Please use these tree papers and write the paper , expecting a paper of 8 pages excluding title and references
Michelle_Michy
20200822164045rural_urban_student_difference1.pdf
INFORMATION TO USERS
This manuscript has been reproduced from the microfilm master. UMI
films the text directly from the original or copy submitted. Thus, some
thesis and dissertation copies are in typewriter face, while others may be
from any type o f computer printer.
The quality o f this reproduction is dependent upon the quality of the
copy submitted. Broken or indistinct print, colored or poor quality
illustrations and photographs, print bleedthrough, substandard margins,
and improper alignment can adversely affect reproduction.
In the unlikely event that the author did not send UMI a complete
manuscript and there are missing pages, these will be noted. Also, if
unauthorized copyright material had to be removed, a note will indicate
the deletion.
Oversize materials (e.g., maps, drawings, charts) are reproduced by
sectioning the original, beginning at the upper left-hand com er and
continuing from left to right in equal sections with small overlaps. Each
original is also photographed in one exposure and is included in reduced
form at the back o f the book.
Photographs included in the original manuscript have been reproduced
xerographically in this copy. Higher quality 6” x 9” black and white
photographic prints are available for any photographs or illustrations
appearing in this copy for an additional charge. Contact UMI directly to
order.
UMI A Bell & Howell Information Company
300 North' Zed) Road, Ann Arbor MI 48106-1346 USA 313/761-4700 800/521-0600
Reproduced with permission of the copyright owner. Further reproduction prohibited w ithout permission.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
RURAL-URBAN STUDENT DIFFERENCES ON THE STRONG INTEREST INVENTORY
FOR A CAREER COUNSELING CENTER SAMPLE
by
Susan L. Pauly Master o f Arts, University o f North Dakota, 1992
A Dissertation
Submitted to the Graduate Faculty
o f the University o f North Dakota
in partial fulfillment o f the requirements
for the degree o f
Doctor o f Philosophy
Grand Forks, North Dakota December
1996
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
UMI Number: 9721216
UMI Microform 9721216 Copyright 1997, by UMI Company. All rights reserved.
This microform edition is protected against unauthorized copying under Title 17, United States Code.
UMI 300 North Zeeb Road Ann Arbor, MI 48103
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
This dissertation, submitted by Susan L. Pauly in partial fulfillment o f the requirements for the Degree o f Doctorate o f Philosophy from the University o f North Dakota, has been read by the Faculty Advisory Committee under whom the work has been done and is hereby approved.
(Chairperson)
This dissertation meets the standards for appearance, conforms to the style and format requirements o f the Graduate School o f the University o f North Dakota, and is hereby approved.
4 / j j u a J $ ________ Dean o f the Graduate School
Date
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
PERMISSION
Title Rural-Urban Student Differences on the Strong Interest Inventory For a Career Counseling Center Sample
Department Counseling
Degree Doctor o f Philosophy
In presenting this dissertation in partial fulfillment o f the requirements for a graduate degree from the University o f North Dakota, I agree that the library of this University shall make it freely available for inspection. I further agree that permission for extensive copying for scholarly purposes may be granted by the professor who supervised my dissertation work, or in her absence, by the chairperson o f the department or the dean o f the Graduate School. It is understood that any copying or publication or other use o f this dissertation or part thereof for financial gain shall not be allowed without my written permission. It is also understood that due recognition shall be given me and to the University o f North Dakota in any scholarly use which may be made o f any material in my dissertation.
Signature
Date
111
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
TABLE OF CONTENTS
LIST OF TABLES..................................................................................................................... v
ACKNOWLEDGMENTS........................................................................................................ vi
ABSTRACT.............................................................................................................................vii
CHAPTER
I. INTRODUCTION....................................................................................................... I
II. REVIEW OF THE LITERATURE...........................................................................3
III. METHOD.................................................................................................................. 47
IV. RESULTS..................................................................................................................56
V. DISCUSSION........................................................................................................... 64
REFERENCES........................................................................................................................ 71
iv
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
LIST OF TABLES
Table Page
1. Demographic Characteristics o f the Sample.............................................................49
2. Iachan Congruence Index Calculation Codes...........................................................52
3. Sample Means and Standard Deviations on SII Scales and for ACT, GPA, Iachan Index Scores, and Profile Differentiation Scores................... 57
4. Summary o f Hierarchical Regression Analysis for Variables Predicting GPA- Equation 1...................................................................................................................59
5. Summary o f Hierarchical Regression Analysis for Variables Predicting GPA- Equation 2...................................................................................................................60
6. Analysis o f Variance for the Iachan o f Congruence by Gender and Rural /Urban S tatus...............................................................................62
7. Results o f Independent t-tests for z-indifference Scores......................................... 63
v
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
ACKNOWLEDGMENTS
I would like to express my sincere appreciation to the members o f my dissertation
committee for their helpful feedback and flexibility. I am especially grateful to Denise
Twohey, Ed.D., my chairperson and mentor, who provided editorial advice as well as
emotional support throughout my graduate program. George Henly, Ph.D. deserves
special recognition for his continued guidance and practical advice. I would also like to
thank my colleague and friend, Michael Ewing, Ph.D. for his assistance on the statistical
portions o f this dissertation. I would also like to express my sincere thanks to Drs.
Deborah Betsworth and Richard Grosz o f the UND Counseling Center for granting my
access to the archival career counseling data. In addition, Dr. Tim Driscoll, Division o f
Student Affairs Research and Evaluation, provided important support for this project.
I would also like to express my deepest gratitude to my husband, Richard L. Pauly,
and my children, Julie, Jason, and Matthew. Without their cooperation and support, I
would not have been able to do any o f this.
vi
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
ABSTRACT
The purpose o f this study was to explore whether differences existed between rural
and urban students’ responses on the Strong Interest Inventory (SII). The subjects for this
study were 665 students who had taken the SII as part o f career counseling services
received at the University o f North Dakota. The subjects were classified as either rural or
urban based on the population o f their hometown. The rural and urban groups’ scores
were compared on the SII General Occupational Theme (RIASEC), Academic Comfort
and Introversion/Extroversion scales, the Iachan index o f congruence between RIASEC
scores and college major, measures o f indifference in responding, and profile
differentiation. No mean differences between rural and urban subjects were detected on
any o f the comparisons. The study also examined whether cumulative grade point
average (GPA) could be predicted by factors including Academic Comfort, gender, being
from a rural/urban environment, and ACT composite scores. Results indicated that when
ACT scores are included in the prediction equation, the effects o f rural/urban disappear.
