PREDICTING JOB SATISFACTION THROUGH BIODATA:
THE DEVELOPMENT OF A CLASSROOM SATISFACTION
Lucy H. Wenzel
B.A., University of Colorado, Boulder, 1986
A thesis submitted to the
Faculty of the Graduate School of the
University of Colorado at Denver
in partial fulfillment
, of the requirements for the degree of
! Master of Arts
This thesis for the Master of Arts
Lucy H. Wenzel
has been approved for the
Wenzel, Lucy H. (M.A., Psychology)
Predicting Job Satisfaction through Biodata: The Development of a
Thesis directed by Professor Kurt Kraiger
The purpose of the present investigation was to develop and
evaluate a biodata questionnaire designed to predict classroom
satisfaction. This was done by conducting a thorough review of the
satisfaction literature, developing the Biodata Questionnaire (BQ), and
collecting empirical evidence on the reliability and construct validity of
The study was conducted in an urban university, using 115
subjects drawn from five undergraduate psychology classes.
Paper-and-pencil questionnaires were administered in class on the
first day of class and again during the 12th week of the semester.
The BQ was developed with 15 conceptually-derived sub-scales.
A scale analysis revealed several of the scales to be highly
intercorrelated; therefore, they were combined to create a single scale.
Other scales were dropped because they had low reliability and no
compelling theoretical rationale for retention. The final BQ scales
measured general satisfaction, internal motivation, expectations,
recognition and quality of instruction.
The fullrscale BQ had adequate internal consistency (alpha = .81)
and test-retest reliability (r = .74). Construct-related evidence of
validity was supported through significant correlations with positive (r
= .56) and negative affectivity (r = -.37) and efficacy (r = .31), as well as
evidence that the BQ significantly predicted end-of-term classroom
satisfaction (R = .39). A hierarchical multiple regression analysis
revealed the BQ to be a better predictor of classroom satisfaction than
expected grade,; a traditional predictor.
The implications of predicting satisfaction were discussed. Using
biodata to predict satisfaction may be helpful in classroom settings, in
organizations to maximize the effects of interventions, and to assist in
the organizational placement process. The findings of the study also
support the dispositional argument of satisfaction by indicating the
temporal and situational consistency of satisfaction and by showing
that different individuals reported different levels of satisfaction in
Suggestions were offered for extending use of the biographical
information to predict satisfaction in an organizational setting.
This abstract accurately represents the content of the candidate's
thesis. I recommend its publication.
This thesis is lovingly dedicated to my family and friends for their
support and patience.
Ultimate thanks go to my advisor, Kurt Kraiger. Kurt's faith in me,
his guidance and his ability to send just the right e-mail at just the right
time, made him more than a mentor, it made him a friend.
Special thanks to Chris Hornick of my advisory committee for
going far above and beyond the call of duty in lending his statistical
expertise to my analyses. Without Chris' brilliant statistical insight,
quick wit and uncanny ability to ferret out a sunny deck with a great
margarita, this project would never have been completed.
Thanks also to Jim Nimmer of my advisory committee. Jim's
attention to detail and support throughout my thesis and my time at
UCD is greatly appreciated.
My love and .gratitude to George Kern for his amazing patience
with late night print-out runs, with constant modem connections and
disconnections, and for listening to more than he ever wanted to know
about everything within these pages.
The impact you've all made on my work and in my life will be with
me forever and for that I am most grateful.
I. INTRODUCTION AND LITERATURE REVIEW .
Literature Review .
Job Satisfaction ....
Data Collection ....
College Student Satisfaction Questionnaire
Positive and Negative Affect Schedule .
Expected Course Grade .
Data Analyses .....
vi i i
Descriptive Statistics . . . .37
Test of Hypotheses . .39
IV. DISCUSSION, IMPLICATIONS, LIMITATIONS,
Discussion . . .44
Summary of Results .44
Implications . . . . .47
Biodata as a Predictor of Classroom
Satisfaction . . .47
Biodata as a Predictor of Job Satisfaction . 49
Predicting Satisfaction and the
Dispositional Argument . .51
Limitations . . .53
Validity of the Measures .54
Confounding Variables .55
Future Directions .56
A. Biodata Questionnaire Sub-Scales .60
B. Biodata Questionnaire
C. College Student Satisfaction Questionnaire
D. Positive and Negative Affect Schedule
E. Efficacy Measure ....
INTRODUCTION AND LITERATURE REVIEW
The underlying causes of the elusive construct of job satisfaction
have been the focus of thousands of articles and studies
(Cropanzano & James, 1990; Gerhart, 1990; Judge & Hulin, 1991;
Locke, 1976; Staw, Bell, & Clausen, 1986). Job satisfaction is
defined as the tendency towards enjoyment, gratification or
contentment in one's work. Industry has been interested in securing
the secret to satisfied workers because individual well being has
been linked to critical organizational interests such as turnover,
absenteeism and morale (Alster, 1989; George, 1989).
In the past, research has been directed chiefly toward task-based
and situational causes of satisfaction. This emphasis placed attention
on workplace hygiene and job design and redesign as the answer to
employee dissatisfaction (Hackman & Oldham, 1976; Herzberg,
Mausner, Peterson, & Capwell, 1957; Salancik & Pfeffer, 1977).
Consequently, organizations experiencing low morale and
dissatisfaction among employees have attempted to change the
workplace (i.e. improvements in the cafeteria, work areas, restrooms,
etc.); redesign the job so as to provide more autonomy in the task
performed; or redesign compensation and benefits programs
(Hackman & Oldham, 1976). Although billions of dollars have been
poured into such projects, the results have often been disappointing
(Bycio, Alvares, & Hackett, 1990; Staw & Ross, 1985). As a result,
researchers have begun to consider individually-based determinants
of job satisfaction. More specifically, theorists have postulated
dispositional influences on job satisfaction (Arvey, Bouchard, Segal &
Abraham, 1989; Staw et al., 1986; Staw & Ross, 1985; Cropanzano &
James, 1990; George, 1989; Judge & Hulin, 1991; Tait, Padgett &
Baldwin, 1989). Their work suggests that people carry a
predisposition toward being satisfied or dissatisfied, regardless of
their environment (Arvey et al., 1989; Cropanzano, James &
Konovsky, 1991; Staw et al., 1986). If true, this claim suggests that
employers may be better served to consider disposition along with
situational demands when attempting to affect the satisfaction of
employees (Bouchard, Arvey, Keller & Segal, 1992; Cropanzano et
Although recent research implicates the role of dispositions in job
satisfaction, no research has yet attempted to use this construct to
predict satisfaction. If negative and positive affectivity, assumably
relatively long-term in nature (Cropanzano et al., 1991; George,
1989), do affect job satisfaction, then the degree of the satisfaction
one is likely to experience at work should be predictable, based on
prior satisfaction (Mumford & Owens, 1982). The key lies in
accurately measuring previously experienced or biographical
satisfaction.. The biodata questionnaire is a valid measure of
biographical or background information (Mumford & Owens, 1982).
Biodata has been found to be a valid predictor of future behavior
and attitudes through its ability to assess past behaviors and attitudes
(Hammer & Kleiman, 1988). Because level of satisfaction seems to
have temporal stability (Staw & Ross, 1985), past satisfaction should
predict future satisfaction, making the biodata method a valid
Although the predictability of job satisfaction has been suggested
(Arvey et al., 1989; Buss & Craik, 1983; Staw & Ross, 1985), there has
been no research or known theories suggesting the use of biodata for
predicting satisfaction. The present research tests the efficacy of the
technique by attempting to predict classroom satisfaction.
To accomplish this goal, several steps were taken. First, a
thorough review of the job satisfaction, classroom satisfaction, and
biodata literature was conducted.
Second, the Biodata Questionnaire (BQ) was developed, guided
by information gathered in the review. The methodology for
developing the questionnaire included the synthesis of information
from both job satisfaction and classroom satisfaction literature for an
indication of the parameters of the construct and therefore the type of
items that needed to be included. Items were formulated and
grouped into sub-scales.
Finally, empirical evidence of the reliability and construct validity of
the biodata scale was collected. It is hoped that the results of this
research will produce a biodata scale that can be adapted for
organizations and used in future research efforts.
The importance of finding satisfaction in one's profession has
become paramount to the American workforce (Hall, 1986; Offerman
& Gowing, 1990). Baby-boomers, born between 1946 and 1964,
make up the largest portion of today's workforce (Hall & Richter,
1990). In contrast to previous generations, these workers place a far
greater emphasis on satisfaction, autonomy and flexibility (Alster,
1989), further increasing the need for understanding the determinants
of the job satisfaction construct.
Not only has intrinsically satisfying, stimulating work become more
important to the American worker and to the American employer
(Griffin, 1988), research has also found job satisfaction to be linked to
individual well being (Tait et al., 1989). Both physical health and
mental well being have been found to be correlated to job satisfaction
(Spector & Jex, 1991; Tait, Padgett & Baldwin, 1989). The well being
of the individual has, in turn, been linked to important organizational
outcomes such as turnover, absenteeism and morale (Locke, 1976).
Job satisfaction has been found to be significantly correlated with
absenteeism, performance and turnover (George, 1989; Spector &
Jex, 1991) as wpll as affective organizational commitment
(Cropanzano et al., 1991). Finally, organizational involvement and
intrinsic motivation have been found to be affected by job satisfaction
(Hall & Richter, 1990; Hoiberg & Pugh, 1978).
The issues of morale, turnover, absenteeism and organizational
commitment are important to organizations. It is in the organization's
best interest to maximize employee job satisfaction. To address this
problem, a broad spectrum of theories regarding the determinants of
job satisfaction have been proposed.
Need theories. Historically, most satisfaction models have
emphasized the job itself as the principal source of job satisfaction.
