTERISTICS:

HIGH TENURE LCC SAMPLE

75

mo. v m.
..umoo. A Auooa. .m. a1+moo. A .+ooao .m. “Alumna. + ..moeev A Aluooqo + ..mxae. «Hen
.mucodomuucou ocqnmmouu venoouuou cooiuon .N. can mCOADoaouuou mzocounoc>m Oouowuuou coozuon Au. ou:mu«udccuw Co umoPm

 

 

 

 

 

 

 

.m-.v mam. Awmn.u ova. Ame.

ohm. ooo. man. he he on mo moo. .vm~.. «on. .moo.v wmo. , .Hb.
muozuo Seam xuwncowm

amo~.. ovu. Aom~.. how. fink.

on. mod. Nmo. mm Om we we owe. Ano~.. mmm. Amn~.v mom. Ans.
Aoaoum possum. moo

A~o~.. mum. A~o~.v ovm. amp.

own. Hoo.n vmo.~ ad on mm mv ooo. Ao-.~ mea. Ano~.. oma. Ada.
wUCMUAuAcOAm xmce

Acm~.w hma. Amh~.v and. Ame.

cho~.o ooo. c00m.m ov mm om mm “00. Amvm.vsmoN. A~m~.v mmm. Ade.
>uausopa xmme

Adh~.. men. ona.. mod. Amp.

ooh. mac. and. hm pm on on mod. .mv~.v QAN. .mam.~ vHN. Ade.
now EOuu xumncoom

Ahm~.. Had. .A-.. wow. Ame.

00m.~ nmo. ohm. ON on am on cowu. .unn.. av“. .~m~.. noa. Ade.
>uofiua> Haaxm

.~m~.. nod. Ahm~.. oh~. Ame.

om“. ohm.~ mom. mm am pm we mmo. .mmn.. “ma. Amo~.. «om. .ae.
>Eoc0u=<

Am. AN. An. ”nomad .ovwoo .nvuo: .+.ux%_ ”Nb. . AAE.
mamas oposkumco mowocmsveum ALKJ. mcouumaouuou dauucou uo msuoq «doom wow
2 ugeoc>o vouuwuuoo
Auouucou mo msuo;. Anamozucouma cw vouuouuoocav
muaamom mnOnm mcouuoaouuou waoc0u£oc>m can
cpovomqnmmOLU pouUOLLOO
can u z

mJL£<m AUQAV fixazmfi Iva: "MUHSmnmeU<x<:U mow Qm>uuummm Dz< AOmPZOU LO mDUOJ 2mm3fimm w~m>4<2< J<WD<U

O x~02mma<

APPENDIX H

CAUSAL ANALYSIS BETWEEN LOCUS OF CONTROL AND
PERCEIVED JOB CHARACTERISTICS: NEW HIRE-LOW

TENURE (LCC) COMBINED SAMPLE

76

..uwopo A .quS. .n. ax+mono A .+qu. .w. “I.-mno. + A+monoo A 1.1oQa. + .+uoo.o .Ho
.mucuwowuuooo omanmmOuo vouuonuou cootuen Aw. vac acouuuaouuou moccoucoc>n vauooquU comluon Adv oocuoauwcowm mo an»?

mo. v m.
n

 

 

 

 

 

a

Amm~.. vmm. .m-.. onm. Auk.

mom. can. one. no mm am «0 oeo. .hw~.. oaw. AAQN.. ANN. Rafi.
uuozuo scum xoanvoom

Anv~.. nmw. noo~.. boa. Ana.

boo. who. nmm. vm oo om on «boa. .-~.. ova. .~v~.. can. Ade.
Aoacom coalamo moo

.om~.. ova. .o-.. omm. flue.

who. Nov. o~o. om hm mm ov o~o.: Avma.. and. Avo~.. ~ma. .Ha.
oocuouuwcuwm xuma

Aaoa.v Mme. Anoo.. moo. Ame.