Rural-urban status and gender were not found to moderate the relationship between
Academic Comfort scores and GPA.
vii
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 1
Introduction
As part o f graduate training, I was privileged to work as an individual career
counselor at the University o f North Dakota Counseling Center. In that role, I worked
with many career clients and interpreted a number o f Strong Interest Inventories (SII).
Over the course o f this time, I began to notice some trends in the students' scores on the
SII which are o f interest to me, and I believe, also, to the psychological community.
While conducting vocational interviews and interpreting SII's, I observed that students
from rural areas tended to have lower scores on the SII Academic Comfort (AC) scale
than did those from larger, urban areas. Also, I noticed that rural students tended to
respond with a higher percentage o f "dislike" than either "indifferent" or "like" responses
to the SO items, especially items in the areas o f occupations and school subjects. This
resulted in SII profiles with low scores on the General Occupational Themes (GOTs) and
no significant differences between the six GOT scores. For interpretative purposes, this
low, undifferentiated profile is problematic because it means that the test-taker’s
personality pattern, as identified by their three highest GOT scores, cannot be clearly
delineated. Therefore, the number o f occupations that match the person’s pattern is
reduced. Stated in practical terms, this means that when a student takes the SII to help
them identify careers which match with their interests, and the end result is a low,
1
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2
undifferentiated profile, the student is often disappointed to find few career matches for
consideration.
There are a number o f reasons why rural students obtain flat SII profiles. Perhaps
rural students genuinely dislike many o f the occupations and school subjects represented
in the SII questions, and the profile correctly reflects their lack o f interests. Alternatively,
rural settings may not give students enough exposure to different occupations, school
subjects and leisure activities to permit them to respond knowledgeably to many o f the
items. In the latter instance, the SII must be interpreted differently for rural students,
something o f which career counselors should be made aware.
The SO is one o f the most widely used vocational assessment instruments (Hansen,
1992). Its applicability has been studied for many populations. One for which it has not
been studied is the rural population. No references were found for any research exploring
the use o f the SII for rural individuals. Therefore, it is the intent o f this study to explore
whether rural and urban students respond differently on the SII.
Reproduced with permission of the copyright owner. Further reproduction prohibited w ithout permission.
CHAPTER 2
Review o f the Literature
The review begins with a definition o f interests and their relation to occupational
choice. Next, theories about the development o f occupational interests and stereotypes
are considered, as these are the foundation upon which career interest inventories rest.
The history and structure o f the SII, and its use with college populations, is discussed
next. Because those from rural populations may be considered as a minority group,
literature on the use of the SII with minority and ethnic groups is also reviewed. Finally,
a review o f research on career concerns for rural populations is presented.
Definition o f Occupational Interests
In order to understand how vocational interests develop, it is important to have some
understanding about what the word “interest” means in relation to occupational choice.
Super (1957) delineated four types o f interests. He defined “expressed interests” as
expressions o f preference for a vocation or activity. The SII, then, measures expressed
interests because it requires test-takers to indicate their preferences for school subjects,
leisure activities, and occupations. Super (1957) also believed that interests could be
inferred from our actions as evidenced by participation in activities. These interests he
called “manifest interests”. He also stated that interests could be measured in two ways.
When a scoring algorithm is used to estimate interests based on individual responses
3
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
4
regarding the person’s likes, dislikes and preferences for one activity versus another
(expressed interests), Super called them “inventoried interests.” Therefore, Super would
suggest that the interests measured by scores on the SII scales are inventoried interests.
He also believed that manifest interests could be measured under controlled conditions,
such as the amount o f time spent examining resources for specific careers. He called
these interests “tested interests”. Super (1957) further stated that there are six categories
within each o f these types o f interests. He classified them as: (a) scientific, technical or
material; (b) humanistic or social welfare; (c) systematic or business detail; (d) business
contact; (e) literary; and (f) musical/artistic.
Dawis and Loftquist (1984) defined interests in a manner similar to Super (1957).
They believed that interests stem from underlying values and abilities. They suggested
that there is a complex relationship between our values and abilities and that we strive to
express this relationship in a variety o f ways. One expression o f the relationship o f
values and abilities is through the statement o f preferences for various activities. For
example, an individual might say, “I like to read books.” According to Dawis and
Loftquist (1984), this preference is termed an “expressed or stated interest.” At times, we
may want to create an instrument which inquires about preferences for a comprehensive
sample o f activities that have been experienced by a majority o f people. When an
individual responds to such an instrument (e.g., the SII), their scored responses are termed
“measured interests” (p. 18, Dawis & Loftquist, 1984). Dawis and Loftquist (1984) also
referred to preferences for activities which may be observed by others through our
participation in the activity. They call this “exhibited interest” and state that congruence
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
5
between these exhibited interests and measured interest is called “validated interest’* (p.
19, Dawis & Loftquist, 1984).
Although Holland’s (1985) description o f interests is similar to those o f Super (1957)
and Dawis and Loftquist (1984), his definition places more emphasis on interests as a
manifestation o f personality. He stated that preferences for activities are based on an
interplay o f personal and environmental forces such as heredity, peers, family, social
class, culture, and the physical environment. These activity preferences develop into
specific interests which, over time, cause a person to develop special competencies.
Holland (1985) believed that the combination o f interests and competencies creates a
personal disposition which leads an individual to think and respond in characteristic
ways. He called this their personality type. The six pure personality types Holland
defined (Realistic, Investigative, Artistic, Social, Enterprising and Conventional) are
remarkably similar to the interest classifications noted by Super (1957). Therefore.
Holland (1985) saw interests and personality type as essentially equivalent constructs.
Knowledge o f how interests are defined and how they relate to personality is important
to understanding the structure and meaning o f the SII. Essentially, the instrument is
based on the assumption that an individual has preferences or expressed interests in
activities, school subjects, working with certain types o f people and in certain
occupations. The test is also based on the idea that these expressed interests, when
compiled permit inferences about basic interest or personality types which can be used to
classify both the individual expressing them and the environment in which the individual
works. Ultimately, the scores obtained from the SII are inventoried interests that measure
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
6
what Holland calls personality type. These scores are then used to help match the person
with congruent occupations. Essentially, then, the ideas o f Super (1957), Holland (1973.
