THE EFFECTS OF ORGANIZATIONAL COMMITMENT, PROFESSIONAL COMMITMENT, LIFE-SPAN CAREER DEVELOPMENT, AND SELF-MONITORING ON JOB SATISFACTION AND JOB PERFORMANCE AMONG STAFF ACCOUNTANTS by PETER JOHN POZNANSKI, B.S., M.S. A DISSERTATION IN BUSINESS ADMINISTRATION Submitted to the Graduate Faculty of Texas Tech University in Partial Fulfillment of the Requirements for the Degree of DOCTOR OF PHILOSOPHY Approved May, 1991
187
Embed
THE EFFECTS OF ORGANIZATIONAL COMMITMENT ... - TDL
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
THE EFFECTS OF ORGANIZATIONAL COMMITMENT, PROFESSIONAL
COMMITMENT, LIFE-SPAN CAREER DEVELOPMENT, AND
SELF-MONITORING ON JOB SATISFACTION AND JOB
PERFORMANCE AMONG STAFF ACCOUNTANTS
by
PETER JOHN POZNANSKI, B.S., M.S.
A DISSERTATION
IN
BUSINESS ADMINISTRATION
Submitted to the Graduate Faculty of Texas Tech University in
Partial Fulfillment of the Requirements for
the Degree of
DOCTOR OF PHILOSOPHY
Approved
May, 1991
So/ rz
e<
Copyright, 1991, Peter John Poznanski
ACKNOWLEDGMENTS
The participation of the professional accountants at
the office of the public accounting firm which
participated in this study is appreciated. Comments on
early drafts of this study from the accounting doctoral
students at Texas Tech University, notably Michael Shaub,
Randy Stitts, and Gary Caplan, are also appreciated.
Dr. Donald Clancy and Dr. James Wilcox were of great
assistance in the initial development of this study and
the data analysis, even though they were not committee
members.
A study such as this could not be successfully
completed without the assistance and encouragement of an
advisory committee, and the leadership of a committee
chairman. Dr. Don Finn's contributions as chairman have
made the completion of this study possible.
Finally, the continual support from family and
friends throughout the process is greatly appreciated.
Thank you all very much.
11
CONTENTS
ACKNOWLEDGMENTS ii
ABSTRACT v
LIST OF TABLES vii
LIST OF FIGURES ix
I. INTRODUCTION AND BACKGROUND 1 Need for the Study 3 Purpose of the Study 4 Organization of the Study 5
II. LITERATURE REVIEW 7 Job Satisfaction and Job Performance 8
Job Satisfaction in the Accounting Profession 9
Job Performance in the Accounting Profession 14
Organizational and Professional Commitment, Life-Span Career Development, and Self-Monitoring 21 Organizational and Professional Commitment in the Accounting Profession 21
Life-span Career Development 26 Self-Monitoring 32
Chapter Summary 42
III. THE THEORETICAL MODEL, RESEARCH MODEL, AND FORMULATION OF HYPOTHESES 45 The Theoretical Model 45 The Research Model 46 Formulation of the Hypotheses 48 Chapter Summary 54
IV. RESEARCH METHODOLOGY 56 Research Design 56
The Population for the Study 57 Development of the Research Instrument 58 Job Satisfaction Measurement 59 Job Performance Measurement 60 Organizational and Professional Commitment Measurement 62
111
Life-Span Career Development Measurement 63 Self-Monitoring Measurement 64 Research Instrument Distribution and Collection Technique 65
Pretesting the Research Instrument 66 Statistical Technique 71
LISREL Model Specification 72 Chapter Summary 79
V. EMPIRICAL ANALYSIS OF THE DATA 81 Response Information 81
Response Rates 81 Test for Nonresponse Bias 82
Statistical Analysis 84 Testing for Identification 84 Testing the Measurement Model 86 Hypothesis Testing 102 The Structural Model 103 Sequential Chi-square Difference Tests 106 Evaluating the Theoretical Model 114
Summary of the Data Analysis 123
VI. SUMMARY AND CONCLUSIONS 127 Summary of the Findings and Implications 127
Research Objective One 128 Research Objective Two 131 Research Objective Three 132
Contributions 134 Limitations 135 Future Research 136
REFERENCES 138
APPENDICES
A. RESEARCH INSTRUMENT 150
B. RESEARCH INSTRUMENT APPROVAL , 162
C. DEMOGRAPHIC DATA AND DESCRIPTIVE STATISTICS 164
D. LISREL RESULTS FOR INDIVIDUAL HYPOTHESES AND PARAMETERS 167
IV
ABSTRACT
Increasing the level of employee job performance and
job satisfaction is a goal of most organizations. No one
theory or variable has been found to explain job
performance and job satisfaction; however, research
evidence has indicated some correlation between these two
constructs (for example, see Vroom, 1964; Lawler and
Porter, 1967; Ross and Bomeli, 1971; Benke and Rhode,
1980, and Bullen and Flamholtz, 1985).
The importance of maximizing job satisfaction and job
performance among employees within organizations is
continually sought by managers. Maximizing job
satisfaction and job performance may lead to higher
productivity, lower overall costs for the firm, and
reduced turnover of employees.
This study investigated constructs that can affect
job satisfaction and job performance, as well as the
relationship between these two constructs. The research
objectives of this study were to: (1) investigate the
effects of organizational commitment, professional
commitment, life-span career development, and self-
monitoring on job satisfaction and job performance for
staff accountants in public accounting firms; (2)
investigate the relationship between job satisfaction and
job performance of staff accountants in public accounting
firms; and, (3) investigate the relationship among
organizational commitment, professional
commitment, and life-span career development for staff
accountants in public accounting firms.
VI
LIST OF TABLES
2.1 LIFE-SPAN CAREER DEVELOPMENT STAGES AND AGES 28
4.1 SUMMARY OF QUESTIONNAIRE RESPONSE RATES FROM THE PRETEST FOR ALL SUBJECTS 68
5.1 SUMMARY OF QUESTIONNAIRE RESPONSES AND ELIMINATIONS 83
5 . 2 NONRESPONSE BIAS TEST RESULTS 85
5.3 MEASUREMENT STATISTICS FOR THE TWENTY-ONE COMPOSITES 90
5.4 MEASUREMENT STATISTICS FOR THE SIXTEEN COMPOSITES 97
5 . 5 PEARSON CORRELATION COEFFICIENTS 99
5 . 6 POLYSERIAL CORRELATION COEFFICIENTS 100
5.7 PSEUDO TEST STATISTIC FOR THE NULL AND SATURATED MODELS 105
5.8 COMPARISON OF THE NULL -AND SATURATED MODELS 107
5.9 COMPARISON OF THE THEORETICAL AND SATURATED MODELS 110
5.10 COMPARISON OF THE THEORETICAL AND CONSTRAINED MODELS 113
5.11 COMPARISON OF THE THEORETICAL AND UNCONSTRAINED MODELS 115
5.12 GOODNESS-OF-FIT INDICES AND ROOT MEAN SQUARE RESIDUiALS 117
5.1 Respecified Measurement Model For Y Indicants 92
5.2 Respecified Measurement Model For X Indicants 93
5.3 Respecified Measurement Model For Composite Y Indicants 95
5.4 Respecified Measurement Model For Composite X Indicants 96
5 .5 Decision-Tree Framework For SCDTs 108
5.6 Research Model—Results Of Hypothesis Testing.... 126
IX
CHAPTER I
INTRODUCTION AND BACKGROUND
Increasing the level of employee job performance is a
goal of most organizations. No one theory or variable has
been found to explain job performance; however, research
evidence has indicated some correlation between job
performance and job satisfaction (for example, see Vroom,
1964; Lawler and Porter, 1967; Ross and Bomeli, 1971;
Benke and Rhode, 1980, and Bullen and Flamholtz, 1985).
Job satisfaction is a concept that differs across
individuals. Research evidence on job satisfaction can be
divided into categories such as leadership, psychological
needs, effort-reward, management ideology and values, and
job design factors and the content of work. Vroom [1964]
describes job satisfaction as having positive attitudes
towards one's job.
Linking the relationship between job performance and
job satisfaction is difficult because of the numerous
factors that affect the direction and magnitude of the
relationship [Vroom, 1964]. In their study on job
satisfaction and employee turnover, Bullen and Flamholtz
[1985] stated:
Other factors which may affect job satisfaction and turnover that have not been tested by the researchers include performance, ability, personality, and certain other characteristics.
geographical location, and general economic conditions, [p. 289]
A job satisfaction study concluded that job
performance causes job satisfaction [Vroom, 1964, p.
182]. Lawler and Porter [1967] took a position contrary
to Vroom [1964] and hypothesized that job satisfaction
directly affects job performance. In a subsequent study.
Porter and Lawler [1968] then agreed with Vroom [1964]
that job performance causes job satisfaction. Finally,
after offering conflicting theories on the causality
between the two constructs, Lawler [1971] discussed the
difficulty of testing and developing theories of job
satisfaction and job performance.
There have been few studies on the causality between
job satisfaction and job performance among accountants.
Using accountants as subjects for a study on job
satisfaction and job performance, Ross and Bomeli [1971]
stated:
. it is not clear whether high satisfaction with autonomy needs results in high performance and quality or whether high performance and quality result in high satisfaction. [p. 384]
In another study on job satisfaction among accountants,
Benke and Rhode [1980] stated:
. job satisfaction and performance are not necessarily positively related. . . . Consequently, job satisfaction [and] job performance . . . may be interrelated in a complex manner, [pp. 187-88]
While an investigation of causality for the job
satisfaction and job performance question is an important
one, it is an empirically unresolved issue. As a result
of the conflicting evidence the question of causality
between job satisfaction and job performance was not
resolved by this research study. Rather, the relationship
between these two constructs was investigated.
Need for the Study
The importance of maximizing job satisfaction and job
performance among employees within organizations is
continually sought by managers. Maximizing job
satisfaction and job performance may lead to higher
productivity, lower overall costs for the firm, and
reduced turnover of employees. Public accounting firms
experience annual non-partner turnover rates as high as
45%, primarily at the entry-level staff accountant
position [Bao, Bao, and Vasarhelyi, 1986]. Rhode,
Sorensen, and Lawler [1976] reported turnover rates of 85%
for initial hires over a ten year period. Lcimpe and
Earnest [1984] reported an annual turnover rate of 23.9%
for staff accountants who had one to three years'
experience with a firm. Their reported annual turnover
rate was higher than any of the prior five year annual
turnover rates. This indicates that accountants are
leaving firms at earlier stages in their career. The cost
of replacing an entry-level staff accountant has been
estimated at between $10,000 and $15,000 [Bullen and
Flamholtz, 1985]. While attrition may be necessary in the
public accounting profession due to the limited number of
supervisory and managerial positions, early detection of
job dissatisfaction and low levels of job performance,
which can be caused by other factors, can diminish the
premature or unexpected attrition of employees whom the
firm would like to retain. In addition, better
understanding of the relationship between job satisfaction
and job performance may lead to minimizing expenditures to
increase job performance, job satisfaction, and costs
associated with employee turnover.
Purpose of the Study
This purpose of this study was to contribute to the
academic literature in the areas of job satisfaction and
job performance, and provide useful information to
practitioners in public accounting firms. This will be
accomplished by investigating constructs that can affect
job satisfaction [Bullen and Flamholtz, 1985] and job
performance cimong staff accountants in a public accounting
firm. The effects on job satisfaction and job performance
of organizational commitment, professional commitment,
life-span career development, and self-monitoring were
examined. These constructs are discussed in depth in
Chapter II.
The relationships among these constructs are
presented in three research objectives: (1) investigate
the effects of organizational commitment, professional
commitment, life-span career development, and self-
monitoring on job satisfaction and job performance for
staff accountants in public accounting firms; (2)
investigate the relationship between job satisfaction and
job performance of staff accountants in public accounting
firms; and (3) investigate the relationship among
organizational commitment, professional commitment, and
life-span career development for staff accountants in
public accounting firms.
Organization of the Study
This study is further divided into five chapters.
Chapter II is a review of the relevant works for this
study relating to job satisfaction, job performance,
professional and organizational commitment, life-span
career development, and self-monitoring. Chapter III
contains the theoretical model, the research model, and
the development of the hypotheses. In Chapter IV, the
research design, the pretest results of the research
instrument, and statistical tools for this study will be
presented. Chapter V contains the statistical analysis of
this study. In Chapter VI the summary and conclusions are
presented.