For example, some of the earlier job satisfaction research was drawn
from Maslow's Hierarchy of Needs Theory (Hackman & Oldham,
1976; Lawler, 1982; Maslow 1970). Need theories were based on
the idea that, as individuals try to achieve self-fulfillment, their
satisfaction in the workplace will be directly related to the
environments ability to satisfy these needs (Maslow, 1970; Herzberg
et al., 1957; Hackman & Oldham, 1976). Accordingly, job design and
redesign were the suggested courses of action for increasing the
workers' satisfaction (Hackman & Oldham, 1976).
However, the validity of need theories has been questioned
(Salancik & Pfeffer, 1977). Many need theories were characterized
as being vague and unwarranted; unsupported by empirical evidence
for an actual hierarchy of needs (Salancik & Pfeffer, 1977). These
criticisms and others led to the virtual abandonment of need theory as
an explanation for job satisfaction (Weiss & Adler, 1984).
Environmental determinism. The ensuing movement was towards
"environmental determinism." The environmental determinism
movement placed primary focus on situational factors (Weiss & Adler,
1984). Emphasis was placed on situations surrounding the job,
rather than the individual within the position. The research of this
movement directed efforts toward determining what kinds of
environments foster job satisfaction, then reproducing those settings
in the workplace. This research was divided into two camps, those
advocating job enrichment and those advocating social information-
processing (Hackman & Oldham, 1976; Salancik & Pfeffer, 1977).
Job enrichment theory posited that to increase employee satisfaction
the job task must be enriched (Hackman & Oldham, 1976). As in
earlier job design approaches, emphasis was placed on the satisfying
properties of the position and not on the individual. Those endorsing
social information-processing believed that people view things as
necessary or unnecessary based on the way they process that item in
their own unique social environment (Salancik & Pfeffer, 1977). Thus,
the way in which individuals perceive their environment must be
manipulated to increase satisfaction (Salancik & Pfeffer, 1977).
Although the theories of job enrichment and social-information
processing differ in perspective, they both define the subject of job
satisfaction primarily from a situational/environmental approach
(Spector, 1988). They imply that the situation in which people are
placed determines the level of satisfaction or dissatisfaction. To affect
satisfaction, the environment must be manipulated. This perspective
resulted in research that was situationally-based, essentially ignoring
the individual involved (Salancik & Pfeffer, 1977). Critics of these
theories found them to be oversimplified and incomplete (Zalesny &
Ford, 1990). Social information processing theories were found to
have few empirical linkages to satisfaction, lacking the inclusion of
individual differences in their equations (Zalesny & Ford, 1990).
Similarly, an attempt to give empirical support to the effectiveness of
job enrichment failed to note the relatively short time period before
subjects regressed to original attitudes (Gerhart, 1990). The study
found interventions significantly increased satisfaction but recidivism
to baseline satisfaction occurred within a year (Gerhart, 1990).
Dispositional approach. More recent theorists have proposed that
job satisfaction comes not merely as a reflection of the environment in
which one performs, but also from within the individual (Arvey et al.,
1989; Cropanzano et al., 1991; George, 1989; Judge & Hulin, 1991;
Staw & Ross, 1985; Staw et al., 1986). Their findings are consistent
with early studies of job satisfaction (Hoppock, 1935; Munsterberg,
1913). Hoppock (1935) determined that dispositional considerations
were at least as important as the situational factors involved with
satisfaction. Prior to that, Munsterberg (1913) stated that disposition
was as much of a cause of dissatisfaction as the work environment.
However, research in the 1970's and early 1980's was so focused on
situational variables that dispositional variables were abandoned or,
at best, treated as confounding variables (Salancik & Pfeffer, 1977).
Recently, however, interest has again been focused on the
individuals within the work environment (Staw & Ross, 1985). This
new approach, known as the dispositional approach, is based on the
hypothesis that individuals carry with them dispositional attitudes,
predetermining whether they view their environments or objects in
their environment, in a positive or negative fashion (Costa & McCrae
1980; Cropanzano & James, 1990; George, 1989). It is suggested
that, because there is temporal stability to dispositional traits, one's
positive or negative view would be transferred from situation to
situation (Arvey et al., 1989; Costa & McCrae, 1980).
The dispositional approach is founded on two basic assumptions:
1) Different individuals perceive the same situation with different
levels of optimisfn or pessimism, and 2) Individuals show attitudinal
consistency across time and situation (Schneider, 1983; Schneider &
Dachler, 1978; Staw & Ross, 1985). These assumptions are
Different people perceive things differently. Most job situations
and environments are fairly ambiguous (Salancik & Pfeffer, 1977).
That is, the "reality" of the work environment is open to considerable
differences of interpretation. Salancik and Pfeffer (1977) propose that
"reality" is interpreted through the socialization and background each
person has experienced and carries into every situation. O'Reilly,
Parlette and Bloom (1980) found that even jobs having identical job
descriptions were viewed disparately by different individuals. An
identical work situation may be interpreted differently and therefore
reacted to, both behaviorally and affectively, according to distinct
individual proclivity (Salancik & Pfeffer, 1977). Support for this
proposition comes from a study by Kraiger, Billings, & Isen (1989),
who found that affect influences subjects' ratings on both task
characteristics and satisfaction. Kraiger et al. investigated the effect of
affect on task by measuring task satisfaction and task characteristics
after manipulating subjects' mood. They found that affect influenced
subject's perception of the task and that it was affect and not the
characteristics of the task that determined the satisfaction
Attitudinal consistency across time and situation. There is also
evidence for temporal and contextual consistency of job satisfaction.
Staw et al., (1986) found a significant correlation of .35 for job
satisfaction over lengthy periods of time. Individuals were found to
have significantly correlated levels of job satisfaction ranging over the
entire 40-year span of their 1986 study. Their findings held true for
individuals varying in sex and socio-economic backgrounds. This
intergenerational, longitudinal study supports the temporal stability of
A meta-analysis, conducted by Tait, Padgett & Baldwin (1989),
looked at the correlation between job satisfaction and general life
satisfaction over 34 studies with a combined n=19,811. They found a
highly significant correlation of .44 for job and life satisfaction;
indicating that the level of satisfaction experienced in one's job is
highly correlated to the over-all satisfaction experienced in life. This
relationship is consistent with the existence of a dispositional factor to
Pulakos & Schmitt (1983) also found evidence of temporal stability
of disposition in their study involving the satisfaction of high school
students. They found significant correlations between satisfaction
measures taken over several years of high school, as well as
measures taken when the students entered the workplace. Their
sample included those who changed schools and those who worked
in more than one position during the study (Pulakos & Schmitt, 1983).
In their 1985 landmark study, Staw & Ross looked at the stability of
job satisfaction over time. They found significant correlations of job
satisfaction, measured over several years. Even subjects who had
left their original jobs, reported consistent levels of satisfaction over
time. Longitudinal data were obtained through the Longitudinal
Survey of Maturb Men collected by the Center for Human Resource
Research at Ohio State University. Individual's job attitudes were
sampled several times over a five year period. Consistency in job
satisfaction was shown at both the two year and the five year
measurement. As expected, the two year findings were more highly
correlated with original measurements of satisfaction; however, the
results were still significant at the 5 year measure.
Finally, job satisfaction has been investigated from a genetic
prospective. In a study of monozygotic twins, raised apart with no
selective placement, Arvey and his associates (1989) estimated the
genetic component of job satisfaction by comparing monozygotic and
dizygotic twin pairs, raised apart, while controlling for differences in
job situations. A sample consisting of monozygotic twins (twenty-five
female twin pairs and nine male twin pairs) were compared with a
sample of dizygotic twins (N=146). Job satisfaction was measured
with the Minnesota Job Satisfaction Questionnaire (Weiss, Dawis,
England & Loquist, 1967). By comparing within-group differences to
between group differences, Arvey et al. suggested that 30% of
satisfaction in the workplace is due to the genetic make-up of the
individual. This 30% was calculated after controlling for pay and
other situational influences. Although Arvey et al. did not claim that
DNA causes job satisfaction, they do propose that job satisfaction is
In summary, the evidence for a dispositional factor to job
satisfaction is significant (Arvey et al., 1989; Cropanzano & James,
1990; Cropanzano et al., 1991; George, 1989; Judge & Hulin, 1991;
Staw & Ross, 1985; Staw et al., 1986). There does appear to be a
temporal constancy to satisfaction (Staw & Ross, 1985) as well as an
individual proclivity towards viewing similar situations from different
perspectives (O'Reilly et al., 1980). Based on this evidence, the
current study attempted to predict future satisfaction based on past
tendency toward satisfaction, using biodata.
Biodata is a quasi-longitudinal, quantitative self-report method,
used to predict future behavior and attitudes, based on past behavior
and attitudes (Mumford & Owens, 1987). The applicant is asked to
remember and report important life events through standardized,
structured multiple choice and/or true-false questions (Mumford &
Biodata techniques date back to 1894 when standardized
questions on life history were proposed for the use of hiring life
insurance agents (Mumford & Owens, 1987). Biodata was first used
around World War I and then again in World War II where it was found
to be accurate in predicting successful Army pilots (Mumford &
The techniques used in present biodata tests have evolved since
World War II. However, the primary assumption remains the same:
Biodata's predictive validity is based on the fact that an individuals'
past behavior and attitudes will predict their future behavior (Hammer
& Kleinman, 1988; Mumford & Owens, 1987; Reilly & Chao, 1982).
The content of the questions in a biodata questionnaire is similar
to that of a structured interview (Mumford & Owens, 1982). Questions
are asked of the applicant regarding accomplishments, attitudes, and
experiences that he or she may have had in the past (Asher, 1972).
These questions are designed to extract information deemed to be
important and job relevant.
Using biodata to predict job satisfaction. Biodata's most basic
assumption is that past behaviors and attitudes are predictive of future
behaviors and attitudes (Mumford & Owens, 1987). It is more useful
for predicting stable attitudes or behaviors than those that change
constantly or are strictly situationally dependent (except however, for
predicting possible patterns of how attitudes and behaviors might
change or how one might feel or behave, given that exact same
situation) (Hoiberg & Pugh, 1978). Therefore, to show that biodata
can be used to predict job satisfaction, two propositions must be
supported: 1) job satisfaction has a degree of constancy across time
and situation, and 2) future job satisfaction can be predicted by past
job satisfaction. The temporal consistency of satisfaction of job
satisfaction was established in the previous segment. Therefore this
discussion will be limited to predicting future job satisfaction based on
past satisfaction on the job.