1985) and Dawis and Loftquist (1984) help explain how the SII works, and why it is
successful.
Models o f Vocational Interest Development
In addition to understanding how interests are defined, it is also important to
understand how career interests develop. Many theories have emerged about career and
interest development. The major theories for consideration may be classified as
developmental, personality, social learning, and valence instrumentality models.
Developmental Career Theories. Super’s (1953, 1957, 1980) theory is called
developmental because it considers career decision making across the life-span. Super
(1957) based his theory on twelve propositions. In the first three, he stated that the
process o f vocational development was ongoing, irreversible, orderly, predictable, and
dynamic. Super (1980) specified the orderly stages o f the process as growth, exploration,
establishment, maintenance, and decline. He described the growth stage as occurring
until about age fifteen and one in which the focus is on physical and psychological
development as well as acquiring experiences which will give background knowledge
about the environment, including the world o f work (Super, 1957). The exploration stage
may last until the individual is around twenty-five years o f age and starts with the
realization that work is a part o f life. In this stage, many choices are fantasized about, but
the range o f choices is eventually narrowed down to only those realistically attainable.
The establishment phase is typified by initial work experiences and attempts to access
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
7
earlier vocational decisions. During the maintenance phase, which lasts until about age
sixty-five, the individual continues to enhance the vocational situation by adjusting both
pleasant and unpleasant aspects o f the job. The declining phase begins prior to retirement
and focuses on meeting work standards and simply retaining the job till retirement
(Super, 1957).
In proposition four, Super (1957) introduced the notion o f self-concept formation.
Definition of a clear self concept is the primary task in career development. It is the self
concept which must be translated into occupational terms before it can be implemented in
a career choice. This idea is considered to be the central tenet o f Super’s developmental
theory. In propositions five and six, Super (1957) suggested that personal and societal
reality factors, including the parental relationship, have increasingly profound effects on
occupational choice. Super (1957) also stated in propositions seven and eight that
vocational choice is affected by intelligence, SES, needs for status, values, interests,
interpersonal skills, education, economic supply and demand, role models, and available
community resources. In proposition nine, Super (1957) spoke to the work environment
requirements for specific skills, abilities, and interests. Propositions ten and eleven
referred to work satisfaction as a function o f the congruence between the individual’s
interests, abilities, values and personality traits and those provided by the work
environment. Furthermore, the degree o f satisfaction is measured by the degree to which
self-concept is implemented in the workplace. Finally, in proposition twelve, Super
(1957) suggested that work is an expression o f personality.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
8
Super (1980) also believed that across the life-span, there are a variety o f roles that an
individual plays. He described the life roles as child, student, leisurite, citizen, worker,
spouse, homemaker, parent and pensioner. He believed these roles are carried out in four
theaters which he labeled, the home, the community, the school, and the workplace. He
integrated the concepts o f life stages, roles, and theaters in a graphic representation he
called the Life-Career Rainbow (Super, 1980). He added to this the concept o f decision
points which he believed occurred at times o f transition between roles. At these decision
points, choices are affected by the interaction o f all personal and environmental forces
(Super, 1980). Ultimately, Super (1957) viewed the development o f occupational
interests as an interplay between personal and environmental determinants, and the
developmental goal as the implementation o f the self-concept.
Gottfredson (1981) also proposed a developmental theory o f career choice. She
believed this development is highly influenced by social environment as well as cognitive
development. According to Gottfredson, the most relevant influences on career
development are gender, social class, intelligence, values, competencies and interests.
Her theory stated that these elements become a part o f individual self-concept at different
stages o f cognitive development.
Gottfredson’s (1981) four stage theory begins with the “orientation to size and power”
which occurs at ages three to five. In this stage, children are first able to understand the
concept o f being an adult someday. In the next stage, “orientation to sex roles,” children
confirm the concept o f gender. This stage occurs at about six to eight years o f age. At
age nine to thirteen, children enter the stage o f “orientation to social valuation” where the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
9
concepts o f social class and ability become important influences on social expectations
and behavior. The final stage o f development is called “orientation to the internal, unique
s e lf ’ and typically begins at about age thirteen. Here, adolescents are increasingly able to
deal with emotional stress as well as more abstract and complex cognitive tasks. Also,
teens in this stage are more aware o f their own inner feelings and personal capabilities
(Gottfredson, 1981).
In practical terms, Gottfredson (1981) postulated the following developmental
progression. The preschool child has a fairly positive view o f the occupations in his/her
awareness; this view is modified first by ruling out those jobs which are incongruent with
gender role, and later by eliminating occupations which do not meet requirements for
status and prestige or that exceed requirements for individual effort. Finally, in
adolescence, individuals begin to consider personal capabilities, interests and values as
they further restrict the range o f career choices. This narrowing o f choices is also
affected by environmental opportunities and barriers. Gottfredson (1981) suggested that
the choice o f career is a compromise between the fulfillment o f adolescent dreams and
the reality o f barriers to employment. She maintained that the pattern o f compromise will
begin with sacrificing vocational interests first, followed by job level, and then sex type.
Poole and Cooney (1985) proposed a model o f career decision-making based on
personal possibility theory. While their theory is not strictly a developmental theory, it
contains elements similar to Super’s notion o f self concept and Gottfredson’s ideas about
concepts related to career choice. They argued that at each stage o f individual
development, people are faced with a huge range o f choice possibilities. Internal
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
10
cognitive structures then affect the possibilities by screening, organizing and creating
experiences (Poole & Cooney, 1985). Therefore, to exercise choice, a person must have
an awareness o f available options. The researchers then postulated that occupational
perceptions may be influenced by their environments (Poole & Cooney, 1985).
Specifically, they examined whether sex, social class and ethnicity affected the awareness
o f occupational possibilities and found evidence to support the influences o f each o f these
factors.
Holland’s Personality Theory o f Career Development. Holland (1973) stated that
personality types develop as a result o f a variety o f forces including genetic, cultural,
personal, and environmental. Specifically, Holland (1985, p. 16) suggested that “types
produce types” meaning that individuals become a certain personality type based on
parental influences, physical and psychological factors, as well as the availability o f
environmental opportunities. This combination o f influences leads an individual to
develop likes and dislikes for certain activities and these preferences are continually
evaluated as the individual participates in different settings such as school, college, or
job. The interaction between preferences and environments subsequently works to create
a personality type that exhibits a predictable set o f behaviors and characteristics as well as
skills and coping styles (Holland, 1985).