CHAPTER II
LITERATURE REVIEW
Much of the research on job satisfaction and job
performance use these constructs as dependent variables
and investigate the impact of numerous independent 1
variables. As for the research evidence, subjects were
chosen typically based on the field or profession under
study. For example, accounting researchers used
accountants, psychologists tended to use students or the
general public, and management researchers used managers
as the subjects for their studies.
Except for self-monitoring, which has been used
primarily in psychological research, the constructs for
this study were selected from existing accounting research
evidence. For example, to determine the effects of
budgetary participation on attitudes and behavior,
Chenhall and Brownell [1988] examined the intervening
effect of role ambiguity on job satisfaction and job
performance. In a study where auditors were the subjects,
Choo [1986] used personality variables to examine their
1 Edwin Locke estimated a conservative figure of 3,350
articles and dissertations on job satisfaction from the period 1957-1972 [Dunnette, 1983]. This indicates the lack of conclusive evidence and continued interest in the area of job satisfaction.
relationship with role stress, and the effect of role
stress on job performance. Rasch and Harrell [1990]
investigated the impact of personal characteristics such
as job satisfaction, on determinants of job turnover.
Ferris [1981] investigated the effect of organizational
commitment on job performance eimong accountants in public
accounting firms. Harrell and Eickhoff [1988] researched
the effect of commitment on the level of job satisfaction
among auditors. The emphasis on accounting related
literature is because the sample under investigation for
this study were staff accountants in a public accounting
firm, and the relationship between this study and prior
research evidence was on constructs that may affect job
satisfaction and job performance.
This literature review is presented in the following
order: (1) job satisfaction and job performance;
(2) professional and organizational commitment, life-span
career development, and self-monitoring; and (3) a
discussion of how new empirical research evidence may
contribute to the accounting body of knowledge and the
accounting profession.
Job Satisfaction and Job Performance
Lawler and Porter [1967] recognized a consistent
relationship between job satisfaction and job performance
This relationship has been explained by the introduction
8
of such variables as intrinsic and extrinsic rewards
[Vroom, 1964], positive and punitive reward behavior [Sims
and Szilagyi, 1975], situational performance constraints
[Phillips and Freedman, 1984], and leadership style and
to be associated with the job itself, environmental
factors, and individual factors. Landy, Zedeck, and
Cleveland [1983] used organizational considerations,
individual considerations, and sociopolitical
considerations as the primary factors affecting job
performance. Different theories behind the relationship
between job satisfaction and job performance have been
well-documented (for example, Vroom, 1964; Lawler and
Porter, 1967; Porter and Lawler, 1968; Lawler 1971; Ross
and Bomeli, 1971; Benke and Rhode, 1980; Harrell and
Stahl, 1984; Bullen and Flamholtz, 1985). There is no one
theory to explain this relationship because of the large
number of variables that affect job satisfaction and job
performance.
Vroom [1964] discussed the difficulty of describing
the relationship between job performance and job
satisfaction because of the numerous factors that affect
the direction and magnitude of the relationship. In this
study, the factors of interest are the constructs
45
organizational commitment, professional commitment, life
span career development, and self-monitoring.
The model developed by Bullen and Flamholtz [1985]
shows the complexity of attempting to explain the causes
of job satisfaction because of the numerous variables
affecting this construct. Figure 3.1 is a portion of the
theoretical model developed by Bullen and Flamholtz.
Overall Assessment of
Job Satisfaction /
\
\
/
Area Level
Career Goals Demographic s Other Factors
Figure 3.1. The Theoretical Model
"Other Factors" include performance, personality, and
certain other personal characteristics [Bullen and
Flamholtz, 1985]. Investigating the interrelationships
among the constructs is the primary purpose of this study.
The Research Model
Researchers have theorized and empirically tested the
relationship among job satisfaction, job performance, and
other variables [for example, Bagozzi, 1980a]. The
research model used in this study was adapted from the
Bagozzi [1980a] research model, and based on the theory
proposed by Bullen and Flamholtz [1985]. The first
46
research objective of this study was to investigate the
hypothesized (H) effects of organizational commitment,
professional commitment, life-span career development, and
self-monitoring on job satisfaction and job performance
for staff accountants in public accounting firms. Figure
3.2 is the research model for these relationships.
Professional Commitment Organizational Commitment Life-Span Career Development
Self-Monitoring
Job Satisfaction
Job Performance
Figure 3.2. The Research Model for H1-H8
The second research objective was to investigate the
relationship between job satisfaction and job performance
of staff accountants in public accounting firms. This
relationship is illustrated in Figure 3.3.
Figure 3.3. The Research Model for H9
47
Research objective three was to investigate the
hypothesized relationship among organizational commitment,
professional commitment, and life-span career development
for staff accountants in public accounting firms. These
relationships are illustrated in Figure 3.4.
Professional Commitment
/
\
\
/
Life-Span Career
Development
\ /
\ /
Organizational Commitment
Figure 3.4. The Research Model for H10-H12
Formulation of the Hypotheses
The research objectives stated in Chapter I were to:
(1) investigate the effects of organizational commitment,
professional commitment, life-span career development, and
self-monitoring on job satisfaction and job performance
for staff accountants in public accounting firms; (2)
investigate the relationship between job satisfaction and
job performance of staff accountants in public accounting
firms; and (3) investigate the relationship among
organizational commitment, professional commitment, and
48
life-span career development for staff accountants in
public accounting firms.
The hypotheses for the first research objective are
predicated on the basic theory that there are many
variables such as personality and certain other personal
characteristics that can affect job satisfaction and job
performance. The first four hypotheses are based on the
effects of organizational commitment and professional
commitment on job satisfaction and job performance.
Models illustrating commitment have antecedents that
precede commitment, and job-related outcomes resulting
from the level of commitment individuals possess [Hunt,
Chonko, and Wood, 1985]. These job-related outcomes
include performance and satisfaction [Hunt, Chonko, and
Wood, 1985]. Ferris [1981] hypothesized that
organizational commitment preceded job performance among
auditors. He reported that junior-level auditors'
performance was influenced by a willingness to exert
effort on behalf of the organization. Few empirical
accounting studies have examined the effects of both
organizational and professional commitment on job
satisfaction. Norris and Niebuhr [1983] reported that
organizational and professional commitment were strongly
related to job satisfaction for individuals employed in
public accounting firms. However, the results reported by
49
Norris and Niebuhr [1983] were for all levels of employees
in all functional areas of the firm. The hypotheses, in
alternate form, for the relationship between
organizational and professional commitment with job
satisfaction and job performance are:
HI: Job satisfaction and organizational commitment will be positively related for a staff accountant.
H2: Job performance and organizational commitment will be positively related for a staff accountant.
H3: Job satisfaction and professional commitment will be positively related for a staff accountant.
H4: Job performance and professional commitment will be positively related for a staff accountant.
Super [1981] stated that there is a lack of empirical
studies on vocational maturity in adulthood. To
investigate vocational maturity in adulthood, these
hypotheses are based on the effects of life-span career
development on job satisfaction and job performance. The
rate and level of an individual's development in his/her
career—vocational maturity—has an effect on satisfaction
[Osipow, 1983]. The relationship between job performance
and life-span career development was proposed by Super
[1957] when he stated that,
. the learning of job or position performance is necessarily much more specific than in the learning of occupational performance (and) virtually all workers have some problems of induction and
50
transition when they begin a new job, and in the inexperienced worker these are often of some significance, [p. 119]
The implementation stage of life-span career
development was investigated. Theoretically, the
implementation stage will occur between the ages of
twenty-one to twenty-five. This is the usual age in which
an individual will enter the accounting profession as a
staff accountant, and go through the induction and
transition processes. Super [1957] theorized a
relationship between life-span career development and both
job satisfaction and job performance during the
implementation stage when he stated that
. [recent college graduates] change in search of something more satisfying to themselves [job satisfaction] and in which they are more satisfactory to employers [job performance], [p. 129]
The hypothesized relationships, in alternate form, among
life-span career development, job satisfaction, and job
performance are:
H5: A positive relation exists between job satisfaction and the level of the implementation stage in a staff accountant's life-span career development.
H6: A positive relation exists between job performance and the level of the implementation stage in a staff accountant's life-span career development.
Hypotheses seven and eight are based on the
relationship between job satisfaction and self-monitoring.
51
and job performance and self-monitoring. Self-monitoring
is used in psychology to explain different attitudes and
behaviors involving interactions.
Because of staff accountants' numerous interactions
with others, self-monitoring should be a construct for
inclusion in this study on job satisfaction and job
performance. If staff accountants who are low self-
monitors do not feel that they fit in a continually
changing work environment, they may be less satisfied with
their job. Their perceived level of job performance can
also be lessened by their perceived lack of adaptability.
Douglas [1983] reported that high self-monitors view
conversations (interactions) as being strategic. Staff
accountants who are high self-monitors should be more
strategic and adaptable than low self-monitors, which can
lead to greater levels of job satisfaction and higher
perceived job performance. Ferris [1982] reported that an
accounting organization's ability to adapt to the outside
environment was directly related to employee job
performance. The ability of staff accountants to adapt to
their environment may also affect their level of job
performance and job satisfaction. Hypotheses seven and
eight are presented below in alternate form.
H7: A positive relation exists between the level of self-monitoring and job satisfaction for a staff accountant.
52
H8: A positive relation exists between the level of self-monitoring and job performance for a staff accountant.
The results of previous empirical research studies
show that job satisfaction and job performance tend to
correlate significantly [Bagozzi, 1980]. This hypothesis
results in theorizing a nonrecursive model in which the
interrelationship between the two constructs is
investigated. This differs from a recursive model in
which one of the constructs precedes or "causes" the
second construct. A nonrecursive model is used because
the theories regarding the causal relationship between job
satisfaction and job per*formance are uncertain and
conflicting [Vroom, 1964; Lawler and Porter, 1967; Porter
and Lawler, 1968; Lawler, 1971; Ross and Bomeli, 1971;
Benke and Rhode, 1980]. The second research objective
hypothesizes, in alternate form, the following:
H9: There is a positive nonrecursive relation between a staff accountant's job satisfaction and job performance.
As stated previously, professional commitment has
been found to precede organizational commitment,
supporting hypothesis ten [Aranya, Lachman, and Amernic,
1982; Lachman and Aranya, 1986]. Hypotheses eleven and
twelve are based on the level of satisfaction an
individual has with the career choice of becoming a
certified public accountant. If an individual is
53
satisfied with his/her career choice, professional
commitment and organizational commitment should be
significantly related in a positive manner with life-span
career development. The last research objective
hypothesizes, in alternate form, the following:
HIO: A staff accountant's organizational commitment will be directly predicted by professional commitment.
Hll: There is a positive relation between professional commitment and life-span career development for a staff accountant.
H12: There is a positive relation between life-span career development and organizational commitment for a staff accountant.
Chapter Summary
Figure 3.5 is the research model for the twelve
hypotheses presented in this chapter. These hypotheses
are based on the theoretical model developed by Bullen and
Flamholtz [1985]. In addition to the twelve hypotheses
under investigation, demographic data such as sex, age,
and years as a staff accountant were collected. While
there are no hypotheses presented for the demographic
data, analyses of the data will be conducted. Chapter IV
contains the research design, the results of pretesting
the research instrument, and the statistical analytical
technique for this study.
54
Figure 3.5. The Research Model for all Hypotheses
55
CHAPTER IV
RESEARCH METHODOLOGY
Having discussed the continuing interest in the
relationship between job satisfaction and job performance,
and the constructs that affect job satisfaction and job
performance, this chapter describes the development of the
approach to test the research model. The analyses of the
data are presented in Chapter V.
The three purposes of this chapter are to: (1)
explain the research design; (2) discuss the pretesting of
the research instrument; and (3) introduce the statistical
techniques that were used to analyze the data.
Research Design
This research study is a field study. Kerlinger
[1986] defines field studies as "nonexperimental
scientific inquiries aimed at discovering the relations
and interactions among sociological, psychological, and
educational variables in real social structures" (p. 372).
He further refines his definition to include studies that
pursue relations and test hypotheses that are done in life
situations such as businesses. Field studies can be
either exploratory, seeking what is occurring, or
hypothesis-testing. This study comes under the latter
56
category and its aim is to uncover relationships among the
constructs.
Field studies have several strengths such as realism,
significance, strength of variables, theory orientation,
and heuristic quality. Weaknesses of field studies
include the large number of variables and variance, which
can influence the study and the lack of precision in
measurement [Kerlinger, 1986]. An attempt was made to
control for extraneous variables that may affect the study
by using a single office of a large public accounting
firm. Using only one office of the firm controlled for
the variation that could occur due to spatial
autocorrelation, a problem common to studies that sample
firms in different geographical locations. Types of
variation that are controlled for by using only one firm
include personnel differences, the leadership style of the
managing partner, and the office environment. Measurement
instruments were refined to maximize the precision of the
instruments (see Chapter V).