Past job satisfaction as a predictor of future job satisfaction. The
literature points to the possibility of predicting job satisfaction. Staw
and Ross (1985) found prior attitudes regarding job satisfaction to be
a "strong predictor" of future attitudes towards job satisfaction. Arvey
et al. (1989) also suggest the use of present and past levels of
satisfaction to anticipate future satisfaction. Finally, in their "Actuarial
Approach" to job satisfaction, Buss & Craik (1983), theorize that past
attitudes can be used to predict future attitudes over time and
Why biodata? If job satisfaction can be predicted, is biodata a
valid tool to predict it? There is evidence that suggests it is preferable
to other measures such as personality tests. Although personality
instruments may get at the construct of satisfaction, there are findings
that show emplpyees are less open to taking a personality test than
filling out a biodata questionnaire (Hough, 1984). This has been
found particularly true of professionals who prefer to be assessed
based on their past record and performance rather than on what they
view as a one-shot, invasive measure of personality (Hough, 1984).
Findings also suggest that many personality tests are not as accurate
at predicting traits as a well designed biodata questionnaire (Hoiberg
& Pugh, 1978). Biodata appears to be a valid and perhaps preferred
method of predicting job satisfaction.
Using the construct of classroom satisfaction to simulate iob
satisfaction. To demonstrate the potential validity of biodata for
predicting satisfaction, the current study will examine its efficacy for
predicting satisfaction in a more accessible environment, the college
classroom. The use of classroom satisfaction is warranted on three
grounds. First, theories of job satisfaction and classroom satisfaction
propose many of the same determinants or outcomes. Both job
satisfaction theorists and classroom satisfaction theorists have
emphasized quality of life, performance, turnover and absenteeism as
outcomes of satisfaction (Aitken, 1982; Betz, Klinginsmith & Menne,
1970; Bean & Bradley, 1986; Cropanzano et al., 1991; Edwards &
Waters, 1982; George, 1989; Locke, 1969). Betz and her colleagues
developed an instrument entitled, the College Student Satisfaction
Questionnaire (CSSQ), to measure college student satisfaction,
under the assumption that the constructs of job satisfaction and
classroom satisfaction are interchangeable (Betz et al., 1970). The
dimensions of classroom satisfaction measured by the questionnaire
are taken from the job satisfaction literature (Herzberg et al., 1957).
These dimensions include, compensation (grades versus studying),
working conditions, social life, policies and procedures, recognition,
and quality of education (work experience) (Betz et al., 1970). The
CSSQ (Betz et al., 1970) has not only been shown to be a valid
measure of student satisfaction (Strong, 1978), it continues to be one
of the most widely used tools for measuring student satisfaction (Okun
& Weir, 1990).
Secondly, several studies have found that satisfaction in one
environment is highly correlated with satisfaction in another (Pulakos
& Schmitt, 1983; Schmitt & Bedian, 1982; Tait et al., 1989). Further,
these authors have suggested that satisfaction in all environments are
due to a single general satisfaction component. Therefore, it is likely
that satisfaction experienced in the classroom may be similar to that
experienced in the workplace.
Finally, classroom satisfaction is an important variable in its own
right and it has received considerable attention in the education
psychology literature. For example, Astin (1977), argued that student
satisfaction with the college experience is equally important as any
other outcome of the educational experience. Noel (1985), in
studying student retention, found student satisfaction to play a key
role in reducing student turnover; confirming the findings of previous
work (Munro, 1981; Pascarella & Chapman, 1983; Terenzini &
Pascarella, 1980). Finally, student satisfaction has been linked to
important educational processes and outcome factors such as student
persistence and academic achievement (Aitken, 1982; Bean &
Bradley, 1986; Cohen, 1987).
This study attempts to develop a biodata questionnaire, the
Biodata Questionnaire (BQ), which is capable of predicting classroom
satisfaction. In so doing, it is hoped that further support will be
provided for the dispositional argument for job satisfaction as well as
create a tool to measure it.
In attempting to achieve the goals of validating the Biodata
Questionnaire (BQ) and demonstrating it's predictive abilities, the
following four hypotheses are offered.
The first hypothesis concerns the relationships between prior
satisfaction and negative and positive affectivity. Positive and
negative affectivity are generally considered as degrees of
pleasantness and unpleasantness, accordingly (Watson & Tellegen,
1985). Negative affectivity refers to general aversive psychological
states including distress, anger, frustration and anxiety (Watson et al.,
1988). Positive affectivity is considered a separate construct and
refers to generally pleasant engagements including excitement,
enthusiasm and alertness (Watson, et al.,1988).
Hypothesis 1 is based on previous research, showing a
relationship between the positive and negative affectivity variables
and satisfaction (George, 1989). George (1989) as well as
Cropanzano et al., (1991) found a significant correlation between
satisfaction and affectivity.
It is expected that affectivity will be related to past satisfaction.
Therefore, a significant positive correlation is predicted
between the BQ and the Positive Affectivity Scale (PAS) of the
Positive and Negative Affect Schedule (PANAS) and a
significant negative correlation between the BQ and the
Negative Affectivity Scale (NAS) of PANAS.
Hypothesis two investigates the relationship between efficacy and
satisfaction as measured by the BQ. Efficacy refers to the degree to
which an individual feels confident in her or his ability to perform.
In addition to affectivity, subjects' confidence of their ability to
succeed (efficacy) has also been shown to be related to satisfaction
(Okun & Weir, 1990). Okun & Weir (1990) found students who had
performed well in the past and who had confidence in their ability to
perform well in the future, showed greater satisfaction. Efficacy was
also found to be positively related to satisfaction in a 1989 study of
mathematics students. Those students who were more confident in
their ability to succeed were more satisfied than those who
questioned their own abilities (Hackett & Betz, 1989).
Subjects who experienced high satisfaction will have stronger
efficacy than those who were not satisfied in the past.
Therefore, a positive correlation is expected between the
Biodata Questionnaire and the Efficacy measure.
Related to the predictive abilities of the BQ is the question of
biodata's relative ability to predict satisfaction in comparison to other
variables measuring related constructs. This will help support the
divergent-related evidence of validity of the BQ. This will be analyzed
using two variables; expected course grade (a traditional predictor of
classroom satisfaction), and PANAS, which measures an individual's
tendency toward positive and negative affectivity.
The BQ is expected to predict variance in classroom satisfaction
that is unaccounted for by expected course grade. A meta-analysis
conducted by Cohen (1982) looked at the relationship between
expected grade and course rating and found expected course grade
to be a significant predictor of classroom satisfaction. Hudson (1989)
also found a significant correlation between expected course grade
and satisfaction. Biodata is expected to be a superior predictor on
two grounds: 1) previous expectations around grades are
incorporated into the measure of past satisfaction and therefore the
biodata instrument should account for more variability than the
measure of current expectation of grade alone, and 2) the Biodata
Questionnaire was designed for prediction whereas expected grade
In addition, the BQ is expected to predict classroom satisfaction
above and beyond the personality measure taken with the PANAS
scale. Research has shown personality measures to be related to
satisfaction (Hough, 1984). However, it was also found that subjects
perceive personality measures to be more invasive than biodata or
biographical questions (Hough, 1984). In addition, the BQ is
expected to be a better predictor than PANAS because, again, the BQ
was designed specifically for predicting classroom satisfaction,
whereas PANAS was not.
Therefore, the following hypothesis is offered.
Biographical satisfaction will add significant prediction of
satisfaction above and beyond measures of similar constructs.
3a. The BQ is anticipated to predict satisfaction beyond the
prediction of the Expected Class Grade measure.
3b. The BQ is expected to predict satisfaction beyond the
prediction of the PANAS scale.
One of the primary concerns of the current research is to
investigate the possibility of predicting satisfaction using biodata.
Theorists have suggested that satisfaction can be predicted (Arvey et
al., 1989; Buss & Craik, 1983; Staw & Ross, 1985), however there has
been no research around using biodata to do so. The fourth
hypothesis addresses this critical research question.
Satisfaction with past university experiences will be predictive
of future college satisfaction. It is expected that the
biographical satisfaction information from the Biodata
Questionnaire, collected on the first day of class, will be
positively correlated with the CSSQ, administered near the end
of the semester.
A sample of 192 students from an urban University was surveyed.
The participants were drawn from five upper class courses to provide
a broad classroom base and to help assure all had previous college
experience. Twenty-five were sophomores, 74 were juniors, 81 were
seniors, 3 were at the graduate level and 9 were non-degree students.
The average subject tenure at the present University was 2.5 years.
The mean age of the sample was 26.95 (s=7.72) with a range from
18 to 52. Forty-seven of the subjects were male and 145 were female.
Unfortunately, there was a high attrition rate between the first and
second survey administrations. This was largely due to a scheduling
conflict which resulted in the inability to collect data on the second
occasion from an entire class of 50. Subjects were also lost because
students dropped a course in which data were collected or were not in
attendance the day of survey administration. Accordingly, the final
number of subjects (for which data were available on both occasions)
Subjects were sampled at two times during the semester. Data
were collected on the first day of class and again approximately 12
weeks into the 16-week semester. Surveys were administered during
class period; subjects were assured of confidentiality and offered
extra credit for participating in the study. They were notified that to
receive full extra credit, participation was required in both parts of the
Figure 1 lists the questionnaires administered on each occasion.
The surveys included the Biodata Questionnaire, PANAS (Watson et
al., 1988), Efficacy, Expected Grade, and the CSSQ (Betz et al., 1970)
at the second collection time (see Table 1). All surveys were
randomly ordered within administrations.