Holland (1973) believed there are six basic personality types and developed a
hexagonal model describing the relationship between them. The underlying theory for the
model was based on four assumptions. First, people can be described in terms o f their
resemblance to six different types: realistic, investigative, artistic, social, enterprising, and
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
11
conventional. Also, work environments can be classified in terms o f the same six types,
with the corresponding types o f people dominating the environment. Furthermore, people
will seek an environment that is congruent with attitudes, values, skills, and abilities and
which is stimulating and satisfying to them. Finally, a person's behavior is determined by
the interaction between the individual and the environment and this influences such
factors as work stability, job performance, and job satisfaction (Holland, 1973).
Holland’s (1973) model is described as hexagonal in structure based on the similarities o f
the six personality types. This model is discussed in greater detail later in this
dissertation because of its direct relevance to the development o f the SII, the instrument
used in this study.
Gati (1979) developed his hierarchical model in response to concerns he had about
the hexagonal model o f Holland (1973). Although Gati (1979) did not take issue with
Holland’s basic ideas regarding interest development, his model did differ dramatically
from Holland’s in structure. Assumptions o f the circular and hexagonal models are that
the adjacent fields are equidistant, and that certain fields tend to be grouped together.
Gati (1979), however, cited a number o f studies by other researchers which did not
support either premise. Gati acknowledged that there are similarities between adjacent
fields, but argued that their spatial representation in a hexagonal or circular manner
greatly oversimplifies their empirical relationships. Instead, Gati (1979) proposed that
occupations could be best represented as collections o f attributes such as features o f the
work environment or the level o f social relationships on the job. He also stated that the
contrast model (Tversky. 1977 in Gati. 1979) served to explain the similarities and
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
12
differences among the vocations. Based on this model, it is assumed that the similarity
between occupations increases as common features are added or as divergent features are
deleted (Gati, 1979). Finally, Gati (1979) believed that the similar relationships among
occupations could best be represented by a hierarchical tree structure, with a similar
pattern for corresponding vocational interests.
At the highest peak o f the hierarchy rests interests in all occupations (Gati, 1979). The
second level is comprised o f two major groups called interests in the “soft sciences” and
interests in the “hard sciences” (Gati, 1979). The third level is divided into four minor
groups which flow from the major hard sciences/soft sciences groups. Radiating from the
interests in soft sciences are two groups: interests in service, social and cultural
occupations; and interests in business, enterprising, organization, and conventional
occupations. Flowing from the interests in hard sciences are two groups: interests in
technological and realistic occupations; and interests in science and investigative
occupations. The four minor groups flow into fields o f occupations and then into the final
level o f specific occupations (Gati, 1979).
Gati (1979) set out to test his model by reanalyzing the variables presented in the
intercorrelation matrix reported by Lunneborg and Lunneborg (1975, in Gati, 1979). The
analysis Gati (1979) described in the literature is complex. However, the end result is his
conclusion that a hierarchical structure o f interests provides a better representation than
does a circular or hexagonal one. In a practical sense this means that when someone is
attempting to implement a career decision, they first decide on a major field o f hard or
soft science and then select one o f the four minor fields. From there they move to an
Reproduced with permission of the copyright owner. Further reproduction prohibited w ithout permission.
13
occupational field and, thence, to a specific job. If, perchance, the individual does not
like the chosen job, he/she can travel back up the hierarchy to a previous stage and
regroup. However, in a circular or hexagonal model the individual who is unsatisfied
with their career choice must ultimately leave the field for an adjacent field (Gati, 1979).
Theories o f Career Development Based on Social Learning Principles. Mitchell and
Krumboltz (in Brown & Brooks, 1984) based their theory o f career development on
Bandura’s (1977) social learning theory o f behavior. This theory rests on classic
behavioralism and reinforcement theory and states that our individual personalities are
based on our learning experiences. However, social learning theory adds the component
that individuals also use their capabilities to act on the environment around them
(Bandura, 1977).
The social learning theory o f career development (Mitchell & Krumboltz, in Brown &
Brooks, 1984) was developed to address why people choose certain careers, change
careers, and express various occupational preferences. The theory examined the impact
o f genetic endowment; environmental conditions and events; learning experiences and
individual performance on the career decision making process. Mitchell and Krumboltz
included in genetic endowment inherited characteristics such as race, sex, and
appearance. They felt that special abilities were likely to be a product o f genetics as well
as exposure to the environment. Environmental conditions could include social, cultural,
political, economic and natural forces. Examples o f such forces could include natural
disasters, labor laws, availability o f job opportunities, and limits on family resources for
education and training.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
14
Krumboltz and Mitchell also considered the impact o f learning experience on career
decision making. They suggested that with instrumental learning experiences, the
individual acts on the environment to produce certain outcomes. They described
associative learning experiences as ones in which individuals recognized connections
between certain stimuli and the environment. With respect to career decisions, certain
stimuli are viewed as having either positive or negative associations and may, therefore,
influence the decision making process. They also considered the interactions among
learning experiences, environmental factors and genetic endowment in their concept o f
task approach skills. Such skills might include values and standards for performance and
work habits.
Also included in their theory is the concept o f self-observation generalizations (SOG).
These refer to the generalizations people constantly make about their skills and abilities
after observing and evaluating their own performance. Parallel to these are world-view
generalizations, observations made about the environment based on past learning
experiences. These generalizations are then used to make predictions about future events.
Krumboltz and Mitchell (in Brown & Brooks, 1984) stated that the emphasis on the
learning experience as well as the consideration o f interactional forces makes their theory
applicable to both males and females and minority cultures.
Hackett and Betz (1981) also based their theory on the work o f Bandura. They
described a self-efficacy model o f vocational preference in which preference is based, in
part, on self perceptions about the individual’s capability o f performing on the job. This
theory is based on Bandura’s social modeling theory ( in Hackett & Betz, 1981).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
15
Basically, the theory stated that perceptions o f low self-efficacy may inhibit an
individual’s entry into a desired profession even though the occupation is perceived as
providing desired outcomes. Hackett & Betz (1981) provided empirical support for this
premise with respect to the underrepresentation o f women in certain vocations such as
law, medicine, management, and engineering. Essentially, they suggested that although
females have the ability to perform in these types o f professions, their self-efficacy
perceptions are affected by gender-role socialization experiences and the lack o f social
role models for these jobs (Hackett & Betz, 1981).