The Population for the Study
The subjects used in this study are employed at a
national public accounting firm located in a large
midwestern city. The firm categorizes levels within the
organization as partner, manager, or staff. Within the
level of staff are the categories senior and assistant.
57
To test the hypotheses of this study, staff
accountants in a public accounting firm who had at least
six months of public accounting experience were selected.
This limitation on subject selection is imposed because
some training and relevant employment—employment specific
to the chosen career—should occur during the
implementation stage of career development [Osipow, 1983].
Development of the Research Instrument
A research instrument was used in this study to
ensure quantifiable data, maintain equivalency of
responses, and because of the sensitive nature of the
material. In addition to the measurement scales used for
the constructs under study, the research instrument also
included requests for demographic information such as age,
sex, position with the firm, years as a staff accountant,
years with the firm, professional certifications, and
education. The demographic data were used as a screening
tool to ensure that the respondents held both the staff
position in the firm and had at least six months of public
accounting experience, as required for this study.
Participation in the study was voluntary, and no attempt
was made to identify respondents by the demographic data
supplied. A copy of the research instrument, a letter to
the subjects informing them of the study, and a letter
58
from a partner in the firm encouraging them to participate
in the study are presented in Appendix A.
Federal regulations require material that will be
used on human subjects to be reviewed. It is a Texas Tech
University policy to review all research involved with
human subjects. The research instrument was submitted to
the Texas Tech University Committee for the Protection of
Human Subjects for review and approval was received. A
copy of the letter granting approval is presented in
Appendix B.
Job Satisfaction Measurement
The Minnesota Satisfaction Questionnaire (MSQ)
[Weiss, Dawis, England, and Lofquist, 1967] was used to
measure job satisfaction. Price and Mueller [1986] stated
that the MSQ was developed to link both measurement and
theory. This scale has been used in accounting studies to
measure job satisfaction (for example, Brownell, 1983).
Weiss et al. [1967] report construct validity based on
the results of testing the instrument within occupations,
and between different occupations. A general internal
reliability coefficient of .90 is reported [Weiss et al.,
1967, p. 23].
59
Job Performance Measurement
Employee performance can be actual or perceived. It
can be further subdivided into historical, expected, or
aspired performance. In public accounting firms, employee
performance is usually based on a supervisor's judgments
and is supported by the employee's (subordinate)
historical actions.
National public accounting firms are reluctant to let
confidential information such as performance evaluations
be seen or used by individuals outside of the firm. An
alternative is to use self-reported performance
evaluations. Self-reported performance evaluations have
been used in accounting studies (for example, Brownell and
Mclnnes, 1986; Choo, 1986; Chenhall and Brownell, 1988).
A potential bias in the use of self-reported performance
evaluations is the inflation of one's performance or a
leniency in self-evaluation [Prien and Liske, 1962;
Lawler, 1968; Thornton, 1980]. Other studies have shown
the opposite effect in self-rating performance evaluations
[Heneman, 1974; Teel, 1978].
There is also the possibility of bias when
supervisors' ratings are used to measure job performance.
Ramanathan et al. [1976] and Jiambalvo [1982] have
reported that there are differences in the perceived
importance of the criteria used in performance
60
evaluations. This may indicate that in studies where
individualistic factors are related to job performance, an
individual's perceived job performance may better relate
to these factors than a superior's job performance rating.
Ferris and Larcker [1983] reported that rewarded
performance (compensation) was a function of both the
interpersonal attraction between the auditor and his/her
supervisor and the task-related ability of the individual.
Therefore, rewarded job performance may be biased based on
the interpersonal attraction between the auditor and
supervisor. It appears that there is potential for bias
using either supervisor or self-rating performance
evaluations [Landy and Farr, 1983]. Since supervisory
performance evaluations were not made available, and there
is a potential for bias in either self-reported or
study was developed by Choo [1986], specifically to
measure accountants' job performance. There was no
statistically significant difference found between self-
reported performance evaluations and supervisor
evaluations. The mean score for self-reported performance
evaluations was .50 less than the mean score reported by
61
superiors [Choo, 1986, p. 26]. This indicates a lack of
leniency bias in the instrument and reported scores.
Organizational and Professional Commitment Measurement
Organizational commitment was measured using the
Organizational Commitment Questionnaire (OCQ) developed by
Porter, Steers, Mowday, and Boulian [1974]. Mowday,
Steers, and Porter [1979] were critical of organizational
commitment scales because they believed most were created
on an a priori basis, with little, if any, reliability and
validity data presented. Using the OCQ, a variety of
analyses were carried out using different samples [Mowday
et al., 1979]. Acceptable levels of reliability and
convergent, discriminant, and predictive validity were
evident for the OCQ. This scale has been used in the
accounting literature for studies measuring organizational
commitment (for example, Aranya et al., 1981a; Aranya et
al., 1982; Norris and Niebuhr, 1983; Aranya and Ferris,
1984).
The Professional Commitment Scale (PCS) developed by
Aranya et al. [1981a] was used in this study. The PCS was
primarily adapted by taking the OCQ and substituting the
word profession for organization. In their study, Aranya
et al. [1981a] reported a coefficient alpha of .87.
62
Nunnally [1978] recommends alpha values of .80 or above to
be good indicators of reliability.
Since the organizational and professional commitment
scales are similar, there is the possible effect of common
method variance on the correlation between the two
measures. Aranya and Ferris [1984] used these two scales
in a study and found the correlation between them did not
indicate a strong response bias.
Bline, Duchon, and Meixner [1991] investigated the
psychometric properties of the professional and
organizational commitment scales used in this study.
Their results provide support that the scales do measure
two different constructs.
Life-Span Career Development Measurement
An individual can be satisfied or dissatisfied with
his or her career choice. Osipow [1983] reports that
dissatisfied individuals would be more likely to consider
a career change than satisfied individuals (p. 220).
Personal interviews or longitudinal studies have been used
to determine an individual's life-span career development
[Osipow, 1983].
An alternative to the interviewing process is to
develop a scale that would measure the level of an
individual's career stage. This study used subjects who
should be in the implementation stage of their career
63
development. Therefore, the scale must contain questions
that will allow the researcher to ascertain whether the
individual is in the implementation stage. The scale
developed for this study is based on Super's [1957] six
descriptors for adjusting to work requirements in the
implementation stage. The six descriptors are technical
competence, the routine or tempo of the work, the work
load, work attitudes and roles, security, and time-in-job
before advancement. The questions were derived from the
characterizations given by Super [1957] for the six
descriptors. The characterizations used for the
questionnaire were limited to those that dealt with career
development, to avoid questions that may measure job
satisfaction.
Self-Monitoring Measurement
Snyder's [1974] Self-Monitoring Scale was used to
measure an individual's level of self-monitoring. This
scale requests the subjects to answer twenty-five true-
false self-descriptive statements. Subjects are then
categorized into a low group or high group based on their
responses to the questions. The technique used to
distinguish between high and low self-monitors is to use a
median cutoff (for example, Snyder, 1974; Douglas, 1983).
The upper fifty-percent are considered high self-monitors
64
of the sample, the lower fifty-percent are considered low
self-monitors of the sample.
Research Instrument Distribution and Collection Technique
The researcher was not allowed the time needed to
meet with all of the subjects at the branch and home
offices, as many of them went to the offices only once a
month because they were at remote areas working on audits.
Therefore, for both the pretest and study, a
representative of the firm distributed the research
instrument, which included a letter from the researcher
describing the study, a letter from a partner in the firm
encouraging subjects to participate in the study, and a
self-addressed stamped envelope to return the research
instrument. This was done through the office manager at
the branch office, and a partner at the home office. For
subjects who did not pick up their mail within-one week of
the distribution of the research instrument, the firm
mailed the research instrument and self-addressed stamped
envelope to their home address. The researcher verified
with the firm's office manager and partner assigned to the
study that this procedure was done. The office manager
and partner sent a follow-up memo requesting completing
and mailing the research instrument to all subjects in
65
both the pretest and study. All completed questionnaires
were mailed directly to the researcher's home.
The distribution of the questionnaires by a
representative of the firm, rather than by both the
researcher and a representative of the firm, controls for
potential experimenter effects. These effects can occur
if the researcher distributes and collects some of the
questionnaires, and a representative of the firm
distributes and collects some of the questionnaires.
Subjects may respond differently to questionnaires based
on the individual who distributes the questionnaires.
This distribution and collection technique has been used
in previous accounting research at public accounting firms
[Sutton and Lampe, 1988].
Pretesting the Research Instrument
Hunt, Sparkman, and Wilcox [1982] list three
categories of items that should be pretested: "(1) items
about the questionnaire itself, (2) items about specific
questions, and (3) items about data analysis" (p. 269).
Converse and Presser [1986] expanded these three
categories into ten specific purposes for pretesting a
questionnaire. The first four purposes are tests for
specific questions: variation (data analysis), meaning,
task difficulty, and respondent interest and attention.
The last six purposes bear on the questionnaire as a
66
whole: flow and naturalness of the sections, the order of
questions, skip patterns, timing, respondent interest and
attention (overall), and respondent well-being.
The research instrument was pretested in a branch
office of the national firm used in this study. The
branch office was located approximately ninety miles from
the home office where this study was conducted. The close
proximity between the offices decreased the possibility of
spatial autocorrelation, which could result from using two
or more offices or firms geographically far apart.
Table 4.1 is the summary of the research instrument
response rates from the pretest. Twenty research
instruments were distributed by the branch office manager
to the staff accountants. Initially, the subjects were
not made aware that they were participating in a pretest;
however, at the end of completing the research instrument,
subjects were requested to list any comments they had
regarding the questions and the overall research
instrument. Subjects were instructed to complete the
research instrument and mail their responses in self-
addressed stamped envelopes, which were provided. These
procedures were used at the branch office to simulate the
conditions that would be employed at the home office for
this study. Within nine days, sixteen of the
questionnaires were returned, of which fourteen were
67
TABLE 4.1 SUMMARY OF QUESTIONNAIRE RESPONSE RATES FROM THE
PRETEST FOR ALL SUBJECTS
1 1 Staff Managers Partners Total
Questionnaires distributed
No Response
Responses
Responses not usable for data analysis
Responses usable for data analysis
20
4
16
2
14
6
1
5
1
4
3
1
2
0
2
29
6
23
3
20
Response rate based on questionnaires distributed 80% 83% 67% 80%
Research instruments were distributed to managers and partners to pretest the life-span career development measurement instrument.
68
usable. Two questionnaires were not usable because the
subjects were employed at the firm for substantially less
than six months. Four managers and two partners completed
questionnaires as part of the pretest. This was to
compare mean responses on the measurement instrument for
life-span career development. Hunt et al. [1982, p. 270]
reported that the size of the pretest sample is not fixed,
and that sample sizes of twelve, twenty, and thirty have
been reported to be satisfactory.
The first purpose of pretesting the specific
questions, variation (data analysis), was met by examining
Pearson correlation coefficients for the constructs
professional commitment, organizational commitment, life
span career development, job satisfaction, and job
performance. No changes were made to the research
instrument based on this analysis. Since self-monitoring
is measured on a dichotomous scale, high or low, it was
excluded, from this analysis. PRELIS and LISREL (linear
structural relations) VII, the statistical packages used
in this study, allow the use of continuous and dichotomous
constructs in the Scime model. However, the number of
responses in the pretest was too small to use LISREL VII.
This posed no problem for this study, since the number of
responses from the home office was large enough to use
LISREL VII.
69
The remaining three purposes of pretesting the
specific questions—meaning, task difficulty, and
respondent interest and attention—were analyzed by
reading the comments provided by the subjects. There were
no comments regarding these three areas of the pretest.
The last six purposes of the pretest—flow and
naturalness of the sections, the order of questions, skip
patterns, timing, respondent interest and attention
(overall), and respondent well-being—were also analyzed
by reading the comments provided by the subjects. Based
on the subjects' responses, two areas that rec[uired
changes in the research instrument were the ordering of
the questions, and the form requesting demographic data
and job titles.
Initially, the questions for organizational and
professional commitment were sequential. Some of the
subjects assumed they were the same questions because of
the similarities between the questions. This was changed
for the research instrument distributed to the home office
by separating the two sets of questions, and inserting a
statement that the questions were not the same as the ones
asked previously. A second change was made regarding the
ordering of the demographic data, and the wording of the
job titles.