mmo. who.” 000. on do on um nod. Aom~.. cod. ANQH.. ova. Ana.
auuucovu xmoa

.oo~.. cmw. Anon.. mpg. .Nao

oma. 0mm. awe. on be «m vm «mmu. .wv~.. mod. Abou.. v-. Ada.
non souu sunnvoom

.o-.. ooa. .~o~.. mpg. awe.

hao. oao. moo. mm nm om om moo. Ammu.. ecu. «om. . cam. Aabv
>uowuu> Hauxm

A~o~.. «ha. ~o-.. non. «may

nnm. pom.“ m~o.~ hm mm on «v snag. Amo~.. mod. .Hoa.. NAN. “day
>30:Ou=<

.m. .m. 1H. .-omoe 1+.moo .-eooo .+.oos Ame. Ads.
names opmsgmuwzu woqocmscoum fiXK: mc0auc~ouuou acuucoo uo mono; oamom won

9

CFQvWFAIFL'nlnu lfldnuuhuidffl.

mJaxtw szHmZOU AUUJV NdDzzP IOJImmH: 3mz

 

.aOuucou mo macaq.
mufizmom QIUIL

1 ud «ill. Hufiflv'uIL .Ilih. Irvhl. ..

oaadc>o couoouuou

 

Hmm n z

z xH02mmm<

 

Anamozucouco cw couoouuoocoo
mcoflumHouuou moccOunoc>m can

upooooqommOuo vouoouuou

.lin! HANNWIW ~3F!‘.Ugfil

Ifl“.g1ln "‘4“ fl"

umUHFmHmm60<m<=U n05 Qm>Hmuxmm 02¢ AOKPZOU m0 WOOD; zmwlbwm m~w>d¢2¢ atmatu

APPENDIX I

CAUSAL ANALYSIS BETWEEN SELF-ESTEEM AND
PERCEIVED JOB CHARACTERISTICS:

NEW HIRE SAMPLE

77

me. x d.
.7221.» 1-94: .moux.§:o A .rx2. .m.a_.-§:; . 1:57.».._.-%s: . lixzoo 2.;
.mucogogwuoou vogtmmoLo cauooupou coczuon AN. pct mccduogouhoo unoccunuc>m wagoouuoo CQCBucn Ag. cccmugogcrom Co uwzkc

 

 

 

I I .‘ ‘-'alll"--l 1|.O I

 

 

 

 

 

 

 

.~g~.. ogN. Aco~.. new. .NP.

omo.~ mpg. mvv. mm em mm on smog. Achg.. mmg. Aog~.. ggm. .geo
95.3: So.— C xocbrcoi

.omn.. New. A~m~.. ngm. Ame.

mom. cm~.m woo.m vm mg on mm moo. .ov~.. ohm. Aomg.. mgw. Age.
gogmow ocEEsm. moo

.mOg.. moo. ~ggg.. ogg. ANS.

goo.g bom.g coo. gm cw om vm chem. .ogg.o 50g. novo.. one. Age.
ouccogudcogm xmah

Amm~.. ohg. Aooo.o moo. ANS.

oom.0g gog.~ gmm. on Nm Vg en chew. .o-.. vmw. “nog.. ~m~. .ghv
soaocoog shoe

Aegn.. «mm. A0n~.. ogm. Amp.

moo.g Nmo.~g ~h0.~ ow og mg me omg. Aoo~.. ggN. .hhg.. onw. .gk.
non scuu xocnpowm

Abo~.. gmg. Aoog.. meg. Auk.

ggm.o «mo.g pom. cm mN mg on -o.u .mmg.. ogm. govo.o boo. .gh.
>uoguu> gggxm

.mhn.. nmw. .~o~.. NmN. .Nk.