College Student Satisfaction Questionnaire
Scores on the College Student Satisfaction Questionnaire (Betz et
al., 1970) were the primary dependent measure. The instrument
Table 1. Administration Schedules
Time 1: Biodata Questionnaire
Efficacy in Course
Positive And Negative Affect Schedule
Time 2: Biodata Questionnaire
College Student Satisfaction Questionnaire
Estimated Course Grade
measures six logically-derived and empirically-supported dimensions
of satisfaction (Betz et alM 1970). The dimensions were originally
generated based on previous job satisfaction research (Herzberg et
al., 1957) and included: policies and procedures, working conditions,
compensation, quality of education, social life and recognition. The
responses were recorded on a 5-point Likert scale ranging from
1 ."Very Dissatisfied" to 5, "Very Satisfied." Subjects were asked to
indicate the response that best reflected their present satisfaction in
that area. Sample items included, "The practice you get in thinking
and reasoning," "The amount of work required in most classes," and
"The amount of personal attention students get from teachers" (See
Appendix C for complete survey). Betz and her colleagues (1970)
found the CSSQ sub-scales to be highly reliable with Cronbach
Alphas ranging from .85 to .92.
Because the CSSQ was originally intended for use with a
traditional student body and campus setting, its use in an urban,
nontraditional university setting for this research created some unique
problems. It became apparent during data collection and data entry
that the social sub-scale was not valid for many in my sample.
Specifically, the average age of this sample was 26.95 with a range
from 18 to 52. In addition, many of these subjects were married with
families. However, the items on the CSSQ social scale were written
for unmarried students in a traditional college campus setting: "The
choice of dates here" or "The activities that are provided to help you
meet someone you might like to date." These items were confusing
for many participants (indicated by comments, and questions during
testing, missing data and a lack of variability on most items in the
social life sub-scale). Since this was an exploratory study, and the
social life scale had dubious validity, it, and the corresponding social
scale on the Biodata Questionnaire were dropped from the analyses.
The Biodata* Questionnaire (BQ) is a 54-item survey developed to
assess the tendency towards satisfaction with the college experience.
A thorough literature review was conducted in the areas of job
satisfaction, classroom satisfaction, intrinsic motivation and positive
and negative affectivity to provide parameters for item development.
With the assistance of my advisor, I generated items based on the
dimensions found in the literature review.
Items were generated in 15 sub-scale categories: general college
satisfaction, expectations, environment, social life, recognition, quality
of education, skill variety, task significance, task identity, autonomy,
feedback, alertness, interest, patience with system and self-
handicapping. The constructs for the sub-scales were selected to
reflect the job satisfaction-related sub-scales of the CSSQ, the
dimensions of intrinsic motivation (Hall & Richter, 1990; Hoiberg &
Pugh, 1978) and areas we felt contributed to satisfaction based upon
our experience in academia. Items were rated on a 5-point Likert
scale reflecting how often subjects were satisfied with the item in the
past. The anchors ranged from 1, "almost never" to 5, "almost always"
Sample items included "In the past I've felt satisfied with my
performance in school," "In past college classes I have received the
grade I deserved," and "Past classes have offered me the freedom
and independence to determine how I accomplished the work
required" (see Appendix B for the complete questionnaire). Scoring
was done by sub-scale with higher scores indicating higher past
Table 2: Biodata Questionnaire Sub-Scales
Quality of Instruction
Patience with System**
* Collapsed into the Intrinsic Motivation scale
** Dropped from analyses
A scale analysis was performed to analyze the sub-scales. The
analysis generated item-to-scale total correlations, correlations
among the scales and scale reliabilities. The five intrinsic motivation
scales (skill variety, task significance, task identity, autonomy and
feedback) were all found to be significantly intercorrelated with an
average correlation of .39 (calculated using Fisher's z). Therefore,
they were combined to create a single intrinsic motivation scale for
subsequent analyses (reliability was .85).
The reliability for five scales was low: environment (.07), alertness
(.16), interest (.55), patience with system (.13) and self-handicapping
(.31). Given these low reliabilities and the fact that there was not a
compelling theoretical rationale for retaining these scales (these were
not based on the CSSQ sub-scales or on intrinsic motivation
dimensions), they were dropped from subsequent analyses.
The remaining five scales, general satisfaction, internal motivation,
expectations, recognition and quality of instruction were then
analyzed using the SPSSX reliability program. No scale was shown
to be improved by removing items. On the grounds of homogeneity of
individual scales as well as the theoretical basis on which they were
developed, these five scales were retained for analyses, in their
entirety (see Appendix A for items within scales).
Demographic information was collected along with the BQ.
Participants were requested to indicate their age, tenure with the
university, year in school (eg., senior, junior, etc.) and gender.
Because of a coding/data entry error, subjects could only be identified
by gender at the aggregate level; thus, data could not be correlated
with sex. Analyses were run on age, tenure at the University and year
in school. No significant differences were found on any variables,
therefore analyses were run only on the entire sample.
Positive And Negative Affect Schedule
The Positive and Negative Affect Schedule (PANAS) is a 20-item
scale developed by Watson et al., (1988) to measure positive
affectivity (PA) and negative affectivity (NA).
The PANAS requires subjects to rate single word items as to what
extent each item described how they felt. The 5-point Likert rating
scale ranged from 1, "very slightly or not at all" to 5, "extremely." The
sub-scales for PA and NA were scored separately, with high scores
indicating a high degree of positive and/or negative affectivity
PANAS is designed so that the emotions or feelings can be rated
by how subjects feel at the moment, today, past few days, week, past
few weeks, year or generally, by altering the instructions. Because we
wanted to get a general sense of the participant's affectivity
orientation, we asked subjects to rate "to what extent you generally
feel this way, that is, how you feel on the average". Sample items
included "interested," "upset," "ashamed," "jittery," and "active" (see
Appendix D for the complete survey form). Watson et al. (1988)
reported alpha reliabilities of .88 for the PA scale and .87 for the NA
Efficacy was measured by a 3-item scale I developed with the help
of a member of my committee. Efficacy is defined as how confident
one is in her or his ability to perform. Based on this definition, subjects
were asked to rate how confident they were that they would receive a
particular grade in the up-coming class. The rating was indicated on a
scale that ranged from 0% confident to 100% confident. For example,
"How confident are you that you will receive at least a C in this
course?" (See Appendix E for complete survey). Scores were
generated by summing the percentage response to each item.
Therefore, if a subject was 100% confident that she or he would
receive an A, the score would be 300 (100 for the C item, 100 for the B
item, and 100 for the A item). If a subject felt 100% confident of getting
a C, 80% confident of getting a B, and 60% confident of getting an A,
the score would be 240.
Expected Course Grade
On the second administration, subjects were asked to provide their
anticipated grade in the course. Subjects selected one from among
options ranging from F to A.
Data were first analyzed for descriptive information, including
means, standard deviations, ranges and univariate correlations. The
hypotheses were then analyzed using the following methods.
The first hypothesis (the BQ will be positively related to PA and
negatively related to NA) was tested by computing Pearson Product
Moment correlations between the PANAS sub-scales and a unit-
weighted composite of the Biodata Questionnaire sub-scales.
Hypothesis 2 ;
Hypothesis two (the BQ will be positively related to efficacy) was
measured using a Pearson Product Moment correlation between the
efficacy scale and a unit-weighted composite of the BQ sub-scales.
Hypothesis 3 ;
Hierarchical regression analyses of Expected Course Grade and
Biodata on CSSQ, and PANAS and Biodata on CSSQ were
performed to examine hypothesis three. Expected grade was first
regressed bn CSSQ (Betz et al., 1970); at the second step, Biodata
was added; the 'change in R2 at the second step was investigated to
test the hypothesis. The second analysis mirrored the first,
substituting PANAS for expected course grade. In this manner, the
BQ's ability to predict classroom satisfaction over and above
measures of similar constructs was assessed.
Hypothesis 4 ;
Hypothesis Ffour posits that scores from the first administration of
the BQ will be positively related to scores on the CSSQ. This was
tested via a Pearson Product Moment correlation between unit-
weighted composites of the BQ and the CSSQ. The sub-scales of the
BQ were also correlated with the CSSQ composite to assess the
relationship of each individual sub-scale and the composite CSSQ.
Finally, a canonical correlation was conducted to estimate the upper
limit of the degree of relationship between sub-scales of the two
Descriptive statistics for major variables can be found in Table 3.
Table 3. Major Study Variables : Correlations, Means, Standard
Deviations, Minimum and Maximum Values.
1 2 3 4 5 6 7 8
Composites (1) BQ (2) CSSQ BQ Sub-Scales .39**
(3) General .82** .19*
(4) Expectations .74** .26** .62**
(5) Recognition (6) Internal .67** .40** .41** .46**
Motivation .93** .26** .62** .57** .58**
(7) Quality of Instr. CSSQ Sub-Scales .71** .40** .52** .43** .44** .59**
(8) Quality of Ed. .32** .72** .26** .15 .32** .29** .24**
(9) Compensation (10) Availability of .25** .61** .23** .15 .21* .19* .27** .55**
Help (11) Policies and .17 .75** .05 .14 .26** .12 .31** * 00
Procedures (12) Working .20* .76** .10* .21* .30** .13 .24** .34**
Conditions .19* .63** .13 .17 .32** .12 .16 .28**
(13) PAS .56** .29** .52** .39** .34** .50** .41** .24**
(14) NAS -.37** -.12 -.31** -.28** -.25** -.33** -.24** -.14
(15) Efficacy (16) Expected .31** .08 .22** .31** .31** .28** .13 .28**
Grade .15 .28** .25** .24** .10 .02 .10 .08
Mean 132.9 114.5 29.8 19.7 11.2 61.6 10.6 26.3
SD 14.5 16.6 4.1 2.2 1.7 7.5 1.9 4.4
Minimum 85 76 20 13 7 33 6 13
Maximum 166 158 39 25 15 83 15 35
minimum n=92. maximum n=192
Table 3. (continued)
! 9 10 11 12 13 14 15 16
(10) Availability of Help (11) Policies and CO
Procedures (12) Working ' .25** .46**
Conditions ; .27** .24* .45**
(13) PAS .20* .23* .13 .15
(14) NAS ' -.24** -.15 .11 -.07 -.21**
(15) Efficacy ".31** .02 -.04 .09 .10 .33** -.12
(16) Expected Grade.24** .16 .13 .28** .23* .18 -.07 .17
Mean i 20.9 21.5 25.8 19.8 36.6 19.7 261.9 9.9
SD 3.9 5.6 5.8 4.4 5.5 5.9 38.3 1.9
Minimum : 6 9 11 8 13 10 110 1
Maximum 29 35 41 30 48 38' 300 12
minimum n=92; maximum n=192
The major variables, including independent and dependent
variables, had acceptable ranges and standard deviations. The
inter-correlations among the sub-scales tor both the BQ and the
CSSQ were found to be moderate to high.