Lent, Brown, and Hackett (1994) proposed a theory o f career development based on
Bandura’s (1986) general social cognitive theory. Bandura put forth an interactive model
called “triadic reciprocality” which suggests that interaction occurs in a bidirectional
fashion between: (a) personal characteristics such as physical appearance, and cognitive
and affective states; (b) external factors in the environment; and (c) overt behavior.
In addition to the concept o f person-situation interaction (Bandura, 1986), Lent et.al.
(1994) placed emphasis on three social cognitive mechanisms with particular import for
the process o f career development. These components were self-efficacy beliefs,
outcome expectations, and goal representations. Lent et.al. (1994) viewed self-efficacy as
a dynamic trait which allows for adjustment o f one’s personal view o f his/her response
capabilities depending on the particular performance requirements o f the environment.
Outcome expectations referred to personal beliefs about the probable consequences o f
performing a specific behavior (Lent et al., 1994). Also, according to the theory o f Lent
et al. (1994), goals were important determinants o f behavior because they provided
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
16
organization, guidance, and sustenance for behavior even over long time periods and with
little external reinforcement. Therefore, goals can help increase the chances that the
desired outcome will be achieved (Lent et al., 1994).
Lent et al. (1994) made several key assumptions that genetic endowment, special
abilities, environmental conditions and learning (including operant, associative, and
vicarious) serve to influence career development in an interactive fashion. Their theory
essentially took a constructionist approach to career development, one which underscores
the importance o f anticipation, forethought, and the assignment o f meaning in the person-
environment interactional process. They believed that this interactional process is most
active up until late adolescence, at which point interests tend to stabilize. However, they
also noted that change can occur at various developmental points across the lifespan, such
as with the loss o f employment, birth o f a child, decline in health, or innovations in
technology (L en tetal., 1994).
Based on these tenets, Lent et al. (1994) developed a causal model o f person,
contextual and experiential factors affecting occupational choice behavior. They also
developed a causal model o f task performance which highlights the roles ability, self-
efficacy, outcome expectations, and performance goals. They noted that the model can be
adapted to fit the career concerns o f women and minorities by expanding the emphasis on
certain aspects o f the model. For example, extra emphasis might be placed on outcome
expectations which may be affected by “glass ceiling” obstacles or lack o f affirmative
action hiring policies (Lent et al., 1994).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
17
Valence Instrumentality and Subjective Expected Utility Theories o f Career Decision
Making. Mitchell and Beach (1976) described expectancy theory and decision theory as
models that can be used to predict vocational preference and choice. They referred to the
work o f Vroom (1964, in Mitchell & Beach, 1976) who stated that occupational choice
depends on the degree to which a certain choice is viewed as having a greater likelihood
than any other choice to lead to an attractive outcome. Vroom developed models to
predict both valence and force towards behavior. Valence refers to the anticipated
satisfaction which is associated with a particular outcome. Algebraically, it is the sum o f
the products o f the valences o f all other outcomes and the individual’s perception o f the
instrumentality o f the specific outcome for obtaining o f the other outcomes (Vroom, 1964.
in Mitchell & Beach, 1976). Force towards behavior is a function o f the sum o f the
products o f the valences for all the outcomes and the strength o f the individual’s
expectancies that the action will be followed by obtaining the outcomes (Vroom, 1964. in
Mitchell & Beach, 1976). Mitchell and Beach (1976) named these two models the
valence model and the choice model, respectively. They stated that the valence model
has been used to predict job choice and satisfaction while the choice model has been used
to predict job effort.
Mitchell and Beach (1976) also reviewed subjective expected utility decision theory
which is based on the principle o f maximization o f expectation. The theory states that
“the expectation for any action is the algebraic sum, across potential outcomes, o f the
values o f each o f the possible outcomes o f that action and o f their respective probabilities
of occurrence should the action be performed (p. 237. Mitchell & Beach, 1976). In
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
18
practical terras, this means that each individual will consider all possible occupational
alternatives and assess the probability that each one would lead to various outcomes as
well as the value the individual assigns to the outcome. Mitchell and Beach (1976) then
reviewed a number o f research studies which supported the utility o f the aforementioned
models.
Wheeler and Mahoney (1981) also espoused an expectancy model approach to
vocational preference and choice. In this model, a distinction is made between
occupational preference and choice. This distinction is based on attraction, expectancies,
and costs. Occupational valence is an important component o f this theory. Specifically,
valence is a function o f an individual’s attraction to an occupation. It is assumed that
occupations with the most potential for provision o f rewards and important outcomes will
be more attractive and preferred, and thus, have the most positive occupational valence
(Wheeler & Mahoney, 1981).
Vocational choice is also affected by expectancies about entering an occupation
(Wheeler & Mahoney, 1981). These expectancies are colored by real world realities. For
instance, even though an individual may have the ability to pursue a career as a realtor,
the risk o f failure due to reliance on commission as the sole income may lower the
individual’s expectancies for success and serve to prohibit entry into this career. Wheeler
and Mahoney (1981) also stated that preparation costs will influence vocational choice.
They define these costs in terms o f time, effort and financial resources necessary to
prepare for entry into the chosen field. For example, though a career in medicine is
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
19
highly desirable for a person, they may forego it because o f the vast amount o f necessary
schooling and the lack o f financial resources.
The Role o f Occupational Stereotypes in Interest Development. Underlying all the
theories o f career development is the notion that individuals have developed a fund o f
stereotypic knowledge about the vast array o f occupations. Included in this fund o f
knowledge is information about likely occupational outcomes and the probability o f their
attainment, and occupational requirements and the likelihood o f meeting them. These
occupational stereotypes then become a basis for statements o f occupational preference
and choice.