gmm.n vmo.h non. gm 0g 5g nv cogm. .omw.. new. .ovg.. ogm. Agko
>50c0u=<

“no .m. .g. A-.moo A+Vmoo Aaooca ..ouog Ame. .gk.
nmumOP euasvmugzu mogocosvoum fx4_o mcoguogouuou gOuucou uo waved wgoum moo
ugsoc>o vouuouuou
”genucoo mo msoog. Amgmwnucouon :g vououuuoocoo
magomom anonm mcogucgouuou naoc0uzoc>m vow
accooagunmouu mouuouuoo
o~g u z

MALI/xv. “5:. 3x2 “MU—ngmkroixzsz 13.. on._>gx.omx._ CZ< Im...—.—.mm:...._..._m ZEHIEQ m~m>4¢z< J<m2<U

g xgczmaa<

APPENDIX J

CAUSAL ANALYSIS BETWEEN SELF-ESTEEM AND

PERCEIVED JOB CHARACTERISTICS: TOTAL

LCC VOCATIONAL EDUCATION SAMPLE

..ImOfiv A A1004.

78

u.+moo. A .+uoa. AN.

“*Anmoo. + A+m0522 A *AIUOJ. + A+ano~ Ago
.mucofioguuooo oogumm0uo vouoouuou cooiuon .N. can mcoguogouuou mac:0u£uc>u vouoouuoo coo3uon gg. uochgugcogm no amok

mo. V ..C
D
0
.Nb.

Age.
nnosoo acne xuooooon

.~e.

Age.
Avgoum 00.5.53 wan.

.Nbv

Age.
cocoUgugCOgm xmub

.Ne.

Age.
augusoog anus

Awe.

.ga.
non Iona xuogvoom

Ana.

ggko
huoguu> gggxm

.NP.

Age.
>IO:OD=<

 

gu.mon A+vmofi AuvUCJ g+.UCJ

 

ghhg.. Chg.
Avog.. ooN.
Amgn.v ogn.
gonn.. mvn.
Amo~.. «nu.
.NOg.. ng.
Aoom.v omg.
AFVg.v hog.
gnoN.. vow.
avmg.. ong.
.oo~.. cow.
.ngN.. ogu.
Ago~.. va.
thg.. hog.
awhv

 

 

munch oucsrmugcu goon. mcogucgounou

 

gouucou no «:00;

 

ugfidc>o couoouuoo
AgOLucoU mo maooav
mugsmvx enunm

(FLE‘nUFIt) AL- V

ZC~F<U3CN A<ZCHF<UC> US Apr—Ch.

Angmozucouua cg puuuouuooco.
acoguMgouuou 6:0:0unocxm can
ovooougnnaOMU vauuouhou

FIJI. I'.-

ungFmgszU<¢<ZU mow Qw>gmbxma 02¢ meFWMImAmm 2mm3fimm mgm>4<z< Atmadu

oduuw mad

I- iii ”1'! IE iii-I.

APPENDIX K

CAUSAL ANALYSIS BETWEEN SELF-ESTEEM AND
PERCEIVED JOB CHARACTERISTICS:

LOW TENURE (LCC) SAMPLE

79

.783 A Tooi .8 2+8... A $on .3 £783 + Tmoo: A 2-83 + .53: I:
.mucogoguoooo ougammOHU coauouuoo cowzuen AN. can mcoguogwuuoo m30:0unoc>w puuuouuoo comzuvn Ag. cosngwgcogm no umoeo

mo. v N.
n

 

 

AnoN.. NoN. ggmN.. ooN. 2N9.

hgv. ooo. men. an on 5N oN ooo. Aoog.. mog. AomN.. moN. .gew
muozuo scum xoonvoom

.N0m.. ogm. .00m.ooNov. .Nao

goo.g ooo. ooo. oN mm gm Nn gvg. ANgN.. hog. .NoN.o NhN. .gb.
.ogmom poi-5m. won.

.omg.. ong. .ovn.oaooN. 2N9.

m0m.g ogo.g ooo. on 5N oN on wno.n .vgg.. ooo. AoNN.. gNN. age.
cosmOgugcmgm xuob

.Nog.. NoN. Anng.. ng. .Nb.

mvo.gg ghN. Nho.n hv 5N og Nm meg. Aomg.. cgg. Agog.. meg. Age.
>ugpcoog xmmb

gonn.. hoN. .Ngn.. goN. ANb.