Test-retest &j alpha reliability coefficients for the Biodata
Questionnaire upit-weighted composite and BQ sub-scales can be
found in Table 4.: The test-retest reliability (with an approximate 12
week interval) wds .74, while the coefficient alpha estimate (calculated
on the first questionnaire administration) was .81. Thus the BQ
appears to be stable and internally consistent.
As may be expected, reliability estimates for sub-scales were
somewhat lower (.52 to .72).
Table 4. Classroom Satisfaction Biodata Questionnaire: Reliabilities
Biodata Questionnaire Composite .74 .81
General Satisfaction .72 .66
Internal Motivation .60 .85
Expectations .60 .47
Recognition i .59 .49
Quality of Instruction .52 .52
Test of Hypotheses
The first hypothesis was tested by examining the correlation
between the uniti-weighted composite of the BQ and both of the
affectivity sub-scales. As seen in Table 3. There was a significant,
positive correlation between the BQ and PA scale (r=.56, p<.01). In
addition, there was a significant, negative correlation between the BQ
and NA scale (Â£=-.37 pc.01). These findings support Hypothesis 1
and provide construct-related evidence of validity of the BQ. Subjects
high in positive Jaffectivity were more likely to score high on the BQ
(indicating a greater level of past satisfaction). In addition, subjects
who scored high in negative affectivity experienced less satisfaction in
the past, as indicated by their low score on the BQ.
In support of Hypothesis 2, a significant, positive relationship was
found between the BQ and the Efficacy measure (r=.31, p<.01).
Participants who felt high efficacy experienced a higher degree of
previous satisfaction than those who did not feel as capable. This
again provides construct-related evidence of validity of the Biodata
Two analyses were conducted to analyze part a and part b of
A hierarchical multiple regression was performed to analyze the
ability of the BQ ;to predict the CSSQ after accounting for expected
course grade. In step 1, expected course grade was regressed onto
the CSSQ, generating an R2 of .07 (j2<.01). Thus, the grade students
expected in the course was significantly related to satisfaction with
that course. When the unit-weighted composite of the Biodata
Questionnaire was added to the equation at step 2, R2 increased to
.19 (see Table 5). The change in R2 was significant (pc.OOOl), and, in
fact accounted for the bulk of the prediction, supporting the hypothesis
that the BQ predicts significant variance in satisfaction beyond that
accounted for by expected course grade and providing evidence of
divergent-related validity for the BQ.
Table 5. Hierarchical Regressions on CSSQ Composite
B & Significance of Change in R2
Expected Grade Biodata Composite .26** .44** .07** .19** JK.0001
PANAS Biodata Composite .30** .40** .09** .16** p<.005
A second hierarchical multiple regression analyzed the ability of
the BQ to predict the CSSQ after accounting for the PAS and NAS
scales of the PANAS. In step 1, PAS and NAS were regressed onto
the CSSQ, generating an R2 of .30 (p<.01). Therefore, it was
concluded that the positive and negative affectivity experienced by
subjects was refeted to their course satisfaction. However, adding the
Biodata Questionnaire to the equation at step 2 increased R2 to .40.
The change in R2 was again significant (p<.005), indicating the
variability accounted for by the BQ to be unique to that accounted for
by PANAS and fending further support to the divergent-related
evidence of validity of the BQ.
Hypothesis 4; the ability of the BQ to predict satisfaction as
measured by th$ CSSQ, was assessed by a) forming separate unit-
weighted composites of the predictor and criterion variables; and by b)
computing the correlation between these composites. The result was
a significant positive correlation between the BQ and the CSSQ (i=.39
Â£<.01). Anotheranalysis was conducted to investigate the
relationship of each BQ sub-scale to the CSSQ composite. To
investigate which particular sub-scales were the most highly related to
the CSSQ composite, univariate correlations were computed between
the composite and the sub-scale scores. Results are found in Table 6.
Table 6. Unit-Weighted Composite Correlations
i CSSQ Composite
BQ Composite .3863**
Quality of Instruction .3982**
Expectations ; .2578**
Internal Motivation .2548**
General Satisfaction n=m .1867*
*Â£<.05 *Â£<.01 j
A canonical correlation was also performed between the five
Biodata sub-scales and the five CSSQ sub-scales to estimate the
relationship between the variables. The first canonical correlation
was .50 and was statistically significant, using Bartletts (1941) Chi
Square test; no other canonical correlations were significant.
Inspection of structure coefficients revealed a large number of
suppressor variables, and other results were difficult to interpret.
Given these results, and the small subject-to-variable ratio, it was
decided that the! univariate analysis (using the unit-weighted
composites) wa6 the more trustworthy. Therefore, the canonical
correlation analyses were not considered further.
DISCUSSION, IMPLICATIONS, LIMITATIONS
! AND FUTURE DIRECTIONS
The primary'purpose of the present investigation was to develop
and evaluate a biodata questionnaire in terms of its ability to predict
satisfaction. The research issues were addressed by conducting a
thorough review of the satisfaction literature, developing the Biodata
Questionnaire, guided by the information gathered in the review, and
finally, collecting empirical evidence of the reliability and construct-
related validity of the scale.
Summary of Results
Construct-related evidence of validity of the instrument can include
a) evidence thatithe items are content valid; b) evidence that test
scores are internally consistent or stable over time; c) evidence that
scores of a test are correlated with tests of related constructs; and d)
evidence that scores on a test predict meaningful criteria (Campbell,
The content 'validity of the BQ was ensured by generating items
which were relevant to either the CSSQ or to the general nature of the
satisfaction construct. Initial items were written and 15 sub-scales
were developed; from knowledge of the classroom satisfaction, job
satisfaction, and internal motivation literature (Arvey et al., 1989; Betz
et al., 1970; Cropanzano et al., 1991; Hall & Richter, 1990; Hoiberg &
Pugh, 1978; Locke, 1976; Spector & Jex, 1991).
The reliability of the BQ was assessed by examining the internal
consistency of the unit weighted composite of the instrument and the
sub-scales forming the composite, as well as with test-retest
consistency. Coefficient alphas were very low for several of the sub-
scales. Because there was not a compelling theoretical rationale for
retaining these scales, they were dropped. The remaining sub-scales,
general satisfaction, internal motivation, expectations, recognition,
and quality,of instruction, showed adequate reliability (alpha's
between .52 and .66; test-retest r's between .49 and .72) as did the
unit-weighted composite (alpha = .81, test-retest i = .74). In future
applications, however, some items may be added, modified, or
deleted to improve the reliability of the instrument.
Additional construct-related evidence of validity for the BQ was
investigated by Examining the relationship between the BQ and
affectivity and the relationship between BQ and efficacy. Significant
relationships were found in the directions anticipated. Biodata was
positively related to efficacy and positive affectivity and was negatively
related to negative affectivity. These results support Hypotheses 1
Hypothesis 3 and divergent-related evidence of validity for the BQ
were supported by examining the relationship between the BQ and
the CSSQ after partialing out measures of similar constructs: expected
course grade and the PANAS scale.
A hierarchical multiple regression analysis found a significant
change in R2 when the BQ was regressed on the CSSQ, after
partialing out the variance explained by expected course grade.
Significant results were also found when the BQ was regressed on the
CSSQ after accounting for the positive and negative scales of the
PANAS. The BQ appears to account for unique variance within the
Several analyses were performed to determine the validity of
biodata as a predictor of satisfaction. A significant correlation was
found between unit-weighted composites of the BQ and CSSQ.
Subsequent univariate correlations between the sub-scales of the
BQ and the composite CSSQ revealed the recognition and quality of
instruction sub-scales as the primary predictors of satisfaction. It is
concluded, in support of hypothesis 4, that a biodata measure can
predict end-of-term classroom satisfaction.
Given the steps followed to develop the instrument, the resulting
reliability, and cpnstruct-related evidence of validity, it is concluded
that the BQ is psychometrically sound for use in classroom satisfaction
research. Additional refinements should be made before the
instrument is cohsidered adequate for uses Other uses.
Biodata as a Predictor of Classroom Satisfaction
This study supports biodata as a predictor of classroom
satisfaction. Knowledge of students' tendency towards satisfaction
was effective in predicting their future levels of satisfaction. The ability
to predict satisfaction may be helpful when attempting to affect student
satisfaction. It is known that students' satisfaction is affected by the
quality of instruction, fairness of evaluation, the type or difficulty of the
class, and the grhde received (Cohen, 1982). Instructors try to
"manipulate" many of these variables in an attempt to increase student
satisfaction and the resultant course evaluations. While it is important
to consider those situational factors within the professors control, the
present study implies that it is also important to consider the variance
in individual pre-disposition.
It is unlikely that professors will be able to change students'
dispositions. However, this does not imply that interventions will be
unsuccessful under any circumstances. Gerhart (1990) used the
analogy that, similar to individuals with poor eyesight who can be
helped with glasses, those most inclined to be dissatisfied may be the
most readily helped by intervention. Thus, Gerhart hypothesized and
found significant improvement in reported satisfaction from subjects
whose baseline satisfaction measures were the lowest. However,
even in Gerhart's work, those who were most predisposed toward
negative affectivity were found to change for less than a year, at which
point they reverted back to their original state of dissatisfaction
(Gerhart, 1990). This type of recidivism must be considered when
attempting to manipulate satisfaction. Professors may be able to
increase satisfaction to a degree by altering classroom situations, but
this study indicates that there is a dispositional aspect to satisfaction.