Banducci (1970) set out to explore the relationship between occupational stereotypic
accuracy and variables including socioeconomic status (SES), academic development,
vocational interests, crystallization o f plans, and range o f personal experience. For his
sample o f 679 twelfth-grade boys with a wide range o f SES at three Midwestern high
schools, Banducci (1970) found that SES, academic development, crystallization o f plans,
and vocational interests were related to occupational stereotype accuracy, but that range
o f personal experience had no significant influence on accuracy after controlling for SES
and academic development. Specifically, Banducci (1970) found that there was a
significant positive relationship between academic development and vocational
perceptual accuracy, and that students o f lower SES were less accurate in their vocational
perceptions than higher SES students. To summarize, Banducci's (1970) work supports
the notion that personal developmental variables have an effect on the accuracy o f
occupational stereotypes.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
20
Hollander and Parker (1972) studied occupational stereotypes in relation to self
description. More specifically, they endeavored to test Holland’s (1959) idea that the
implementation o f vocational choice partly involves congruence between self-description
and knowledge o f occupational stereotypes. The subjects in their study were fifty-four
predominantly middle class Caucasian high school sophomores from urban Oklahoma.
The subjects were administered the Adjective Check List (Gough & Heilbrun, 1965 in
Hollander & Parker, 1972) as a measure o f self-description and occupational stereotypes.
They were also given an Occupational Preference List constructed by Hollander & Parker
(1972) to determine vocational preference. The results showed that stereotypes played a
significant role in vocational exploration and choice (Hollander & Parker, 1972). The
results also supported the idea that self-description is related to occupational choice for
adolescents. This is similar to Holland’s (1973) idea that individuals will search for a
work environment that is congruent with their personal characteristics.
Bloch and Rim (1979) also examined occupational stereotypes, and they, too,
suggested that each individual has a system o f implementing vocational choice based on
his/her structure o f occupational stereotypes, and that this structure is influenced by
individual background variables. However, Bloch and Rim (1979) believed that there is a
universal perception o f occupations that is not influenced by individual background
variables such as age, school, SES, intelligence or parental occupation. They designed
research to identify the individual dimensions used for occupational classification as well
as to classify these dimensions into values categories commonly associated with
vocational choice (Bloch & Rim. 1979). The results showed that individual values can be
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
21
organized into a 20-values scheme, and that occupational affinity is not influenced by
individual background variables or the level o f information an individual possesses about
that occupation (Bloch & Rim, 1979). Furthermore, Bloch and Rim (1979) found that
some values are used frequently by the majority o f the population, but that there are
values which are used less often and even some unique values used by select populations.
Finally, their study revealed that boys perceive the world o f work less uniformly than
girls (Bloch & Rim, 1979).
Summary o f the Literature on Career Interests and Theory. In summarizing the
literature on the various theories o f career development and their relationship to
occupational stereotypes and career indecision, it is important to note the similarities and
differences between them. I think it is safe to say that all o f the theories assume that each
individual has a fund o f basic knowledge about themselves and occupations in general.
The developmental, personality and social learning theorists focus more on how interests
develop and their relationship to aspects o f the environment and the self. The valence
instrumentality theorists tend to focus more on how interests are implemented. For the
purpose o f this research study, more emphasis has been placed on the personality theory
o f Holland (1973) since the SII’s General Occupational Themes are based on Holland's
(1973) personality types. However, the explanations about how the SII works as an
instrument to aid in career decision making are very much related to the underlying
notions about how interests are defined by Super (1957), and Dawis and Loftquist (1984).
In this study, it is postulated that rural and urban students may respond differently on the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
22
SII. If this is true, then it is also necessary to begin thinking about how each o f the
theorists might account for this.
History and Structure o f the SII.
The Strong Interest Inventory (SII) is one o f the most widely used career interest
inventories. It is commonly used as an aid in the career decision making process,
especially at college counseling centers. Because o f its popularity, it has also served as
the basis for many research projects. The fourth edition o f the Strong Interest Inventory
(SII) used in this study has evolved from its earliest predecessor, the Strong Interest
Inventory Blank (SVIB) which was first published in 1927 by E. K. Strong (Campbell,
1987; Hansen, 1987). In the nearly 70 years since its inception, the SVIB has undergone
many revisions (Campbell, 1987; Hansen, 1987). Though the SVIB published in 1927
was for men only, Strong later published the first women's form in 1933 (Campbell,
Crichton, Hansen, & Webber, 1974). Although the rationale for creating a separate
women's form was never expressed by Strong or the publisher, Campbell et al. (1974)
surmised that it occurred because o f differences in the available employment
opportunities for men and women as well as gender differences in responses to the items
on the test. Campbell (1974) then undertook the task o f combining the same sex forms as
well as adding the general occupational themes, (based on Holland's theory o f
personality) to the inventory (Holland, 1973). This 1974 merging of forms was in
response to claims that the inventory was gender biased. The theoretical framework o f
Holland was added to assist in the organization and interpretation of the scores and to
address the criticism that the test was atheoretical in nature (Campbell et al.. 1974;
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
23
Anastasi, 1988). At this time, the test became known as the Strong Campbell Interest
Inventory (SCII), Form T325 o f the SVIB (Campbell, 1987).
Since 1974, revisions in 1981 and 1985 expanded the instrument and improved its
format. According to Campbell (1987), the 1985 revision also sought to develop current
reference groups for Women-in-General, Men-in-General, and the Occupational Scales.
Another major goal o f this revision was to expand the usefulness o f the test by including
non-professional occupations in addition to the professions already listed.
The 1985 SII version o f the test used in this study emerged to contain 264 scales
including 6 General Occupational Themes (GOT), 23 Basic Interest Scales (BIS), 207
Occupational Scales, special scales for Academic Comfort (AC) and Introversion-
Extroversion (I/E), and 26 Administrative Indexes (Hansen, 1992).
Presently, the SII contains 325 items about occupations, school subjects, leisure
activities, types o f people, and individual test-taker characteristics. The respondent is
asked to answer the items by choosing from these categories: “ Like,” “ Indifferent,” or
“dislike.” The test is then scored by computer.
Uses o f the SII. The original version o f the SII developed by Strong in 1927 employed
the empirical method o f test construction (Hansen, 1987; Hansen, 1992; Anastasi, 1988).