MNo.m oNh.N NoN. gv VN gN on cog. mog.. NON. .mNN.. th. age.
non Baku xomnvwom

.mmN.. hoN. .oon.vcvnv. 3N9.

hgo.o ooo. Nom.N o9 Nn oN mm ono. . .ogN.v mug. .vgn.. ohN. Age.
sooaua> gggxm

.mmN.. onN. .NoN.. ahN. ANa.

Nov.m NNN.g Ngm. on oN ON on covg. «mog.o vog. Novg.. hmg. .gev
>Boc0u3¢

.3 .3 2. 7.me Aimee TUE :53 .95 2.5

 

 

mama? Ouoarmsgco

n mwgocoavauh

fixuuo mcogomgouuoo
ogEmc>o couoouuou

 

AgOLucou mo msooav
mugawom &IUIL

..WH ~ In ..VLM ..Iu. “4".u1‘:;.w ml.l1 nr "J.,“.«nflflflual!”I...‘hu.fl.Ildid‘.dNI...IN..“(.u|I.Ih_-

gmg n z

 

gouucou no usuog ogmom moo

 

.mgmozucouca cg couoouuoocoo
m:0guogouuoo m20:0u:o:>m can
ocmoomgummOMU vouoouuou

an... "a.“ ”.1-..“ J“ H‘EJ...“ ”rig...“ "tuilfldlnn‘ d. N flit-0.".ln d. u. N I. .L'. I‘Iflfllnu.‘ “ISA“ “1L .

udat<m AQUA. mmasz 30a ungngmmFU<m<zu mow Qm>HmUzmm 024 tfimfimMIhdmm ZNmBPQM mgm>d<2< J<mD<U

2 Xuozmmm<

APPENDIX L

CAUSAL ANALYSIS BETWEEN SELF-ESTEEM AND
PERCEIVED JOB CHARACTERISTICS:

HIGH-TENURE (LCC) SAMPLE

80

mo. \ ma
..-moo. A Auooo. .n. "1+moo. A ..ooo. .N. 11.nmoo. + 1+moeoo A A.-ooq. + ..ooo.. .goo
.mucogoguuoOo ougnmmcLo couoouuou cowauon AN. oco mcoguchCLoo msoc0u£uc>n vouoouuoo cooluon Ag. QUCMUgugcogm mo awoke

 

 

 

 

 

 

 

 

 

Avvg.. mmg. .mmg.. ooo. ANS.
Neg. Ngo. NNo. Nm on he gv ooo. Avog.. th. Ammg.. nog. .gfi.
muozuo Baum xoonooom
ANNn.. oNn. gnoN.. onN. ANS.
50g. mnn.g moN. ov om we on QNg. .omN.o moN. .NQN.. ovN. Age.
Aegoow cesium. moo
AmnN.. ooN. Awhg.. oog. .NB.
ng.g ohm. gum.g om gm on we ong. AegN.. nog. .moN.. th. Age.
oucoogugcogm shoe
ANNN.. ghg. gang.. ONg. 2N9.
mmN.g ggo. ooo. mm me me nv ogg. Acmg.. ogN. .vmo.o oho. .gb.
>ugucoog xmoe
.moN.. ONN. Agog.. meg. .Nh.
mgn.g ooo. hob. Nm ov ov mv song. Ang.. gNg. AVON.. oog. Aghv
non Echo xuunveom
.voN.. omN. AnoN.. beg. “New
Ngo. ooo. ooo. nv om nv av cmNg. .hNN.. th. .ohg.. cog. Age.
>umguo> gggxm
.NhN.. omN. AvoN.. nvN. “Nb.
ggo. Nhg. NNo. 0' vv 1v av gho. .va.v noN. .ong.. hog. .gh.
>EOCOu3<
In. .N. .g. .-omon 1+.mon .n.ooa .+.ooo .Ne. .ge.
munch oucavmngzu mogUCUDUOHL fiXx: mcogungopuou gOuucou uo maqu ogcum moo
n owaoc>o couoouuou
.gouucoo go msoog. .mgmozucouao cg vouoouuoocoo
mugsmom anUum mcoguogeuuou n:0:0uzuc>m pea
cvooocqummOuU vouoouuou
ooN u z