Theoretically it is posited that intervention efforts will be increased
when disposition is considered along with situation.
Biodata as a Predictor of Job Satisfaction.
Although this paper addresses the issue of classroom satisfaction,
a primary interest is in developing a technique that will be effective in
predicting job satisfaction. The support of the third hypothesis
suggests that satisfaction can be predicted, it is hoped that the
techniques used in this study may be effective in organizational
settings. Betz dnd her colleagues (1970) as well as other researchers
(Bean & Bradley, 1986; Edwards & Waters, 1982) have used many of
the traditional dimensions in job satisfaction to define Classroom
satisfaction. In addition, research supports that satisfaction tends to
extend across situation (Schmitt & Bedian, 1982; Pulakos & Schmitt,
1983; Tait et alj 1989). Because of the similarities in the constructs of
job and classroom satisfaction, and the tendency of satisfaction to
cross over situational boundaries, it is not unrealistic to believe that
the technique of,ipredicting future satisfaction from past satisfaction will
be successful injthe workplace.
If prediction is possible, as many researchers suggest (Arvey et al.,
1989; Cropanzano & James, 1990; Cropanzano et al., 1991; George,
1989), there may be far reaching implications for organizations. In the
past when there! have been morale problems within an organization,
one strategy has| been to redesign the job or the department so as to
enrich the position and better attend to the needs of the employees
(Hackman & Oldham, 1976). The findings of dispositional research
indicate that organizations also need to consider employee
disposition if there is to be a lasting effect on the over-all morale and
satisfaction within the company (George, 1989; Cropanzano & James,
1990; Cropanzano et al., 1991; Chen & Spector, 1991; Staw & Ross,
1985). Although it appears possible for some interventions to have an
effect on satisfaction, it may be necessary to alter them or reinforce
them, depending upon the dispositional tendencies of the staff
(Gerhart, 1987; Staw et al., 1986; Weiss & Adler, 1984). In other
words, when developing the content, strategy and timeline of
interventions, effectiveness may be increased by considering the pre-
disposition towards satisfaction of the staff involved.
An additional, implication of the present study is the use of biodata
as a placement tool. Once organizations have selected employees
they are often faced with the difficult task of finding the right "fit" within
the organization. Concurrently, it has been posited that high levels of
satisfaction may be disadvantageous in situations where stressful
decision making and analytical judgement are required (Staw & Ross,
1985). If such situations can be identified, the ability to predict
satisfaction with a biodata instrument may be helpful in identifying
individuals who will be more or less pre-disposed toward working in
them. Ideally, if a causal relationship between satisfaction and
performance could be inferred and an organization had knowledge of
the satisfaction-;performance relationship for all jobs within the
company, the organization could use biodata to increase performance
by placing employees into positions that best suited their tendency
Predicting Satisfaction and the Dispositional Argument
A final implication of this study is the support it lends to the
dispositional approach to satisfaction. The dispositional approach is
based on two premises: 1) Individuals show attitudinal consistency
both temporally and contextually, and 2) Different individuals perceive
the same situation with different levels of optimism or pessimism
(Schneider, 1983; Schneider & Dachler, 1978; Staw & Ross, 1985).
This study demonstrated that present satisfaction can be predicted
from knowledge of past satisfaction. To do this, there must have been
some temporal stability in the satisfaction experienced by the
participants. In fact, the ability to predict satisfaction from knowledge
of past satisfaction implies that satisfaction has a degree of temporal
consistency; supporting the first assumption of the dispositional
In addition, the present study found consistency oyer different
contexts. Past satisfaction was measured over each of the subjects
entire collegiate career. Subjects were asked to recall satisfaction
across classes, years and in some cases, institutions. Many of the
students in the sample were transfer students, as indicated by the
disparity in tenure in the present university and their year in school
(mean tenure = 2.5 years, mean year in school = 3.36 years).
However, the measure of past satisfaction obtained with the BQ was
still able to account for between 14% and 36% of variance in present
satisfaction, as indicated by the unit-weighted composite analysis and
the canonical correlation. This would suggest that something besides
the situation is affecting satisfaction.
The findings of this study also support the second premise of
dispositionality, that different individuals will have differing
perceptions regarding similar situations. Subjects in the same class
reported variable levels of satisfaction. While the range for the CSSQ
was 82, the standard deviation was 16.65, clearly indicating that
different individuals were perceiving similar situations with varying
levels of optimism, supporting the second premise.
This study thus supports the dispositional argument by
demonstrating temporal and situational consistency of satisfaction, as
well as that different individuals viewed similar situations with differing
levels of satisfaction.
There are several limitations to the present study. Of particular
concern are the problems created by the sample, the validity of the
measures and potential confounding variables.
One limitation to the present study was the small sample size. The
final sample analyzed was comprised of 115 students enrolled in
undergraduate psychology courses. The small sample size may have
had negative ramifications for the canonical correlation analysis,
causing the weights and the resultant Rc to be unstable and
uninterpretable. This was apparent in the weights and their
relationship to the structure matrix as well as in the comparison of the
canonical results to the unit-weighted composite findings.
In addition, because of the small sample, I was unable to split the
sample into two groups and compare a criterion sample to a validation
sample as initially planned. This limits the generalizibility of the
Finally, the sample was relatively limited in its representativeness.
The sample contained primarily middle class, white, female subjects.
There is concern that the generalizability of these findings is
minimized because of these limitations.
Validity of the Measures
Concern is raised over the limitations imposed by the dependent
measure, the CSSQ. Of particular concern was the social life scale
which was eliminated from the study because of apparent error
caused by the idiosyncrasies of the sample. The social life sub-scale
was out of context for the particular environment under study.
Although the CSSQ scales appear to have prior validity (Okun & Weir,
1990; Strong, 1978), as well as face validity, further validation of the
CSSQ (Betz et al., 1970) should be conducted in a non-traditional
setting before using it again under similar conditions. Pilot-testing of
the instrument may have uncovered the problems caused by the
idiosyncracies of the present sample. Any unreliable or invalid scales
could then have been eliminated prior to administration in the actual
study, so as to minimize any confounding effects. An alternative
would be to design a satisfaction measure specifically for the study.
Attention should also be called to the biodata questionnaire.
Typically, the development of any biodata instrument is an interactive
process. In the future, more work could be done to refine the items,
questions, scales, and responses.
In reviewing the relationships found between the BQ and the
convergent variables, there exists the possibility of confounding
variables. Qf particular concern is the relationship between efficacy
and past satisfaction. Although the literature supports a relationship
between satisfaction and efficacy (Hackett & Betz, 1989), it is plausible
that the obtained correlation could have been caused by a mutual
relationship with; a third variable. This issue should be addressed in
future research to assure that the relationship is not spurious.
Finally, concern is raised regarding common method variance.
Both the BQ and the CSSQ are paper-and-pencil measures, with a
similar 5-point Likert format. The differential relationship among the
BQ, CSSQ and other variables in the study, as indicated in Table 5,
are inconsistent with such an interpretation. For example the BQ is
significantly correlated with the NAS scale of PANAS and the efficacy
scale (both pa^er-and-pencil, Likert scales); yet neither is significantly
related to the CSSQ. In the subsequent studies, the concern
regarding common method variance could be eliminated by collecting
data through alternative measures such as peer/spousal report or
Many opportunities exist for future research in the area of
predicting satisfaction, including replicating and cross-validating of the
present study, conducting a similar study in an organizational setting,
and considering additional variables.
In replicating the present study, several recommendations are
made. It is suggested that future research efforts be conducted with a
larger sample arjid include a better representation of gender, ethnicity
and socio-economic factors. This will enable findings to be more
stable and generalizable.
In addition, itj would be beneficial to obtain alternative measures of
satisfaction. Because of the difficulties associated with using the
CSSQ, it is recommended that an instrument be used that is less
situationally specific and more valid.
Finally, because consistency of baseline satisfaction is pivotal to
dispositional research, it is suggested that the predictive validity of the
BQ be established for different time intervals. This could be
accomplished by collecting measures at several different times over a
longer period. In addition to supporting the temporal stability
arguments of disposition theory, measuring satisfaction at several
points in time will better define how far into the future biodata can
predict and at what point (if any) it becomes ineffective.
It is also recpmmended that future research design a study that
involves a cross-validation of the findings. Ideally the study would be
conducted in two parts. The first would be used to analyze the
statistical properties of the predictive instrument and to establish
weights for the sub-scales, items or responses. The second part
would then involve validating the weights established in the first study.
The weights established in a cross-validation design with a large
sample would be more likely to be stable and generalizable to the
population and therefore more useful in the future.
To further the support of individuals having a tendency towards
satisfaction, the biodata questionnaire could be broadened to include
general life satisfaction items (satisfaction with life in general, spousal
or peer relationships, television programs, etc.). The generalized
questionnaire could then be used to predict situation specific
satisfaction (ie.; job satisfaction), adding stronger support to the
dispositional argument than the present study (specific items to
Finally, future work should also consider the disposition/
environment interaction in the prediction of job satisfaction. Theorists
have proposed the importance of considering both disposition and
environment when studying job satisfaction (Cropanzano & James,
1990). The present study appears to account for a degree of
dispositional satisfaction, however, cannot account for environmental
variability or the interaction of the two. It would be of interest to
investigate this relationship to determine if knowledge of both
environment and pre-disposition could increase the predictability of
Because satisfaction is so important to individual well-being
(Spector & Jex, 1991; Tait et al., 1989) and organizational interests
(Cropanzano et aU 1991; George, 1989), efforts made to alter
satisfaction should be made as effective as possible (Bouchard et al.,
1992; Staw & Ross, 1985). To do this, it is becoming evident that
organizations must consider individual pre-disposition as well as
situational influences. The ability to predict satisfaction will assist in
assessing the dispositional factors involved. It is hoped that the
findings from this stud/ will be instrumental in furthering research in
this important area.