It was found that people in the same occupation had similar interests, and that these
interests could serve to differentiate people in one occupation from another. Therefore,
the SVIB was developed to compare the similarity o f the test-taker's interest to those of
people employed in various jobs.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
24
In addition to measuring an individual's occupational interests, the current version o f
the SII can also indicate interest in different leisure activities, interest in spending time
with various types o f people, and interest in living or working in selected environments
(Hansen, 1992). And, although the major use o f the instrument is in career counseling, it
can be used in a variety o f ways within that context. For instance, some clients may wish
to expand the repertoire o f occupations they have considered in the past to include new
areas. Other clients may wish to narrow the number o f choices to a select few. Still other
clients may desire to confirm a career choice they have already made (Hansen, 1992).
Regardless o f the client's reason for using the instrument, the SII can assess interests and
help integrate them with the world-of-work. This, o f course, assumes that the client has a
good understanding o f self, interests, and the world-of-work. For some clients (possibly
some o f the rural clients I have worked with), this may not be the case, which is, o f
course, the reason for this study.
Description o f the SII Scales
General Occupational Themes: Campbell and Holland (1972) derived the SVIB
Occupational Theme scales for men and Hansen and Johansson (1972) constructed
similar scales for women, both based on the research o f Holland (1959). Holland
presented his theory in Making Vocational Choices: A Theory o f Careers (19731 As
previously mentioned, a person is classified according to one or more of the six
personality themes o f Realistic, Investigative, Artistic, Social, Enterprising, and
Conventional. Theoretically, there are 720 possible theme combinations (Hansen &
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
25
Campbell, 1985). For practical reasons, most clients are classified by the one to three
types with the highest scores. This is referred to as the personality pattern.
Personality patterns can also be described in terms o f consistency and differentiation.
Consistency refers to whether or not a pattern's related elements have common
characteristics. For example, a pattern o f Realistic-Investigative is considered consistent
because it has common traits such as an orientation towards things rather than people.
Conversely, a pattern o f Conventional-Artistic is considered inconsistent because the
traits are opposites on such factors as desire for control and expressiveness (Holland.
1985). Differentiation in a pattern denotes the numerical difference between a person's
GOT scores. Well differentiated patterns are ones that reflect profiles which resemble
single personality types. Poorly differentiated profiles are called "flat" and represent a
person who equally resembles each theme type and thus cannot be characterized in a
predictable fashion.
Characteristics o f individuals of each o f the six theme types follow. Realistic persons
enjoy practical, concrete activities as opposed to abstract thinking. They tend to be
mechanically inclined, well-coordinated, rugged individuals who enjoy work involving
physical activity. They typically do not enjoy social settings which require them to be
verbally and interpersonally skilled. They prefer occupations such as farmer, fish and
wildlife specialist, mechanic, electrician, and engineer.
Investigative individuals enjoy activities that are scientifically oriented, involve
problem solving, abstract thinking, and intellectual ability. These individuals tend to be
rather unconventional in their values and ideas, and are often introverted and
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
26
independent. They typically enjoy such occupations as college professor, biologist,
chemist, astronomer, or psychologist.
Artistic types prefer unstructured activities that allow freedom o f self-expression and
creativity. Like Investigative types, they are rather unconventional and introspective, but
prefer working with artistic and musical media and value aesthetic quality. Typical job
preferences include artist, musician, writer, photographer, or interior decorator.
Social types are typified by their extroverted, humanistic nature. They enjoy activities
that require social and interpersonal skill. They like group activity, as well as being the
leader o f the group. Examples o f preferred jobs include: teacher, social worker,
counselor, or speech therapist.
Enterprising types are strong leaders and use their verbal and management skills to
dominate, organize, and sell. They prefer activities that do not require sustained periods
o f intellectual inquiry. They are often viewed as popular, powerful, aggressive, social,
and status-seeking. They often choose a job as business executive, real estate
salesperson, insurance agent, or politician.
Conventional types prefer activities that are structured and well-organized. Like
Enterprising types, they identify with those in power and value status and material
possessions, but they typically avoid leadership positions. Instead, they seek work that is
systematic and that allows them to operate in an efficient, practical, and conscientious
manner. They typically seek employment as: accountant, bookkeeper, office worker,
banker, or IRS agent.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
27
The SII GOT scales were constructed by selecting 20 items that represent each type,
based on Holland's descriptions (Campbell & Holland, 1972; Hansen & Johansson.
1972). The scales proved to be correlated with each other, and based on the
intercorrelational strengths, an underlying hexagonal structure emerged. The structure is
such that the types that are directly opposite each other have the weakest correlations
while those next to each other have the strongest correlations (Hansen & Campbell,
1985).
For the 1985 version o f the SII (Hansen & Campbell), the GOT scale was normed on a
sample o f 600 people o f half men and h alf women. They were meant to represent People-
in-General from various professional, technical, and non-professional occupations
represented in all six GOT types (Hansen & Campbell, 1985). The mean age o f the
sample was 38.2 years, and the level o f education ranged from those without a high-
school diploma to those with Ph.D's.
The raw score means and standard deviations for this norming group were used to
create a standardization formula that converts all scores to standard score distributions
with a mean o f 50 and a standard deviation o f 10. Men and women have different
distributions for each theme, and an individual's results should be interpreted based on the
distribution for their sex. Then for each sex, the computer assigns an interpretive
comment corresponding to the percentile band the individual falls in, based on the in
general norming group. The interpretive labels include: very high, high, moderately high,
average, moderately low, low, and very low. These GOT scores are then used to provide
a general view o f the client's occupational orientation (Hansen & Campbell. 1985).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
28
Test-retest reliabilities for the GOT are reported by Hansen and Campbell (1985) to
range between .85 to .93 for a two-week interval, from .84 to .91 for a thirty-day interval,
and from .78 to .87 for a three-year interval. These reliabilities indicate relative stability,
but with some changes.
Intemal-consistency reliabilities o f the six GOT's were also computed using
coefficient alpha. For a male sample o f 1445, the range was from .90 to .95, and for a
female sample o f 1410, the range was reported at .90 to .93. This indicates a high level o f
internal consistency reliability for each GOT scale (Hansen & Campbell, 1985).
Hansen and Campbell (1985) report several validity studies o f the GOT's. One such
study at the University o f Minnesota found high correlations (median =.765) between the
GOT's and the scales on the Vocational Preference Inventory. This indicates construct
validity given that the two tests appear to be measuring similar interest traits.
The Academic Comfort Scale: This scale has undergone many label changes on past
editions o f the SII and has even been eliminated from the most recent version. However,
the Academic Comfort Scale (AC) was a part o f the 1985 version o f the SII, and a scale
which I used regularly in the interpretive process.