70
No other changes were made to the research instrument
after the pretest. On average, it took the subjects
fifteen minutes to complete the research instrument, with
a range of nine minutes to twenty minutes.
Statistical Technique
The constructs under study are latent or unobserved.
In their purest form they are abstract unidimensional
concepts, making them only indirectly measurable [Bollen,
1989]. The latent variable model consists of the
structural equations that summarize the relationships
among the constructs, and the measurement model. The
totality of relationships among the constructs is usually
referred to as the structural equation, and their
measurement as the measurement model. This is misleading
because it indicates that the measurement model is not
structural. While this is not the case, to differentiate
between the relationship among the constructs and the
measurement model, this study will use the traditional
approach of referring to the relationships among the
constructs as the structural model, and the measurement of
these constructs as the measurement model.
The primary statistical tools used to analyze the
data in this study are PRELIS and LISREL VII. PRELIS and
LISREL VII were developed by Joreskog and Sorbom [1988,
1989a, 1989b]. PRELIS is data screening software that can
71
read grouped and patterned data. In addition to data
screening, PRELIS computes correlations and matrices which
can be used as input files for LISREL VII. The necessity
of using PRELIS in this study is that it can compute
correlations and other product moments for continuous and
dichotomous variables that other statistical programs
cannot produce [Joreskog and Sorbom, 1988]. This will
allow the dichotomous construct self-monitoring to be
analyzed with the continuous constructs in this study.
LISREL VII is used to analyze the structural and
measurement models. LISREL VII will minimize the
difference between the sample covariances and the
covariances predicted by the model. An advantage to using
LISREL VII with the data for this study is that it will
integrate measurement concerns with the structural
equation model. This is done by incorporating both the
measured indicator variables with the latent theoretical
concepts into a single structural equation model [Hayduk,
1987, pp. 87-88]. LISREL thus allows for integrating
model development, estimation, evaluation, and
interpretation with measurement concerns [Bohrnstedt,
1983].
LISREL Model Specification
The LISREL model is illustrated in Figure 4.1, and
the LISREL terminology and notations for this study are
72
*l-'n
^t- ^n
^ - ^ i '.6/n
ib.i^yi.i
^)6-'-7
^.6- ^7
Figure 4.1. The LISREL Model
73
presented in Figure 4.2. The model consists of three
latent exogenous constructs (xis)—professional
commitment, life-span career development, and self-
monitoring—which will be predicted outside of the model,
and three endogenous constructs (etas)—organizational
commitment, job satisfaction, and job performance—which
will be predicted within the model. This is done by the
three ecjuations which make up the general structural
equation model. (See Figures 4.3 - 4.5 for the specific
research equations and the general form of these
equations.)
The first equation (Figure 4.3) contains the
hypothesized direct effects among the concepts. These
were estimated by the matrices of structural coefficients
(betas and gammas), the vectors of the endogenous and
exogenous concepts, and a vector of "errors" in the
conceptual model (zeta). The covariances among the zeta
terms will make up matrix psi. Xis will generate a
matrix, phi, which is the covariance among the xis. The
second and third equations (Figures 4.4 and 4.5) link the
conceptual variables to their observed indicators [Hayduk,
1987]. The second equation is the measurement model for
the endogenous indicators, consisting of a matrix of
structural coefficients (lambdas), the vector of the
endogenous concepts, and a vector of "errors" in the
74
S p e c i f i c Terms I n c l u d e d in I h e LISREL Model
^15
><1
P r o f e s s i o n a l Commi tiuen t L i f e S p a n C a r e e r D e v c 1 o p m c n I
S e 1 f - M o n i l o r i ng Orga i i i 2a t i ona i Commi Uiien t J o b S a l i s f a c l i on J o b P e r f o r m a n c e i n d i c a n t s of Piof ess i ona 1 Commi linen I i n d i c a n l s of Life S p a n C a r e e r D e v e l o p m e n t i n d i c a n t s of Se 1 f-Mon i t or i ni; i n d i c a n t s of O r g a n i z a t i o n a l Coimiii tuien t i n d i c a n t s of Job S a t i s f a c t i o n i n d i c a n l s of Job P e r f o r m a n c e
D e f i n i t i o n s of L I S R E L ' S G e n e r a l T e r m s
I Tl X y p q (U n
(xi) (ela)
an unobservable (latent) cxog^enous variable an unobservable (latent) cndogejtous variable an observed indicant of an exogenous variable an observed indicant of an endogenous variable the number of y-variables the number of x-variables the number of T^-variables the number of ^-variables
Parameter Matrices
Ay (Larabda-y)
Ay (Lambda-x)
B (Beta)
r (Gamma)
i (Phi)
Y (Psi)
Sj ( T h e t a - E p s i I o n )
e, (Theta-Dclta)
a pxm among a qxn among a mxm among a mxn among a nxn among a mxm among a pxp among a qxq among
matrix of y ' s and T
matrix of x's and ; matrix of
n matrix of ; ' s and T\ ma t r i X of
the I m a t r i x of
I he
the
the
the
the
the
the T)' s e r r o r m a t r i x of the the y's er r o r m a t r i x of the
the x's er r o r
r e 1 u I i oiisii i ps
re 1 a t i o n s h i p s
r e l a t i o n s h i p s
r e l a t i o n s h i p s
r e i a l i o n s h i p s
r e l a t i o n s h i p s Ierms (C's ) re1ali o n s h i p s lerms ( € ' s ) r e l a l i o n s h i p s Ierms ( 6 ' s )
C34.3 A i ; . . - - - J A . • . . . b • . A . • • . . . . . A . r •
A . 5 •
A i l , • . *••« A 11 • . ' i • • / • I • . , . . . . A I . t
. «... A I • .
/ • I • •
. ' • •• / • I r •
4 • • rf
f
f
f
^( t T
^ E T
r
r
6 ' 0
"11 f U t «
"n
"IE
"11 "1?
"20
" I I
6=' 24
?31
^31 f
6 : :
"34 ^33 ^ ; ; 6 , . hi "3" 'ii "«3
6.; H
" U f
"4 i
Figure 4.5. The Measurement Model for X Indicants
78
measurement model (epsilon). The covariance among the
epsilon terms constitute matrix theta epsilon. The third
equation is the measurement model for the exogenous
indicators, consisting of a matrix of structural
coefficients (lambdas), the vector of exogenous concepts,
and a vector of "errors" in the measurement model (delta).
The covariance among the delta terms constitute matrix
theta delta.
To test the model, a two-step approach is used
[Anderson and Gerbing, 1988]. This approach evaluates the
measurement model separately from the full structural
equation model. The first step is to evaluate the
measurement model for unidimensionality of the measurement
constructs, and respecify the model if necessary [Anderson
and Gerbing, 1988]. The second step is to simultaneously
test the measurement and structural model after
respecification. These steps are presented in Chapter V.
Chapter Summary
The research design, pretesting of the research
instrument, and the statistical techniques used in this
study were presented in this chapter. The research design
is based on field study research [Kerlinger, 1986].
Pretesting the research instrument was conducted as
closely as possible to the distribution and collection
techniques used in this study. There were no significant
79
changes made to the research instrument or planned
distribution and collection techniques based on the
pretest. Based on the constructs being investigated,
structural equation modeling with a two-step approach was
selected to analyze the data, using PRELIS and LISREL VII
Chapter V contains the analyses for this study, including
response rates, testing for nonresponse bias, and
statistical testing of the measurement models and
structural equations.
80
CHAPTER V
EMPIRICAL ANALYSIS OF THE DATA
The study to this point has used the theories of
Bullen and Flamholtz [1985] and Bagozzi [1980] to
construct a research model to investigate the
interrelationships among variables that can affect job
satisfaction and job performance, as well as the
relationship between job satisfaction and job performance
The research design, population, research instrument,
pretest results, and statistical procedures used in this
study were presented in Chapter IV. In the first part of
this chapter the response rates and the test for
nonresponse bias are presented. The second part presents
the statistical analyses used in the study to test the
research model (see Figure 3.5). This analysis includes
testing and refining the measurement models, simultaneous
tests of the measurement and structural models, and tests
of the individual structural coefficients.
Response Information
Response Rates
The study was conducted at a national public
accounting firm located in a large midwestern city. The
distribution and collection of the research instrument
were described in Chapter IV. The firm employed 421
81
accountants classified as staff at the time of this study.
Of these, 312 responded, for a 74% response rate.
Listwise deletion is used in this study, so any
respondents not answering all of the questions, or meeting
the experience requirements for this study were eliminated
(see Statistical Analysis, p. 84). This reduces the data
matrix to a matrix without any missing observations. This
resulted in 31 responses being eliminated for a usable
response rate of 67%. Table 5.1 presents the response and
elimination rates.
Test for Nonresponse Bias
Armstrong and Overton [1977] recommend the best
protection against nonresponse bias is the reduction of
nonresponse itself. However, when nonresponse is present,
subjects who respond late or less readily are assumed to
be more like nonrespondents [Armstrong and Overton, 1977].
In this study, the nonresponse rate was 26%, possibly due
to the distribution and collection techniques. To test
for nonresponse bias, a comparison of the means of early
and late respondents was undertaken. This testing was
done by using multivariate analysis of variance.
Multivariate analysis of variance is used when means on
several different variables are being compared
simultaneously, as in this study [Bray and Maxwell, 1985].
82
TABLE 5.1 SUMMARY OF QUESTIONNAIRE RESPONSES AND ELIMINATIONS
Number Percent
Total Questionnaires Distributed
Questionnaires Returned 1
Questionnaires Eliminated
Usable Responses
421
312
31
281
100%
74%
7%
67%
1 Questionnaires were eliminated for candidates who were
not employed at the firm for at least six months, or who did not answer all of the questions pertaining to the measurement of the constructs. This latter requirement is because of the use of listwise deletion.
83
Table 5.2 presents the nonresponse bias test result.
The null hypothesis for testing nonresponse bias is that
there is no difference in the mean responses between the
early and late respondents. Wilks' Lambda results in not
rejecting the null hypothesis at the .05 level. No
significant nonresponse bias is indicated by the results
of the test statistic.
The demographic data and descriptive statistics are
presented in Appendix C. Statistical tests were conducted
for differences in mean responses for the demographic
data. There were no significant differences between mean
responses based on the demographic data at the .05 level.
Statistical Analysis
Testing for Identification
Equations in a model must be identified if meaningful
results are to be obtained [Fisher, 1966]. Berry [1989]
stated:
An equation is identified if the acquisition of data for a properly chosen infinite sample would be sufficient to determine a unique set of parameters for the equations; if the equation is identified, only one explanation (i.e., one set of parameters) would be consistent with both the data from the infinite sample and the model's restrictions, [p. 25]
LISREL VII checks for the identification of the model by
making tests in the following way [Joreskog and Sorbom,
1989b]:
84
TABLE 5.2 NONRESPONSE BIAS TEST RESULT
Test F-Test PR > F
Wilks' Lambda .9434 .4548
85
The program checks the positive definiteness of the information matrix. (The information matrix is the probability limit of the matrix of second order derivatives of the fit function used to estimate the model.) If the model is identified, the information matrix is almost certainly positive definite. If the information matrix is singular, the model is not identified and the rank of the information matrix indicates how many parameters are identified, [p. 17]
There were no instances of lack of identification in the
model during the statistical analysis using LISREL VII.
Testing the Measurement Model
Anderson and Gerbing [1988] recommend a two-step
approach to be used for structural equation modeling.
This is accomplished by first testing the measurement
model prior to testing the measurement and structural
submodels simultaneously. The reliability and validity of
the measurement instrument can be assessed by testing the
measurement model in a two-step approach.
Because of the large number of indicants used in this
study, composites of the 95 questions used in the
measurement instrument were formed. The use of composites
and single indicators has precedence when using LISREL as
the statistical tool for data analysis (for example, see
Bagozzi, 1980a and 1980b; Gaski, 1986; Howell, 1987).
Questions 1-70 measured the constructs professional
commitment, life-span career development, organizational
commitment, job satisfaction, and job performance. These
86
seventy questions were grouped into twenty composites,
four for each construct. Self-monitoring was measured by
questions 71-95. Based on the responses to these
questions, subjects were numerically categorized as a low
(1) or high (2) self-monitor. Therefore, a total of
twenty-one composite measurement indicants were used for
the six constructs.