BIODATA QUESTIONNAIRE SUB-SCALES
In the past I have enjoyed school.
In the past, I felt satisfied with my performance in school.
I have dropped classes in the past because I was dissatisfied with the class.
I've enjoyed the classes I've taken in college.
I have been more satisfied with past classes than other students.
I feel like college is something to be endured.
I was satisfied with the classes I took last semester.
I don't mind making sacrifices to meet the goals I have set.
In past college classes, I received the grade I deserved.
In the past, I have studied very hard but still not done well in a college class.
Previous classes I have taken have had unfair tests.
I've found classes satisfying even if I didn't do as well as I would have liked.
In the past, classes that I've really liked I've also done well in.
I feel that previous professors I've had in school recognized a job well done:
I am well liked by professors:
Professors I'Ve taken don't care how well I do in their class:
QUALITY OF INSTRUCTION
In past college classes I've taken, I feel that I learned a lot.
Professors I have taken classes from have given me extra help when I have needed it.
I have taken classes from professors who I would never take a class from again.
Classes I've taken in the past have allowed me to demonstrate my knowledge in a
variety of ways (eg., taking tests, writing papers, discussions in class, giving
In other classes I've taken, the professor has used a variety of ways to present new
materials (eg., lectures, assigned readings, demonstrations, hands-on learning, etc.).
In the past, my professors have found novel ways of presenting difficult information.
The classes I've taken in the past have been a waste of time.
The classes I've had will be useful to me in the future.
The skills I've learned in class will have an impact on others.
Classes I've taken will prepare me for other studies.
Classes I've taken will prepare me for my career.
The classes lYe taken in the past have built on the information I've learned in other
In the past, professors have been able to cover all of the information we needed to
know for a class.
INTRINSIC MOTIVATION (continued)
Past classes I've had complement each other in the skills I Ve learned.
In past classes, I've had a say in deciding the topics I did projects on.
Past classes have offered me the freedom and independence to determine how I
accomplished the work required.
I've had control over the pace of my work in other classes I've taken.
I've known, as I've studied, whether I was really learning the information I was studying.
I've known, as I studied, how well I would do on the exam.
I've known, as I studied, how well I would do in the class.
(Please note, the following questionnaire presents items without individual rating
scales because of space considerations. A scale from 1 to 5 was present for each item
during administration of the questionnaire.)
The scale below asks questions about how you feel about your experience in college
so far. The questions refer to classes you've had in general. To answer the
questions, use the scale provided and please circle the number that best indicates
your position on eabh item.
1 2 3 4 5
almost rarely sometimes often almost
In the past, I felt satisfied with my performance in school:
I've known, as I've studied, whether I was really learning the information I was studying:
I've enjoyed the classes I've taken in college:
Professors I have taken classes from have given me extra help when I have needed it:
I've stayed up too late the night before an exam and not done well because of it:
In past college classes, I received the grade I deserved:
In the past, classes that I've really liked I've also done well in:
The classrooms in which Ive taken previous classes have been conducive to learning:
In other classes I've, taken, the professor has used a variety of ways to present new
materials (eg:, lectures, assigned readings, demonstrations, hands-on learning, etc.):
I have made friends'in classes I've taken:
I get down on myself after I do poorly on a test:
No matter how much sleep I've had the night before, I'm alert in class:
I've had control over the pace of my work in other classes I've taken:
I feel that previous professors I've had in school recognized a job well done:
I have been more satisfied with past classes than other students:
I am well liked by professors:
In past college classes I've taken, I feel that I learned a lot:
I get irritated when it's unclear as to where a lecture is leading:
In the past, professors have been able to cover all of the information we needed
to know for a class:
I have taken classes from professors who I would never take a class from again:
Classes I've taken in the past have allowed me to demonstrate my knowledge in a
variety of ways (eg., tests, papers, discussions, presentations, etc.):
School would be more satisfying if the classrooms were more comfortable:
Past classes I've had complement each other in the skills I've learned:
The classes I've taken in the past have been a waste of time:
I've known, as I studied, how well I would do on the exam:
I've found the courses I've had to be very interesting:
The skills I've learned in class will have an impact on others:
I feel like college is something to be endured:
In the past, I have studied very hard but still not done well in a college class:
Classes I've taken will prepare me for my career:
The classes I've taken have built on the information I've learned in other classes:
I've found classes satisfying even if I didn't do as well as I would have liked:
In the past, my professors have found novel ways of presenting difficult information:
I don't mind making sacrifices to meet the goals I have set:
I tend to study the least for my most difficult exams:
My mind has strayed during classes I've had:
I met interesting people in the classes I have taken:
In past classes, I've had a say in deciding the topics I did projects on:
Past classes have!! offered me the freedom and independence to determine how
I accomplished the work required:
I enjoyed working with the people in the group projects I've worked on:
i was satisfied withjthe classes I took last semester:
Professors I've taken don't care how well I do in their class:
I've known, as I studied, how well I would do in the class:
The classes I've had will be useful to me in the future:
In the past I've fallen asleep during class:
Previous classes I have taken have had unfair tests:
College courses have been so interesting I've done further woik on the subject:
I have dropped classes in the past because I was dissatisfied with the class:
Classes I've taken wilt prepare me for other studies:
I've been frustrated by previous professors from whom I've taken classes:
I don't mind waiting in lines:
In the past I have enjoyed school:
In the past, I've registered for more classes than I should have:
I find learning in and of itself interesting:
COLLEGE STUDENT SATISFACTION QUESTIONNAIRE
(Please note, the following questionnaire presents items without individual rating
scales because of space considerations. A scale from 1 to 5 was present for each item
during administration of the questionnaire.)
Below are questions pertaining to your satisfaction with your college experience as a
whole. Using the scale provided, please circle the response for each item that best
reflects your present satisfaction in that area.
1 2 3 4 5
Very Somewhat Very
Dissatisfied Satisfied Satisfied
The chance to take courses that fulfill your goals for personal growth:
The chance to prepare well for your Career:
Your opportunity here to determine your own pattern of intellectual development:
The practice you get in thinking and reasoning:
The quality of education students get here:
The preparation students are getting for their future careers:
The appropriateness of the requirements for your major:
The amount of time you must spend studying:
The difficulty of mofet courses:
The amount of work required in most classes:
Teacher's expectations as to the amount that students should study:
The amount of study it takes to get a passing grade:
The pressure to study:
The chances for men and women to get acquainted:
The choice of dates here:
The chance of having a date here:
The activities that are provided to help you meet someone you might like to date:
The chance to work on projects with members of the opposite sex:
The social events provided for students here:
The ability of most advisors to help students develop their course plans:
The interest that advisors take in the progress of their students:
The availability of your advisor when you need him:
The counselling that is provided for students here:
The amount of personal attention students get from teachers:
The friendliness of most faculty members:
The willingness of teachers to talk to students outside of class time:
The chance to participate in making decisions about school regulations:
The extent that student opinions influence important decisions about the school:
The chance to tell the administration what changes you think are needed in
the coursework here:
The chance for informal contacts between teachers and students outside of class:
The respect that is shown for the ideas of students:
The availability of comfortable places to lounge:
The places you can go just to rest during the day:
The places provided for students to relax between classes:
The concern here for the comfort of students outside of classes:
The chance of getting a comfortable place to live:
The noise level at home when you are trying to study:
The cleanliness of the housing that is available for students:
The availability of quiet study areas for students:
The availability of good places to study:
The chance to live where you want to:
POSITIVE AND NEGATIVE AFFECT SCHEDULE (PANAS)
This scale consists of a number of words that describe different feelings and
emotions. Read each item and then mark the appropriate answer in the space next to
that word. Indicate to what extent you generally feel this way, that is, how you feel on
the average. Use the following scale to record your answers.
1 2 very slightly a tittle or not at all 3 4 5 moderately quite a bit extremely
Below are statements pertaining to how well you believe you will perform in this class.
Using the scale provided, circle one number for each item indicating how confident
you are that you wijl achieve that grade.
0 10 20 30 40 50 60 70 80 90 100
Not At All Somewhat Very
Confident Confident Confident
1) I believe I will get at least a C in this course:
0 10 20 30 40 50 60 70 80 90 100
Not At All Somewhat Very
Confident Confident Confident
2) I believe I will get at least a B in this course:
0 10 20 30 40 50 60 70 80 90 100
Not At All Somewhat Very
Confident Confident Confident
3) I believe I will get an A in this course:
0 10 20 30 40 50 60 70 80 90 100
Not At All Confident Somewhat Confident Very Confident
Satisfaction : A tendency towards enjoyment, gratification or contentment.
Disposition: , Prevailing tendency, mood or inclination. Temperamental makeup.
Biodata: A quasi-longitudinal, quantitative self-report method, used to predict future behavior and attitudes, based on past behavior and attitudes.
Aitken, N.D. (1982). College student performance, satisfaction, and
retention. Journal of Higher Education, 53, 32-50.
Alster, N. (1989). What flexible workers can do. Fortune, 119, 4, 62.
Arvey, R.D., Bouchard, T. J., Segal, N. L., & Abraham, L. M. (1989).
Job satisfaction: Environmental and genetic components. Journal
of Appiied Psychology, 74 (2) 187-192.
Astin, A.W. (1977). Four Critical Years. San Francisco: Jossey Bass.
Asher, J. J. (19^2). The biographical item: Can it be improved?
Personnel Psychology, 25, 251-269.
Bartlett, M.S. (1*941). The statistical significance of canonical
correlations.! Biometrika, 32, 29-38
Bean, J.P., & Bradley, K.K. (1986). Untangling the satisfaction
performance relationship for college students. Journal of Higher
Education, 57, 393-412.