mqmt<m .094. mmszmfi 23g: “mugfimgmmfio<x<:o mOo Dm>gmU£mm 024 IMMPmmlmdmm 2mmxbmm mgw>q<2< J<mD<U

J xgozmoa<

APPENDIX M

CAUSAL ANALYSIS BETWEEN SELF-ESTEEM AND PERCEIVED

JOB CHARACTERISTICS: NEW HIRE-LOW

TENURE COMBINED SAMPLE

81

..quH. A .uooa. An. A.+mono A .+oQo. AN. Ag.nmnn. + .+moo.2 A ..nooa. + .+ooq.o .g.
.nucogoguuoou onguamouu vouoouuoo cooluon AN. can anaguoguuuou naocouzuc>u wouuouuou cooauon Ag. OUCngugcmgm uo uapk

mo. v N-
n

 

 

 

 

 

a

.ogN.. NgN. .ogN.. va. .Na.

mnn.g moo. mgg.g we on av hm goo. .oog.. Nog. .gnN.. onN. Age.
mumnuo noun xocavoum

.oNn.. ohN. .mNN.. opN. .Nb.

hmo.g ovo. ooo.N vv vm om mo smog. .0m~.. mnN. .ooN.. ovN. .gh.
goguum in. man.

.m~g.. mNg. nogN.. ng. .Na.

va.m noo.v Ngo. 05 av vv gs mgo.u Aegg.. ooo. Amng.. veg. Age.
oocnogHNGOgm some

AoNN.. mmg. .oOg.. beg. aNa.

mov.mg noo.N Nev.N Nm 5v on we .nog. .omg.. hog. .mmg.o Nog. .ga.
>ugucovg game

.vnn.. voN. .mvN.. mNN. .Na.

hen.h mom.n ooo. Np vv Nv ms «mgN. .mmg.. NgN. .vmg.. mvN. Age.
non Iona xoanuowm

.moN.. th. .nmN.. omN. .Na.

omo.gg Nam.N mnN.g ms mm on Np cge.: Anog.. cog. .mog.. goN. Age.
suoguo> gggxm

.mNn.. ovN. .ooN.. nmN. .Na.

nmm.m moo.n NNN. vb om ov on .gmg. .whg.. hog. .Nvg.. Nog. Age.
>EOCODS<

An. .N. .g. .-omoo 1+.moo Auoooq .+.ooo .Ne. Age.
mumps aposgm-g:o nogucoagmum AOQSC mcoaomgouuoo gouucoo no cacao ogmom mod

n

 

.gOhucou no m:oOJ.
mogsnom ouout

. «Win Arlyn!) on... a} 1nd,. . 1. ., II‘I...I.... “ .Iu ._,u.n . «lull u a I...

QJANEW omz g EZCU

1”]. a

di Hi

.«zHHII

22:22? 3C4 1mm H : 3m2

ogsdc>o wouuouuou

u..u ..It'lhlh I....Ilt.h.v.l. a. u «11‘ “J.“. P.”.‘“¢fla.iafl H. 4‘...“

gmN n z

 

Amgmenucouam cw concouuoucoo
macauogouuou msocounoc>m can

opooomglmmouu uouoouuoo

"mvgfimgmmfinvésa mOg. Dm>HmUmmm 02¢ zmnmkmmlhqgmw

I xgazmoea

On... it'll

2223920 m H m>ag<2< 449.20

.yyu’l‘.: iii! A“. E ”11'1" “1”!an