Next, the input files were created for analysis using
LISREL. This was accomplished by using PRELIS, the
predecessor analytical software for LISREL. Since self-
monitoring is a dichotomous construct, and the other
constructs are continuous, Joreskog and Sorbom [1988 and
1989a] recommend using the weighted least squares method
for data analysis. This entails the use of listwise
deletion because there are a different number of cases
(observations) involved in the computations if pairwise
deletion is used [Joreskog and Sorbom, 1989a]. PRELIS
computed Pearson product moment correlations among the
constructs professional commitment, life-span career
chi-square difference tests for the five nested models are
performed after it is first assessed whether a structural
model exists that would have an acceptable goodness-of-
fit.
Testing a nested sequence of models using a chi-
square difference test rather than testing individual
paths is preferred [Anderson and Gerbing, 1988] and is
recommended in research literature (for example, Bentler
and Bonett, 1980; James, Mulaik, and Brett, 1987; Bollen,
1989). Anderson and Gerbing [1988] cite four strengths
that the two-step approach has over a one-step approach.
These strengths are:
First, it allows tests of the significance for all pattern coefficients. Second, the two-step approach allows an assessment of whether any structural model would give acceptable fit. Third, one can make an asymptotically independent test of the substantive or theoretical model of interest. Finally, the two-step approach provides a particularly useful framework for formal comparisons of the substantive model of interest with next most likely theoretical alternatives, [p. 422]
The Structural Model
To determine if any structural model has an
acceptable goodness-of-fit, a "pseudo" chi-square test
statistic is calculated (Anderson and Gerbing, 1988,
103
p. 418). The pseudo chi-square test statistic is
comprised of the results from two different models, the
saturated and null models. The chi-square value of the
saturated model and the degrees of freedom from the null
model are used to test for significance. The saturated
model has the smallest value for any structural model
since all of the parameters are free, while the null model
has the largest number of degrees of freedom since all of
the parameters are fixed. If this test statistic is
significant, then there is no structural model that would
give an acceptable fit. This is because any alternative
models would have fewer degrees of freedom than the null
model, with a chi-square value greater than or equal to
the saturated model.
Table 5.7 presents the results of the pseudo chi-
square test statistic. The test statistic is significant
which could indicate a misspecification of the measurement
model [Anderson and Gerbing, 1988]. The measurement model
was reexamined by investigating modification indices
obtained in the LISREL output, and attempting to further
refine the composite indices. No further respecification
could be done to obtain a measurement model that would
produce an acceptable pseudo chi-square test statistic.
Anderson and Gerbing [1988] stated:
. . . a researcher can build a measurement model that has the best fit from a content and
104
TABLE 5.7 PSEUDO TEST STATISTIC FOR THE NULL AND SATURATED MODELS
Saturated Model Chi-square Test Value 436.74
Null Model Degrees of Freedom 106
Significance Level < .001
105
statistical standpoint, where respecification may have been employed to accomplish this, and still provide a statistical assessment of the adequacy of the theoretical model of interest, [p. 419]
In this study the measurement model is deemed to have the
"best fit." The improvement in fit of the null versus the
saturated model is highly significant as illustrated in
Table 5.8 [Bentler and Bonett, 1980]. Also, the goodness-
of-fit index of .990 demonstrates that the remaining
improvement of .01 (1.00-.990) that might be obtained if
the measurement model is improved is not significant from
a practical viewpoint [Bentler and Bonett, 1980]. The
next step after the structural model has been tested for
goodness-of-fit is to compare alternative models using
sequential chi-square difference tests (SCDTs).
Sequential Chi-sguare Difference Tests
After the saturated model is evaluated, a decision-
tree approach is used to test sequential chi-square
differences for the models. Figure 5.5 illustrates the
order of testing the nested models for SCDTs [Anderson and
Gerbing, 1988]. The null hypothesis to compare a set of
SCDTs is that there is no significant difference between
two nested structural models based on the difference in
chi-square values and the difference in degrees of
freedom. For each SCDT in which the null hypothesis is
not rejected, the model that is more constrained (i.e.,
106
TABLE 5.8 COMPARISON OF THE NULL AND SATURATED MODELS
Chi Significance Model Square Difference D.F Level
Null 10693.69 — 106 1
Saturated 436.74 10256.95 88 .001
1 Difference in degrees of freedom, 18, with difference
in the chi-square value, 10256.95.
107
MC - MS
sign
accept MC
accept MI
acceoc MU
MC - Ms — ^ accept He
sign
V ns MI - .MU ;^—^ Respecify MU as alternaie
mcd<l Hu'; then Ni - Mu'
sicn .-.s Mu - MS
sicn
^ accept M;
Relax ccnstrai.-t m MU t. a: is next jicst-lixeiy model Mu2; then Mu: - MS.
ns - not significant at specified crctability i*v€2 Sign - significant at specified profcatiiity level Mt - theoretical model MS - saturated acdel MC - constrained model MU - unconstrained model
Figure 5.5. Decision-Tree Framework for SCDTs
108
more fixed parameters) would be accepted. To test the
null hypothesis, a chi-square test statistic value is
calculated for the saturated, constrained, theoretical,
and unconstrained models.
The first SCDT test is between the theoretical and
saturated model. The chi-square values and associated
degrees of freedom are presented in Table 5.9. The five
relationships not hypothesized in the theoretical model
(SM->PC, SM->LSCD, JS->OC, JP->OC, and SM->OC) represent
the five degrees of freedom difference. Since
constraining these causal links does not make a
statistically significant difference at the .001 level,
the hypothesis that there is no significant difference
between the two models is not rejected. Therefore, the
second SCDT is between the theoretical and constrained
models.
Constraining additional parameters should be done
from a theoretical perspective by seeking a "weak-link" in
the theoretical model. One method of constraining or
unconstraining paths is to examine the modification
indices provided by LISREL. The modification index is "a
measure of predicted decrease in chi-square if a single
constraint is relaxed and the model is reestimated"
[Joreskog and Sorbom, 1989b]. Theoretical justification
should be the prime criterion for constraining or
109
TABLE 5.9 COMPARISON OF THE THEORETICAL AND SATURATED MODELS
Chi Significance Model Square Difference D.F Level
Theoretical 447.94 93 1
Saturated 436.74 11.20 88 < .05
1 Difference in degrees of freedom, 5, with difference in
the chi-square value, 11.20.
110
unconstraining paths between constructs. Bagozzi and Yi
[1988] stated:
Changing a model by indiscriminatly freeing parameters suggested by modification indices is of course an example of capitalizing on chance. One should avoid relying on such a practice except perhaps in the most exploratory of studies. As a matter of fact, models should not be modified unless one has theoretical and/or methodological reasons justifying any modification, [p. 81]
An examination of the modification indices resulted in a
decision not to constrain or free any of the paths based
solely on these indices.
Based on prior research that also used professionals
as subjects, the path job satisfaction—>job performance
will be constrained. Bagozzi [1980], using professional
salespeople as subjects, hypothesized that job performance
will influence job satisfaction, but that job satisfaction
will not significantly affect job performance. His
theoretical rationale for the job performance—>job
satisfaction sequence is that an individual will form
positive or negative feelings (job satisfaction) after
outcomes on the job (job performance) are obtained, and
compared to the expected and obtained rewards. Bagozzi
[1980] believed that the job satisfaction—>job
performance sequence was more difficult to justify for two
reasons. First, a person must be aware of his/her
feelings (job satisfaction) and attribute these feelings
to specific aspects of the job (job performance). Second,
111
feelings have many antecedents that may not necessarily be
related to job performance. Bagozzi [1980] reported that
job satisfaction did not have a statistically significant
impact on job performance, and job performance had a
statistically significant impact on job performance.
Table 5.10 shows the result of constraining the job
satisfaction—>job performance path. There is a
statistically significant result that indicates that this
path is significant in the theoretical model and should
not be constrained. The null hypothesis that there is no
significant difference between the two nested structural
models is rejected. These results differ from those
reported by Bagozzi [1980]. An explanation for this could
be the difficulty in explaining causality between these
two constructs (for example, Lawler and Porter, 1967;
Porter and Lawler, 1968; Lawler, 1971; Ross and Bomeli,
1971; Benke and Rhode, 1980).
The next SCDT is between the theoretical and
unconstrained model. As with the constrained model,
parameters that are freed in the unconstrained model
should be done in a theoretical perspective. The
parameter that is freed in the unconstrained model is the
sequence job satisfaction—>organizational commitment. As
in this study, other studies have used organizational
commitment as an antecedent to job satisfaction (for
112
TABLE 5.10 COMPARISON OF THE THEORETICAL AND CONSTRAINED MODELS
JOB SATISFACTION—>JOB PERFORMANCE CONSTRAINED
Chi Significance Model Square Difference D.F Level
Constrained 1008.85 94 1
Theoretical 447.94 560.91 93 < .001
1 Difference in degrees of freedom, 1, with difference in
the chi-scjuare value, 560.91.
113
example, Aranya, Lachman, and Amernic, 1982; Hunt, Chonko,
and Wood, 1985; Harrell, Chewning, and Taylor, 1986).
Still other studies have used job satisfaction as an
antecedent to organizational commitment (for example.
Steers, 1977; Zahra, 1984). In these latter studies, job
satisfaction was classified as an attitudinal variable
that precedes organizational commitment.
Table 5.11 presents the results of the SCDT between
the theoretical and unconstrained models. Freeing the
parameter between job satisfaction and organizational
commitment did not result in a statistically significant
change in the chi-square value based on the change in the
degrees of freedom. The null hypothesis that there is no
significant difference between the two nested structural
models is not rejected. Since this path is not
significant and does not contribute to the explanation of
the construct covariances, the theoretical model would be
tentatively accepted.
Evaluating the Theoretical Model
Even though the theoretical model has been
tentatively accepted using the decision-tree framework for
SCDTs, further evaluation is needed to justify accepting
the theoretical model. Anderson and Gerbing [1988]
recommend practical in addition to statistical
considerations. Also, relationships that were
114
TABLE 5.11 COMPARISON OF THE THEORETICAL AND UNCONSTRAINED MODELS
Job performance may be increasing based on a learning
curve, or some other factors. While staff accountants may
perceive themselves as performing better over time on the
assigned jobs, overall they may not be satisfied with
their jobs. There could be two reasons for these results.
First, it is possible that the same effects discussed
above on job performance and organizational commitment are
occurring in this relationship. While performance is
increasing over time, so is the likelihood of migration
from the firm. Another explanation is that the tasks may
not be as challenging as a staff accountant would like,
and the repetitive nature of the tasks could lead to job
dissatisfaction. Norris and Niebuhr [1983] reported that
job satisfaction decreases for accountants early in their
careers due to the work, self, and home.
Research Objective Three
The third research objective was to investigate the
relationship among organizational commitment, professional
commitment, and life-span career development for staff
accountants in public accounting firms. Hypothesis twelve
posited a positive significant relation between life-span
career development and organizational commitment. This
hypothesis was rejected and discussed in Chapter V.
"Hypothesis ten stated that a staff accountant's
organizational commitment will be directly predicted by
132
professional commitment. This hypothesis was not
rejected, supporting previous empirical studies in
accounting research (for example, Aranya, Lachman, and
Amernic, 1982; Lachman and Aranya, 1986).
Hypothesis eleven posited a significant positive
relation between professional commitment and life-span
career development. This hypothesis was not rejected as
evidenced by the statistically significant covariance
among these two exogenous concepts in matrix phi. This
covariance "parallels the correlation between predictor
variables in multiple regression" [Hayduk, 1987, p. 94].
This suggests that as staff accountants progress further
through the implementation stage of their career
development, their level of commitment to the profession
should also increase.
The use of structural equation analysis to examine
the hypothesized relationships cimong the constructs is a
positive characteristic of this study. A two-step
approach was used, which tests the measurement model prior
to testing the structural submodels. This approach is
considered superior to the traditional approach of testing
individual paths using structural equations [Anderson and
Gerbing, 1988].
The results of the study should also be considered in
light of its limitations and contributions. The next
133
section will discuss these as they apply to the research
design and results of the data analysis.
Contributions
Caplan [1989] discussed the need to examine the
behavior of individuals in public accounting firms. One
of the contributions of this study is the addition to this
body of literature by investigating the relationships of
professional commitment, life-span career development,
self-monitoring, and organizational commitment with job
satisfaction and job performance among staff accountants
in public accounting firms.
Job satisfaction and job performance are constructs
of interest in empirical studies among numerous
disciplines (see footnote 1, Chapter II). The constructs
self-monitoring and life-span career development have not
been used in published empirical studies using job
satisfaction and job performance as endogenous constructs.
Also, a research instrument was developed to measure the
implementation stage of an individual's career
development.