Betz, E.L., Klinginsmith, J.E., & Menne, J.W. (1970), The
measurement and analysis of college student satisfaction.
Measurement and Evaluation in Guidance, 4, 110-118.
Bouchard, T.J., Vvey, R.D., Keller, L.M., & Segal, N.L. (1992).
Genetic influences on job satisfaction: A reply to Cropanzano and
James. Journal of Applied Psychology, 77 (1), 89-93.
Buss, D. Ml, & C|raik, K. H. (1983). The act-frequency approach to
personality. Psychological Review, 90, 105-126.
Bycio, P., Alvares, K.M., & Hackett, R.D. (1990). Job performance and
turnover: A review and meta-analysis. Applied Psychology: An
Campbell, J.P. (1976). Motivation theory in Industrial and
Organizational Psychology. In M.D. Dunnette (Ed.), Handbook of
Industrial & Organizational Psychology, (pp. 1297-1343). Chicago:
Chen, P., & Spector, P. (1991). Negativity affectivity as the underlying
cause of correlations between stressors and strains. Journal of
Applied Psychology, 76, (3).
Cohen, P.A. (1982). Synthesizing research results on teacher
evaluation using meta-anaiytic procedures, Paper presented at
the Annual Meeting of the American Psychological Association in
Cohen, P.A. (1987). A critical analysis and reanalysis of the
multisection validity of meta-analysis, Paper presented at the
Annual Meeting of the American Educational Research
Association in Washington, D.C.
Costa, P.T., & McCrae, R.R. (1980). Influences of extraversion and
neurotieism bn subjective well-being: Happy and unhappy people.
Journal of Personality and Social Psychology, 38, 668-678.
Cropanzano, R., & James, K. (1990). Some methodological
considerations for the behavioral analysis of work attitudes.
Journal of Applied Psychology, 75, 433-439.
Cropanzano, R.,; James, K., & Konovsky, M.A. (1991). Dispositional
affectivity, work attitudes, and performance. Paper presented at
the Sixth Annual Conference of the Society of Industrial
Psychology in St. Louis, Missouri.
Edwards, J.E., & Waters, L.K. (1982). Involvement, ability,
performance jand satisfaction as predictors of college attrition.
Educational Psychological Measurement, 42, 1149-1152.
George, J.M. (1989). Mood and absence. Journal of Applied
Psychology, 74, 317-324.
Gerhart, B. (1987). How important are dispositional factors as
determinants of job satisfaction? Implications for job design and
other personnel programs. Journal of Applied Psychology, 72,
Gerhart, B: (1990). What is the practical relevance of dispositional
effects on job satisfaction? Working paper #90-06 (Revised).
Griffin, R.W. (1988). Consequences of quality circles in an industrial
setting: A Ibingitudinal assessment. Academy of Management
Journal, 31, 338-358.
Hackett, G., & Betz, N.E. (1989). An exploration of the mathematics
self-efficacy/mathematics performance correspondence. Journal
for Research in Mathematics Education, 20 (3), 261-273.
Hackman, d. R., & Oldham, G. R. (1976). Motivation through the
design of wo;rk: Test of a theory. Organizational Behavior and
Human Performance, 16, 250-279.
Hall, D.T. (1986). An overview of current career development theory
research, and practice. In D. Hall and Associates (eds). Career
Development in Organizations, (pp 1-20). San Francisco: Jossey-
Hall, D. T., & Richter, J. (1990). Career gridlock: Baby
boomers hit the wall. The Executive, 4 (3), 7-22.
Hammer, E. G., & Kleiman, L. S. (1988). Getting to know you.
Personnel Administrator, 86-92.
Herzberg, F, Mausner, B., Peterson, R.O., & Capwell, D.F. (1957). Job
Attitudes: Review of Research and Opinion. Pittsburgh:
Hoiberg, A;, & Pugh, W. M. (1978). Predicting Navy effectiveness:
Expectations, motivation, personality, aptitude, and background
variables. Personnel Psychology, 31, 841-852.
Hoppock, R. (1935). Job satisfaction. New York: Harper.
Hough, L. M. (1984). Development and evaluation of the
"Accomplishment Record" method of selecting and promoting
professionals. Journal of Applied Psychology, 69(1), 135-146.
Hudson, J.C. (1$89). Expected grade correlates with evaluation of
teaching. Journalism Educator, 44 (2), 38-44.
Judge, T.A., & Hulin, C.L. (1991). Job satisfaction as a reflection of
disposition: A multiple source causal analysis. Paper presented at
the Sixth Annual Conference of the Society of Industrial
Psychology in St. Louis, Missouri.
Kraiger, K.* Billings, R.S., Isen, A.M. (1989). The influence of positive
affective states on task perceptions and satisfaction.
Organizational Behavior and Human Decision Processes, 44,12-
Lawler, E.E. Ill (1982). Increasing worker involvement to enhance
organizational effectiveness. In P. Goodman (Ed.), Change in
Organizations, (pp. 280-315). San Francisco: Jossey-Bass.
Locke, E.A. (1976). The nature and cause of job satisfaction. In M.D.
Dunnette (Ed.), Handbook of Industrial & Organizational
Psychology, ;(pp. 1297-1343). Chicago: Rand McNally.
Maslow, A.H. (1970). Motivation and personality (2nd ed ). New
York: Harper & Row.
Mumford, M. D., & Owens, W. A. (1982). Life history and vocational
interests. Journal of Vocational Behavior, 21, 330-348.
Mumford, M. D.-, & Owens, W. A. (1987). Methodology review:
Principles, procedures, and findings in the application of
background data measures. Applied Psychological Measurement,
Munro, B.H. (1^81). Dropouts from higher education: Path analysis
of a national sample. American Educational Research Journal,
Munsterberg, HI (1913). Psychology and Industrial Efficiency.
Boston: Houghton Mifflin.
Noel, L. (1985). Increasing student retention: New challenges and
potential. In Noel, L., Levitz, R., & Saluri, D. (Eds), Increasing
Student Retention (pp. 1-27), San Francisco: Jossey-Bass.
O'Reilly, C.A., Parlette, G.N., & Bloom, J.R. (1980). Perceptual
measures of task characteristics: The biasing effects of differing
frames of reference and job attitudes. Academy of Management
Journal, 23, 118-131.
Offerman, L. R.J& Gowing, M. K. (1990). Organizations of the future:
Changes and challenges. American Psychologist, 45, 95-108.
Okun, M.A., & Weir, R.M. (1990). Toward a judgement model of
classroom satisfaction. Educational Psychology Review, 2 (1), SO-
Pascarella, E.T., & Chapman, D.W. (1983). A multiinstitutional, path
analytic validation of Tinto's model of college withdrawal.
American Educational Research Journal, 20 (1), 87-102.
Pulakos, E.D., & Schmitt, N. (1983). A longitudinal study of a valence
model approach for the prediction of job satisfaction of new
employees. Journal of Applied Psychology, 68, 307-312.
Reilly, R. R., & Chao, G. T. (1982). Validity and fairness of some
alternative employee selection procedures. Personnel
Salancik, G. R., & Pfeffer, J. (1977). An examination of need
satisfaction models of job attitudes. Administrative Sciences
Quarterly, 22, 427-456.
Schmitt, N., & Bedian, A.G. (1982). A comparison of LISREL and two
stage least squares analysis of hypothesized life-job satisfaction
relationship.1 Journal of Applied Psychology, 67, 806-817.
Schneider, B. (1983). Interactional psychology and organizational
behavior. Ini L. Cummings & B. Staw (Eds.), Research in
Organizational Behavior, (Vo15, pp.1-31). Greenwich, CT: JAI
Schneider, B., & Dachler, P.H. (1978). A note on the stability of the
job description index. Journal of Applied Psychology, 63, 650-
Spector, P.E. (1988). Development of work locus of control scale.
Journal of Occupational Psychology, 61 (4), 436-457.
Spector, P.E., & Jex, S.M. (1991). Relations of job characteristics from
multiple data sources with employee affect, absence, turnover
intentions, and health. Journal of Applied Psychology, 76 (1), 28-
Staw, B. M., Bel), N. E., & Clausen, J. A. (1986). The dispositional
approach to job attitudes: A lifetime longitudinal test.
Administrative Science Quarterly, 31, 56-77.
Staw, B. M., & Ross, J. (1985). Stability in the midst of change: A
dispositional iapproach to job attitudes. Journal of Applied
Psychology, 70 (3), 469-480.
Strong, S.R. (1978). Critique of the college student satisfaction
questionnaire. In O.K. Buros, (Ed.), The Eighth Mental
Measurements Yearbook, (pp.524-525). Highland Park: Gryphon
Tait, M., Padgett M. Y., & Baldwin T. T. (1989) Job and life satisfaction:
A reevaluation of the strength of the relationship and gender
effects as a function of the date of the study. Journal of Applied
Terenzini, P.T., ;& Pascarella, E.T. (1980). Voluntary freshman attrition
and patterns of social and academic integration in a university: A
test of a conceptual model. Research in Higher Education, 6 (1),
Watson, D., Clark, L., & Tellegen, A. (1988). Development and
validation of brief measures of positive and negative affect: The
PANAS scales. Journal of Personality and Social Psychology, 54
Watson, D;, & Tellegen, A., (1985). Toward a consensual structure of
mood. Psychological Bulletin, 98, 219-235.
Weiss, H.M., & Adler, S. (1984). Personality and organizational
behavior. In L. Cummings & B. Staw (Eds.), Research in
Organizational Behavior, (Vol 6, pp.1-50). Greenwich, CT: JAI
Weiss, D.J., Dawis, R.V., England, G.W., & Loquist, L.H. (1967).
Manual for the Minnesota Satisfaction Questionnaire. Minnesota
Studies in Vocational Rehabilitation, 22.
Zalesny, M;D., 8j Ford, J.K. (1990). Extending the social information
processing perspective: New links to attitudes, behaviors, and
perceptions. Organizational Behavior and Human Decision
Process, 47, 205-246.