This study used only staff accountants as subjects,
which provides information on a large segment of
professional accountants. For public accounting firms,
this-study provides information that may be useful in
personnel decisions. The importance of maximizing job
134
satisfaction and job performance among employees within
organizations is continually sought by managers.
Maximizing job satisfaction and job performance may lead
to higher productivity, lower overall costs for the firm,
and reduced turnover of employees. While attrition may be
necessary in the public accounting profession due to the
limited number of supervisory and managerial positions,
early detection can diminish the premature or unexpected
attrition of employees whom the firm would like to retain.
In addition, better understanding of the relationship
between job satisfaction and job performance may lead to
minimizing expenditures to increase job performance, job
satisfaction, and costs associated with employee turnover.
Limitations
The subjects were all employees of one office of a
national public accounting firm. As noted earlier, all of
the subjects were staff accountants. This limits the
generalizability of the results to other firms and job
levels.
When nonresponse cannot be eliminated, there is the
possibility of nonresponse bias. Attempts were made to
maximize the response rate by including a stamped self-
addressed envelope, and a letter from a partner in the
firm encouraging participation, with each research
instrument. A follow-up letter was sent to all subjects
135
from a partner in the firm encouraging them to complete
and mail back the research instrument if they had not
already done so. Tests for nonresponse bias were
performed to see if there was a difference between early
and late respondents. While these tests did not indicate
nonresponse bias, it is always possible that nonresponse
bias is present. The overall response rate for this study
was 74%.
Lack of precision in measurement is a possibility in
survey research. With the exception of life-span career
development, the survey questions had been used in
previous empirical studies. While this does not guarantee
reliability and validity, it provides support for the use
of these questions in this study. The questions for life
span career development were created for this study.
Pretesting was undertaken to support the use of these
questions. Attempts were made to respecify the
measurement model to increase the reliability of the
research instrument. Even with due care, the possibility
of measurement error exists when doing survey research.
Future Research
This study can lead to additional research in the
area of job satisfaction and job performance among
accountants. Comparisons can be made between staff
accountants, managers, and partners with the constructs
136
used in this study. Further stages of life-span career
development can be investigated with managers and partners
as subjects. Additional constructs could be used to
investigate job satisfaction and job performance.
The study can be expanded by surveying staff
accountants, managers, and partners employed in public
accounting firms nationwide. This would increase the
generalizability of the results.
Accountants not employed in public accounting could
be used as subjects, and comparisons can be made between
groups. This would include accountants employed in
private industry, not-for-profit organizations, and
education.
Finally, a study of this type could be used to
investigate how the constructs relate to job turnover.
This would entail a longitudinal study, in which a model
predicated on job satisfaction and job performance would
be developed to attempt to predict job turnover.
137
REFERENCES
Ajzen, I., C. Timko, and J. White. "Self-monitoring and the Attitude-Behavior Relation." Journal of Personality and Social Psychology (1982): 426-435.
Alderfer, C. P. Existence, Relatedness, and Growth (The Free Press, 1972).
Anastasi, A. Psychological Testing (Macmillan Publishers, 1976).
Anderson, J. and D. Gerbing. "Structural Equation Modeling in Practice: A Review and Recommended Two-Step Approach." Psychological Bulletin (1988): 411-423.
Aranya, N., A. Barak, and J. Amernic. "A Test of Holland's Theory in a Population of Accountants." Journal of Vocational Behavior (1981b): 15-24.
and K. Ferris. "A Reexamination of Accountants' Organizational-Professional Conflict." The Accounting Review (January 1984): 1-15.
, N., R. Lachman, and J. Amernic. "Accountants' Job Satisfaction: A Path Analysis." Accounting, Organizations and Society (1982): 201-215.
, N., J. Pollock, and J. Amernic. "An Examination of Professional Commitment in Public Accounting." Accounting, Organizations and Society (1981a): 271-280.
Armstrong, J. and T. Overton. "Estimating Nonresponse Bias in Mail Surveys." Journal of Marketing Research (August, 1977): 396-402.
Arrington, C , C. Bailey, and W. Hopwood. "An Attribution Analysis of Responsibility Assessment for Audit Performance." Journal of Accounting Research (Spring 1985): 1-20.
Bao, B., D. Bao, and M. Vasarhelyi. "A Stochastic Model of Professional Accountant Turnover." Accounting, Organizations and Society (1986): 289-296.
Bagozzi, R. "Performance and Satisfaction In An Industrial Sales Force: An Examination of Their Antecedents and Simultaneity." Journal of Marketing (1980a): 65-77.
138
Bagozzi, R. Causal Models in Marketing (John Wiley & Sons Inc., 1980b).
"A Prospectus for Theory Construction in Marketing." Journal of Marketing (1984): 11-29.
and Y. Yi. "On The Evaluation of Structural Equation Models." Journal of the Academy of Marketing Science (1988): 74-94.
Bartol, K. "Professionalism as a Predictor of Organizational Commitment, Role Stress, and Turnover: A Multidimensional Approach." Academy of Management Journal (1979): 815-821.
Bary, J. and S. Maxwell. Multivariate Analysis of Variance (Sage Publications, Inc., 1990).
Benke, R. Jr. and J. Rhode. "The Job Satisfaction of Higher Level Employees in Large Certified Public Accounting Firms." Accounting, Organizations and Society (1980): 187-201.
Berry, W. Nonrecursive Causal Models (Sage Publications, Inc., 1989).
Bline, D., D. Duchon, and W. Meixner. "The Measurement of Organizational and Professional Commitment: An Examination of the Psychometric Properties of Two Commonly Used Instruments." Behavioral Research in Accounting (1991): 1-12.
Bohrnstedt, G. "Measurement." In Rossi, P. et al. , eds. , Handbook of Survey Research (Academic Press, 1983).
Bollen, K. Structural Eguations with Latent Variables (John Wiley & Sons, Inc., 1989).
Brenner, V., C. Carmack, and M. Weinstein. "An Empirical Test of the Motivation-Hygiene Theory." Journal of Accounting Research (Autumn 1971): 359-366.
Briggs, S., J. Cheek, and A. Buss. "An Analysis of the Self-monitoring Scale." Journal of Personality and Social Psychology (1980): 679-686.
Brownell, P. "Leadership Style, Budgetary Participation and Managerial Behavior." Accounting, Organizations and Society (1983): 307-321.
139
Brownell, P. and and M. Mclnnes. "Budgetary Participation, Motivation, and Managerial Performance." The Accounting Review (October 1986): 587-600.
Bullen, M. and E. Flamholtz. "A Theoretical and Empirical Investigation of Job Satisfaction and Intended Turnover in the Large CPA Firm." Accounting, Organizations and Society (1985): 287-302.
Caplan, E. "Behavioral Accounting—A Personal View." Behavioral Research In Accounting (1989): 109-123.
Chenhall, R. and P. Brownell. "The Effect of Participative Budgeting on Job Satisfaction and Performance: Role Ambiguity as an Intervening Variable." Accounting, Organizations and Society (1988): 225-233.
Choo, F. "Job Stress, Job Performance, and Auditor Personality Characteristics." Auditing: A Journal of Practice & Theory (Spring 1986): 17-34.
Chow, C. "The Effects of Job Standard Tightness and Compensation on Performance: An Exploration of Linkages." The Accounting Review (October 1983): 667-685.
Collins, F., P. Munter, and D. Finn. "The Budgeting Games People Play." The Accounting Review (January 1987): 29-49.
Converse, J. and S. Presser. Survey Questions: Handcrafting the Standardized Questionnaire (Sage Publications, Inc., 1986).
Cook, D. "The Effect of Frequency of Feedback on Attitudes and Performance." Journal of Accounting Research (Selected Studies 1967): 213-224.
De Coster, D. and J. Rhode. "Analysis of Certified Public Accounting Subgroup and Accounting Student Personality Traits Using The California Psychological Inventory." Journal of Vocational Behavior (1972): 155-162.
Dillard, J. "Applicability of an Occupational Goal-Expectancy Model in Professional Accounting Organizations." Decision Sciences (1979): 161-176.
Dillon, W. and M. Goldstein. Multivariate Analysis: Methods and Applications (John Wiley & Sons, 1984).
140
Douglas, W. "Scripts and Self-monitoring When Does Being a High Self-monitor Really Make a Difference?" Human Communication Research (1983): 81-96.
Earnest, K. and J.C. Lampe. "An Expectancy Theory Investigation Into Causes of Employee Turnover In Public Accounting." Advances in Accounting (1987): 215-238.
Ferris, K. "Perceived Uncertainty and Job Satisfaction in the Accounting Environment." Accounting, Organizations and Society (1977a): 23-28.
"A Test of the Expectancy Theory of Motivation in an Accounting Environment." The Accounting Review (July 1977b): 605-615.
"Organizational Commitment and Performance in a Professional Accounting Firm." Accounting, Organizations and Society (1981): 317-325.
"Perceived Environmental Uncertainty, Organizational Adaptation, and Employee Performance: A Longitudinal Study in Professional Accounting Firms." Accounting, Organizations and Society (1982): 13-25.
and D. Larcker. "Explanatory Variables Of Auditor Performance In A Large Public Accounting Firm." Accounting, Organizations and Society (1983): 1-11.
Fisher, F. The Identification Problem in Econometrics (McGraw-Hill, 1966).
Freidson, E. The Professions and Their Prospects (Sage Publications, 1973).
Gaski, J. "Interrelations Among a Channel Entity's Power Sources: Impact on the Exercise of Reward and Coercion on Expert, Referent, and Legitimate Power Sources." Journal of Marketing Research (February, 1986): 62-77.
"Commentary on Howell's Observations." Journal of Marketing Research (February, 1987): 127-129.
Gruneberg, M. (Ed.). Job Satisfaction-A Reader (John Wiley and Sons, 1976).
Harrell, A., E. Chewing, and M. Taylor. "Organizational-Professional Conflict and the Job Satisfaction and Turnover Intentions of Internal Auditors." Auditing; A Journal of Practice & Theory (Spring 1986): 109-121.
141
Harrell, A. and R. Eickhoff. "Auditors' Influence-Orientation and Their Affective Responses to The "Big Eight" Work Environment." Auditing: A Journal of Practice & Theory (Spring 1988): 105-118.
and M. Stahl. "McLelland's Trichomoty of Needs Theory and The Job Satisfaction and Work Performance of CPA Firm Professionals." Accounting, Organizations and Society (1984): 241-252.
Hayduk, L. Structural Equation Modeling with LISREL (The John Hopkins University Press, 1987).
Heneman, H., III. "Comparisons of Self- and Superior Ratings of Managerial Performance." Journal of Applied Psychology (1974): 638-642
Herzberg, F., F. Mausner, and B. Snyderman. The Motivation to Work (John Wiley & Sons, Inc., 1959).
Holland, J. L. Making Vocational Choices: A Theory of Careers (Prentice-Hall, Inc., 1973).
Howell, R. "Covariance Structure Modeling and Measurement Issues: A Note on Interrelations Among a Channel Entity's Power Sources". Journal of Marketing Research (February, 1987): 119-126.
Hunt, S., L. Chonko, and V. Wood. "Organizational Commitment and Marketing." Journal of Marketing (Winter, 1985): 112-126.
, R. Sparkman Jr., and J. Wilcox. "The Pretest in Survey Research: Issues and Preliminary Findings." Journal of Marketing Research (1982): 269-273.
Huizinga, G. Maslow's Need Hierarchy in the Work Situation (Welters, 1970).
Ickes, W. and R. Barnes. "The Role of Sex and Self-monitoring in Unstructured Dyadic Interactions." Journal of Personality and Social Psychology (1977): 315-330.
James, L., S. Mulaik, and J. Brett. Causal Analysis: Assumptions, Models, and Data (Sage Publications, Inc., 1987) .
142
Jiambalvo, J. "Performance Evaluation and Directed Job Effort: Model Development and Analysis in a CPA Firm Setting." Journal of Accounting Research (Autumn 1979): 436-455.
"Measures of Accuracy and Congruence in the Performance Evaluation of CPA Personnel: Replication and Extensions." Journal of Accounting Research (Spring 1982): 152-161.
and J. Pratt. "Task Complexity and Leadership Effectiveness in CPA Firms." The Accounting Review (October 1982): 734-750.
, D. Watson, and J. Baumler. "An Examination of Performance Evaluation Decisions in CPA Firm Subunits." Accounting, Organizations and Society (1983): 13-29.
Joreskog K. and D. Sorbom. PRELIS (Scientific Software, 1988).
. LISREL VII (Scientific Software, 1989a).
LISREL VII: A Guide to the Program and Applications (SPSS Inc., 1989b).
Kerlinger, F. Foundations of Behavioral Research (Holt, Rinehart and Winston, 1986).
Kim, J. and C. Mueller. Factor Analysis: Statistical Methods and Practical Issues (Sage Publications, Inc., 1986) .
and C. Mueller. Introduction to Factor Analysis: What It Is and How To Do It (Sage Publications, Inc., 1986) .
Kulik, J. and S. Taylor. "Self-monitoring and the Use of Consensus Information." Journal of Personality (1981): 75-84.
Lachman, R. and N. Aranya. "Evaluation of Alternative Models of Commitments and Job Attitudes of Professionals." Journal of Occupational Behavior (1986): 227-243.
Lampe, J. C. and K. Earnest. "How Motivation Affects Accountants' Productivity and Turnover." Management Accounting (1984): 50-55.
143
Landy, F. and J. Farr. The Measurement of Performance (Academic Press, Inc., 1983).
, S. Zedeck, and J. Cleveland (Eds.). Performance, Measurement and Theory (Lawrence Erlbaum Publishers, 1983).
Lawler, E., III. "Ability as a Moderator of the Relationship between Job Attitudes and Job Performance." Personnel Psychology (1966): 153-164.
"A Correlational-Causal Analysis of The Relationship Between Expectancy Attitudes and Job Performance." Journal of Applied Psychology (1968): 462-468.
Pay and Organizational Effectiveness: A Psychological View (McGraw-Hill Book Co., 1971).
, W. Kuleck, Jr., J. Rhode, and J. Sorensen. "Job Choice and Post Decision Dissonance." Organizational Behavior and Human Performance (1975): 133-145.
and L. Porter. "The Effect of Performance on Job Satisfaction." Industrial Relations (1967): 20-28.
Lengermann, J. "Supposed and Actual Differences in Professional Autonomy Among CPAs as Related to Type of Work Organization and Size of Firm." The Accounting Review (October 1971): 665-675.
Locke, E. "The Nature and Cause of Job Satisfaction." In Dunnette, M. ed.. Handbook of Industrial Psychology (John Wiley and Sons, 1983).
Maslow, A. Motivation and Personality (Harper & Row, Inc., 1954).
McClelland, D. C. The Achieving Society (Van Nostrand, 1961).
Mossholder, K., A. Bedeian, J. Touliatos, and A. Barkman. "An Examination of Interoccupational Differences: Personality, Perceived Work Climate, and Outcome Preferences." Journal of Vocational Behavior (1985): 164-176.
Mowday, R. , R- Steers, and L. Porter. "The Measurement of Organizational Commitment." Journal of Vocational Behavior (1979): 224-247.
144
Norris, D. and R. Niebuhr. "Professionalism, Organizational Commitment and Job Satisfaction in an Accounting Organization." Accounting, Organizations and Society (1983): 49-59.
Nunnally, J. Psychometric Theory, Second Edition (McGraw-Hill Book Company, 1978).
Osipow, S. Theories of Career Development, Third Edition (Prentice-Hall, Inc., 1983).
Philips, J. and S. Freedman. "Situational Performance Constraints and Task Characteristics: Their Relationship to Motivation and Satisfaction." Journal of Management (1984): 321-331.
Porter, L. and E. Lawler III. Managerial Attitudes and Performance (Richard D. Irwin, Inc. and The Dorsey-Press, 1968).
, R. Steers, R. Mowday, and R. Boulian. "Organizational Commitment, Job Satisfaction and Turnover Among Psychiatric Technicians." Journal of Applied Psychology (1974): 603-609.
Price, J. and C. Mueller. Handbook of Organizational Commitment (Pitman Publishing Inc., 1986).
Prien, E. and R. Liske. "Assessments of Higher-Level Personnel Rating Criteria: A Comparative Analysis of Supervisor Ratings and Incumbent Self-Ratings of Job Performance." Personnel Psychology (1962): 187-194.
Rasch, R. and A. Harrell. "The Impact of Personal Characteristics on the Turnover Behavior of Accounting Professionals." Auditing; A Journal of Practice & Theory (1990): 90-102.
Ramanatham, K., R. Peterson, and M. Maher. "Strategic Goals and Performance Criteria in CPA Firms." The Journal of Accountancy (January 1976): 56-64.
Rasch, R. and A. Harrell. "The Impact of Personal Characteristics on the Turnover Behavior of Accounting Professionals." Auditing: A Journal of Practice & Theory (Spring, 1990): 90-102.
Reed, S. and S. Kratchman. "Job Satisfaction: Differences Throughout the Accounting Profession." The Ohio CPA Journal (Autumn 1987): 33-47.
145
Rhode, J., J. Sorensen, and E. Lawler. "An Analysis of Personal Characteristics Related to Professional Staff Turnover in Public accounting Firms." Decision Sciences (October 1976): 771-797.
Ross, T. and E. Bomeli. "A Comment on Accountants' Job Satisfaction. " Journal of Accounting Research (Autumn 1971): 383-388.
Seller, R. and R. Sapp. "Just How Satisfied Are Accountants With Their Jobs?" Management Accounting (March 1979): 18-21.
Senatra, P. "Role Conflict, Role Ambiguity, and Organizational Climate in a Public Accounting Firm." The Accounting Review (October 1980): 594-603.
Sims, H. Jr. and A. Szilagyi. "Leader Reward Behavior and Subordinate Satisfaction and Performance." Organizational Behavior and Human Performance (1975): 426-439.
Snyder, M. "Self-monitoring of Expressive Behavior." Journal of Personality and Social Psychology (1974): 526-537.
Advances in Experimental Social Psychology (Academic Press, 1979).
"The Influence of Individuals on Situations: Implications for Understanding the Links Between Personality and Social Behavior." Journal of Personality (1983): 497-516.
, E. Berscheid, and P. Click. "Focusing on the Exterior and Interior: Two Investigations of the Initiation of Personal Relationships." Journal of Personality and Social Psychology (1985): 1427-1439.
and N. Cantor. "Thinking About Ourselves and Others: Self-monitoring and Social Knowledge." Journal of Personality and Social Psychology (1980): 222-234.
and S. Gangestad. "Choosing Social Situations: Two Investigations of Self-monitoring Processes." Journal of Personality and Social Psychology (1982); 123-135.
146
Snyder, M., S. Gangestad, and J. Simpson. "Choosing Friends as Activity Partners: The Role of Self-monitoring." Journal of Personality and Social Psychology (1983): 961-972.
and T. Monson. "Persons, Situations, and the Control of Social Behavior." Journal of Personality and Social Psychology (1975): 637-644.
and J. Simpson. "Self-monitoring and Dating Relationships." Journal of Personality and Social Psychology (1984): 1281-1291.
and W. Swann Jr. "When Actions Reflect Attitudes: The Politics of Impression Management." Journal of Personality and Social Psychology (1976): 1034-1042.
and E. Tanke. "Behavior and Attitude: Some People are More Consistent Than Others." Journal of Personality (1976): 510-517.
, E. Tanke, and E. Berscheid. "Social Perception and Interpersonal Behavior: On the Self-fulfilling Nature of Social Stereotypes." Journal of Personality and Social Psychology (1977): 656-666.
Sorensen, J. "Professional and Organizational Profiles of the Migrating and Non-migrating Large Public Accounting Firm CPA." Decision Sciences (1970): 489-512.
and D. Franks. "The Relative Contribution of Ability, Self-Esteem and Evaluative Feedback to Performance: Implications for Accounting Systems." The Accounting Review (October 1972): 735-746.
and T. Sorensen. "The Conflicts of Professionals in Bureaucratic Organizations." Administrative Science Quarterly (March 1974): 98-106.
Steers, R. "Antecedents and Outcomes of Organizational Commitment." Administrative Science Quarterly (March 1977): 46-56.
Strawser, R. , J. Ivancevich, and H. Lyon. "A Note on the Job Satisfaction of Accountants in Large and Small CPA Firms." Journal of Accounting Research (Autumn 1969): 339-345.
147
Super, D. E. "Vocational Development in Adolescence and Early Adulthood: Tasks and Behavior." In Super, D. E. et al.. Career Development: Self-Concept Theory (CEEB Research Monograph No. 4, 1963).
The Psychology of Careers (Harper and Row, 1957).
and E. Knasel. "Career Development in Adulthood: Some Theoretical Problems and a Possible Solution." British Journal of Guidance and Counseling (July 1981): 194-201.
Sutton, S. G. and J. C. Lampe. "Monitoring Quality in a Knowledge-Work Environment: Integrating Structured Group Processes with Decision Support System Design." Productivity Management Frontiers (Vol. II 1988): 365-372.
Teas, K. "Supervisory Behavior, Role Stress, and the Job Satisfaction of Industrial Salespeople." Journal of Marketing Research (February 1983): 84-91.
Teel, K. "Self-Appraisal Revisited." Personnel Journal (July 1978) : 364-367.
Thornton, G., III. "Psychometric Properties of Self-Appraisals of Job Performance." Personnel Psychology (1980): 263-271.
Todd, J., P. Thompson, and G. Dalton. "Management Control of Personnel." The Journal of Accountancy (February 1974): 34-40.
Tracy, L. and T. Johnson. "What do the Role Conflict and Role Ambiguity Scales Measure?" Journal of Applied Psychology (October 1981): 464-469.
Turner, R. "Self-monitoring and Humor Production." Journal of Personality (1980): 163-172.
Vroom, V. Work and Motivation (John Wiley & Sons, Inc., 1964) .
Wahba M. and L. Bridwell. "Maslow Reconsidered: A Review of Research on the Need Hierarchy Theory." Organizational Behavior and Human Performance (April 1976): 212-237.
148
Weiss, D., R. Dawis, G. England, and L. Lofquist. Manual for the Minnesota Satisfaction Questionnaire (University of Minnesota, 1967).
Wright, A. "Performance Appraisal of Staff Auditors." The CPA Journal (November 1980): 37-43.
"An Investigation of the Engagement Evaluation Process for Staff Auditors." Journal of Accounting Research (Spring 1982): 227-239.
Zahra, S. "Antecedents and Consequences of Organizational Commitment: An Integrative Approach." Akron Business and Economic Review (Fall 1984): 26-32.
Zytowski D. and R. Hay. "Do Birds of a Feather Flock Together? A Test of the Similarities Within and the Differences Between Five Occupations." Journal of Vocational Behavior (1984): 242-248.
149
APPENDIX A
RESEARCH INSTRUMENT
150
FOR YOOa ACTIOH Memorandum
0 All Persoaael DUtribudoQ
Limiud DLstribudon: Partners Maiugers Audit • f TAX • • .MC • f Admin. 0
SuiT t •
•
0
Secrcttries 0 0 0
° TO: All Professional Personnel
FROM:
DISSERTATION QUESTIONNAIRB
Attached is a questionnaire which all office professionals are being asked to complete in order to assist a doctoral student with his dissertation. A high response rate is critical to the success of the dissertatloo. The Firm has committed Its assistance to this project and therefore, I urge each of you to take the time to complete this questionnaire. Thank you for your cooperation.
151
Peter John Poznanski, CPA 3219 Chadwick Drive Rockford, II. 61109
Certified Public Accountants
Dear
Attached is a copy of a questionnaire I will be using in my doctoral dissertation. Your assistance in answering the questions will facilitate my completing the dissertation required for my degree, and contribute to the body of knowledge in accounting.
All of your responses will be kept in the strictest of confidence, known only to myself. In no way will this questionnaire identify you, nor will any attempt be made to identify you by your responses. The first page requesting the demographic information will be detached from the questionnaire when I receive it, and the questionnaires and demographic data sheet will be destroyed as soon as the data is entered into the computer. You can return the completed questionnaire in the attached self-addressed st£unped envelope.
One of the more frustrating aspects of answering a questionnaire of this nature is not knowing the purpose behind the questions. The reason for this is that unconsciously you may bias your responses towards the answers you feel are appropriate rather than your true feelings. While I believe this point is debatable, it is a requirement imposed upon roe for this study. However, the time you take to answer the questions as best you can will greatly enhance the results of this study. This study will culminate four years of my graduate school work, so your responses are very important and very necessary to me.
The data analysis and the dissertation should be completed early next year. I will be sending
a copy of the completed study. If you are interested in receiving a copy of the results please retain my address and contact roe. I will be happy to share a copy of the results with you. Thank you for taking the time to help roe complete ray schooling and study.
Sincerely,
Peter John Poznanski, CPA
152
Demographic Data
1. Age:
2. Sex:
3
4
5
6
7
Number of years in the public accounting profession: __
Years at current Firm:
Position at current Firm (Please circle the appropriate response): ^^ ^^^ixacc
Partner or Principal
Senior Manager
Manager
Senior Consultant
Supervising Senior, Audit
Supervising Senior, Tax
Senior Accountant, Audit
Senior Specialist, Tax
Consultant
Staff Accountant
Tax Specialist
Assistant Accountant
Assistant Tax Specialist
Intern
Preprofessional
Staff Technician
Paraprofessional
Administrative
Other (please specify):
Number of years in current Position:
College Degree(s) and Major and Minor course work for each degree:
8. Are you currently enrolled in a degree program, emd if so which degree are you working on?
9. Professional Certification(s) and the number of years with the Certification (e.g., CPA, two years):
153
Please c i r c l e the nmnber-frorn one to five--which best represents your gersonal fee l ings towards the accounting
Strongly Disagree
1 Disagree
2 Neutral
3 Agree 4
Strongly Agree 5
I am willing to put in a great deal of effort beyond that normally expected m order to help make my profession successful
I talk up this profession with my friends as a great profession to be associated with
I feel very little loyalty to this profession
I would accept almost any type of job assignment in order to keep working in areas that are associated with this profession
I find that my values and the profession's values are very similar...
I am proud to tell others that I am part of this profession
I could just as well be associated with another profession as long as the type of organization in which I worked were similar
Being a member of this profession really inspires the very best in roe in the way of job performance
It would take very little change in my present circumstances to cause me to work in areas that are not associated with this profession
I am extremely glad that I chose this profession over others I was considering at the time I joined
2
2
2
2
2
4
4
4
4
5
5
5
5
5
154
Strongly Disagree
1 Disagree
2 Neutral
3 Agree
4
Strongly Agree
5
There's not too much to be gained by sticking with this profession indefinitely 1
Often, I find it difficult to agree with this profession's policies on important matters relating to its members 1
I really care about the fate of this profession 1
For me this is the best of all professions to be a member of 1
Deciding to be a member of this profession was a definite mistake on my part 1
2
2
2
3
3
3
4
4
4
5
5
5
Please c i r c l e the number--frora one to five—which best represents your personal feelings on your performance of your present job.
Needs Unsatisfactory Improvement Satisfactory Superior Outstanding
1 2 3 4 5
Maintaining quantity of work
Maintaining quality of work
Communicating orally
Communicating in writing
Accepting responsibility and initiating action
Exercising professional skills and care
Following policies and procedures
Planning and organizing work
2
2
2
2
2
2
2
5
5
5
5
5
5
5
155
Needs Unsa t i s f ac to ry Improvement Sa t i s fac tory Superior Outstanding
1 2 3 4 5
Adapting to new or d i f f e r e n t job s i t u a t i o n s 1 2 3 4 5
G e t t i n g along with o the r s in the f i rm. . 1 2 3 4 5
Dealing with c l i e n t s outs ide the f i r m . . 1 2 3 4 5
Supervis ing others 1 2 3 4 5
P l e a s e c i r c l e the number—from one t o five—which best r e p r e s e n t s your personal feel ings towards your accounting c a r e e r .
St rongly Strongly Disagree Disagree Neutral Agree Agree
1 2 3 4 5
My previous education and training have given me adequate preparation for ray current pos i t ion 1 2 3 4 5
I have completed my training for my career 1 2 3 4 5
Ky current job is relevant to my career goals 1 2 3 4 5
I am considering changing careers 1 2 3 4 5
My current career offers job security.. 1 2 3 4 5
I feel that the amount of time that I will be required to stay in ny current position before I am eligible for a promotion is reasonable 1 2 3 4 5
I am satisfied with my career choice to be an accountant 1 2 3 4 5
I am not prepared for my current position 1 2 3 4 5
156
Please circle the number--from one to five--which best represents your personal feelings towards your present job.
Very Dissatisfied
1 Dissatisfied
2 Neutral
3 Satisfied
4
Very Satisfied
Being able to keep busy all the time...
The chance to work alone on the job
The chance to do different things from time to time
The chance to have some prestige in the community because of my job
The way my supervisor handles his/her people
The competence of my supervisor in making decisions
Being able to do things that don't go against my conscience
The way the job provides for steady employment
The chance to do things for other people
The chance to supervise other people...
The chance to do things that make use of my abilities
The way organizational policies are put into practice
My pay and the amount of work I do.
The chances for advancement ,
The freedom to use my own judgment.
The chance to try ray own methods of doing the job
The general working conditions
157
Very ^ Dissa t i s f i ed Dissat i s f ied Neutral Satisfied Sat is f ied
1 2 3 4 5
The way my co-workers get along with each other X 2 3 4
The feedback I get for doing an e f f e c t i v e job 1 2 3 4 s
The feel ing of accomplishment I get from the job ; 1 2 3 4 5
Please circle the number—from one to five—which best 5??h^wh?^h r^^ E^isonal feelings towards the orga^zation with which you are presently employed. Please"nite that
Strongly Disagree
1 Disagree
2 Neutral
3 Agree 4
Strongly Agree 5
I am willing to put in a great deal of effort beyond that normally expected in order to help this organization be successful ,
I talk up this organization with my friends as a great organization to be associated with
I feel very little loyalty to this organization
I would accept almost any ty];>e of job assignment in order to keep lurking for this organization
I find that ny values and the organization's values are very similar.
I am proud to tell others that I an part of this organization
I could just as well be working for a different organization as long as the type of work was similar
4
4
4
4
4
5
5
5
5
5
158
strongly Disagree
1 Disagree
2 Neutral
3 Agree 4
Strongly Agree 5
This organization really inspires the very best in me in the way of job performance 1
It would take very little change in my present circumstances to cause me to leave this organization 1
I am extremely glad that I chose this organization to work for, over others I was considering at the time I joined. 1
There is not too much to be gained by sticking with this organization indefinitely 1
Often, I find it difficult to agree with this organization's policies on important matters relating to its employees 1
I really care about the fate of this organization 1
For me this is the best of all possible organizations for which to work 1
Deciding to work for this organization was a definite mistake on my part 1
3
3
3
3
4
4
4
4
5
5
5
5
The statements on the following pages concern your personal reactions to a number of different situations. No two statements are exactly alike, so consider each statement carefully before answering. If a statement is TRUE or MOSTLY TRUE as applied to you, circle the word True next to the question. If a statement is FALSE or NOT USUALLY TRUE as applied to you, circle the word False next to the question.
I find it hard to imitate the behavior of other people
My behavior is usually an expression of my true inner feelings, attitudes, and beliefs
True
True
False
False
159
At-parties and social gatherings, I do not attempt to do or say things that others will like True False
I can only argue for ideas which I already believe True False
I can make impromptu speeches even on topics about which I have almost no information True False
I guess I put on a show to impress or entertain people True False
When I am uncertain how to act in a social situation, I look to the behavior of others for cues True False
I would probably make a good actor True False
I rarely need the advice of my friends to choose movies, books, or music True False
I sometimes appear to others to be experiencing deeper emotions than I actually am True False
I laugh more when I watch a comedy with others thain when alone True False
In a group of people I am rarely the center of attention True False
In different situations and with different people, I often act like very different persons True False
I aro not particularly good at making other people like roe True False
Even if I aro not enjoying royself, I often pretend to be having a good time. True False
I'm not always the person I appear to be True False
I would not change my opinions (or the way I do things) in order to please someone else or win their favor True False
I have considered being an entertainer. True False
160
In order to get along and be liked, I tend to be what people expect me to be rather than anything else True False
I have never been good at games like charades or improvisational acting True False
I have trouble changing my behavior to suit different people and different situations True False
At a party I let others keep the jokes and stories going True False
I feel a bit aw)cward in company and do not show up quite so well as I should.. True False
I can look anyone in the eye and tell a lie with a straight face (if for a right end) True False
I may deceive people by being friendly when I really dislike them True False
Please circle your responses to the following two questions.
Are you planning to leave the firm at which you are currently employed?
Yes Undecided No
Axe you planning to remain in the public accounting profession?
Yes Undecided No
Thank you very much for your time and assistance.
161
APPENDIX B
RESEARCH INSTRUMENT APPROVAL
162
TEXASXECH 0«f<t ol t rwifch S«r«<ei
8oi «*,'0/U66oci. r«i4t 71*01. IP 35/(BOH N j j
July 10, 1990
Dr. reter John Pozhanski Accounting Campus
Caar OiT. Pozhanski;
The Te.' as Tech University Committee for the Protection of Human Subjec t s has approved your project, "Tne Effects of Orga.'.izat lonal Commitment, Professional Commitment, Life Span Career Oevelopraent, and Self-Monitoring on Job Sat i s fac t icn and Job Performance Among Staff Accountants." The approval v i l l extend for one year from J u l y 31, 1990. You w i l l be reminded of the pending e x p i r a t i o n one month before your approval expires so that you .-nay request an e x t e n s i o n if you wish.
The best of luc)c on your project .
Sincerely,
Jaoes Saith Chairperson, Hunan Subjects Use Coocittee
The responses, except for self-monitoring, were made on a five point Likert-type scale, where higher responses indicate stronger levels of the construct. The values for self-monitoring were either 1 (for low self-monitoring) or 2 (for high self-monitoring).
166
APPENDIX D
LISREL RESULTS FOR INDIVIDUAL
HYPOTHESES AND PARAMETERS
167
TABLE D.l TESTING INDIVIDUAL PHI
Model Chi-Square Difference D.F
Significance Level
Null (baseline) 10693.69 106
Relationship and Hypothesis Tested:
PC<->LSCD/H11 10333.55 360.14 105 .001
Significance level based on 1 degree of freedom difference and the change in the chi-square value.
168
TABLE D.l TESTING INDIVIDUAL PHI
Model Chi-Square Difference D.F
Significance Level
Null (baseline) 10693.69 106
Relationship and Hypothesis Tested:
PC<->LSCD/H11 10333.55 360.14 105 .001
Significance level based on 1 degree of freedom difference and the change in the chi-square value.
168
TABLE D.2 TESTING INDIVIDUAL BETAS—WITH PHI
1 Chi- Significance
Model Square Difference D.F Level
Null (baseline) 10333.55 105
Relationship and Hypothesis Tested:
0C->JS/H1
0C->JP/H2
JP->JS/H9
JS->JP/H9
10109.23
10321.76
9641.96
9824.16
224.32
11.79
691.59
509.39
104
104
104
104
.001
.001
.001
.001
Significance level based on 1 degree of freedom difference and the change in the chi-square value.
169
TABLE D.3 TESTING INDIVIDUAL GAMMAS—WITH PHI
Model Chi-Square Difference D.F
Significance Level
Null (baseline) 10333.55 105
Relationship and Hypothesis Tested:
PC->OC/H10
PC->JS/H3
PC->JP/H4
LSCD->0C/H12
LSCD->JS/H5
LSCD->JP/H6
SM->JS/H7
SM->JP/H8
5899.77
10264.41
10323.41
5949.34
10294.33
10263.84
10290.87
10333.43
4433.78
69.14
10.14
4384.21
39.22
69.71
42.68
.12
104
104
104
104
104
104
104
104
.001
.001
.005
.001
.001
.001
.001
> .10
Significance level based on 1 degree of freedom difference and the change in the chi-square value.
170
TABLE D.4 LISREL VALUES FOR THE HYPOTHESES
1 Significance
Hypothesis Value Level
0C->JS/H1
0C->JP/H2
PC->JS/H3
PC->JP/H4
LSCD->JS/H5
LSCD->JP/H6
SM->JS/H7
SM->JP/H8
JP->JS/H9
JS->JP/H9
PC->OC/H10
PC<->LSCD/H11
LSCD->0C/H12
8.430
-6.253
-2.974
-.901
4.565
-1.008
1.559
-.962
-14.212
37.076
12.024
51.829
-.722
.001
.001
.001
> .10
.001
> .10
> .05
> .10
.001
.001
.001
.001
> .10
The significance of the values are based on a comparison to a normal distribution table.
171
TABLE D.5 1 LISREL VALUES FOR LAMBDA Y AND LAMBDA X
Lambda Y
OCl 0C2 0C3 JSl JS2 JS3 JPl JP2 JP3
Lambda
OC 1 127 112
X
JS
1 63 67
JP
1 76 77
PC LSCD SM PCI 1 PC2 102 PC3 108 LSCDl 1 LSCD2 43 LSCD3 37 SM 1
The value 1 indicates that the starting values was set at 1 to link the specific indicator to the concept [Hayduk, 1987].