The scale is labeled such because it seems to be measuring comfort in an academic
setting. On the earlier versions o f the SII, the items for this scale differentiated between
good and poor students at the University o f Minnesota's College o f Liberal Arts. Items
corresponding to academic pursuits were weighted positively and items denoting such
things as blue-collar jobs were given a negative weight. When the scale was normed on
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
29
the graduates o f liberal arts colleges, the mean was 50, and Ph.D's usually scored in the
60's (Hansen & Campbell, 1985).
The 1985 version used the same norming groups as the 1974 version. For the men, the
sample included 421 college professors, and for the women, 275 psychologists and 119
mathematicians. It is unclear why they did not sample men and women from the same
occupations. The sample yielded scores with a mean o f 60 and a standard deviation o f 10.
The 1985 version continued to show scores in the 60's for Ph.D's. (Hansen & Campbell.
1985).
The 1985 reference samples showed that the 300 women in the Women-in-General
sample scored higher on AC than the 300 men in the Men-in-General sample, with scores
o f 47 and 44 respectively. The standard deviations were 14.7 and 14.9 respectively
(Hansen & Campbell, 1985). However, Broday and Braswell (1990) found virtually no
gender differences on AC for a sample o f university students. They also found that the
overall mean AC scores were lower for their college counseling center sample than for
professional men and women, and suggested this was due to level o f educational
attainment. Indeed, Hansen (1992b) provided the following standardized ranges: scores
60-65, Ph.D's; 50-55, professional degrees; 45-55, master's and bachelor's degrees; 35-44
associate or vocational/technical degrees; and 34 or lower for high school diplomas.
The reliability o f the AC scale as measured by test-retest ranges from .91 to .85 for
intervals o f two weeks to three years (Hansen & Campbell, 1985). This indicates strong
temporal stability o f the scale scores. The validity o f the AC scale is usually assessed by
using the score to determine some future behavior such as entrance into college or grade
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
30
point average. However, a variety o f studies have shown modest correlations, at best.
For instance, Hansen and Campbell (1985) report AC correlations with college grades
ranging from .10 to .30. A study by Johnson (1969) found a correlation o f .10 between
Academic Achievement (AACH, precurser to AC) and freshman GPA for undergraduate
males. Wagman (1971) found a correlation o f .35 between AACH and GPA for a group
o f undergraduate and graduate males and females. Also, Swanson and Hansen (1985)
found correlations between AC and GPA o f .26 for freshman and .20 for seniors.
In relating AC to college attritional level, research by Wright (1976) found no
differences in Academic Comfort between a group o f students who graduated from
college and a group who did not. Swanson and Hansen (1985) also found that AC seems
to act as a moderator variable when predicting the choice o f a college major. That is,
students with high AC scores in the freshman year were more likely to choose majors
consistent with their SCII profiles.
Broday and Braswell (1990) examined the relationship o f the AC scale to other scales
on the SCII. Among other things, they found that AC was significantly positively related
to the number o f items to which test takers responded "like," and significantly negatively
related to the number o f "dislike" responses. They also found strong positive
relationships between AC and Investigative and Artistic General Occupational Themes
(GOT).
Finally, a study by Tomlinson and Evans-Hughes (1991) found an interaction effect o f
gender and ethnicity on AC scale scores for a group including African Americans,
Hispanics, and Whites. In their sample. White and Hispanic women scored higher on AC
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
31
than did White and Hispanic men. Furthermore, African American men scored higher
than did African American women on the AC scale. They suggest that the relationship
between AC and educational persistence with minority populations is a subject for further
study.
The Introversion-Extroversion Scale: This scale is considered to be an indicator o f a
client's interest in working with things or ideas, versus working with people. The scale
was originally constructed by comparing SVIB responses o f two groups o f students from
the University o f Minnesota: those defined either as extroverts or introverts on the
Minnesota Multiphasic Personality Inventory (MMPI). The 1985 Women-in-General
sample o f 300 had an average score o f 48 with a standard deviation o f 10.4, and the Men-
in-General sample obtained a mean score o f 50 with a standard deviation o f 11.2.
Professional women and men scored 48 and 49, respectively, while non-professional
women and men scored 49 and 50, respectively (Hansen & Campbell, 1985). Test-retest
statistics indicate relative stability over time with correlations ranging from .91 to .82 for
intervals o f two weeks to three years, respectively (Hansen & Campbell,).
The Administrative Indexes: Finally, the test profile reports percentages o f responses
marked "Like,” “ Indifferent,” or “Dislike," for each o f the inventory parts. This index
gives an indication o f the client's response style. While the percentages vary according to
test category and gender, the mean "Like" response percentage for the inventory is 37 and
the standard deviation is 12. When a client has a high level o f any response category, this
will affect the scores across the GOT. BIS, and OS. Clients with a high percentage o f
"Like" responses have elevated scale scores and there could be a variety o f reasons why
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
32
they respond in such a manner. Some are simply curious, have a variety o f interests, or
are enthusiastic. Others may be unfocused, or unable to say "no."
Clients with a high percentage o f dislikes tend to have low scale scores. This response
tendency usually occurs for one o f two reasons. Respondents may already be interested in
a single career and so mark "Dislike" for all the others. Another reason is that they really
do not like a lot o f the available choices. The client who falls in the second category is
challenging and the counselor will need to explore the reasons behind this. Also
challenging is the client who has a high percentage o f "indifferent" responses. These
individuals are often experiencing a high degree o f career indecision or apathy regarding
choice. No clear interests emerge from the inventory, often rendering unhelpful results.
Again, the counselor must ascertain the reasons behind such a response pattern in order to
be helpful.
Advantages and Disadvantages o f the SII
One of the main advantages o f the SII is the fact that it is so widely used as a career-
decision making instrument and a tool for research. It is considered psychometrically
sound, with an excellent manual (Isaacson, 1985). Many researchers have verified the
validity of the instrument for various populations, as well as its reliability over time. The
SII provides information at both general (GOT and BIS) and specific levels (OS)
(Isaacson, 1985). Finally, the SII is relatively simple to complete in a short amount of
time.
Most of the criticisms o f gender inequality on the SII were addressed when the test
was revised. The original pink and blue forms for women and men have been
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
incorporated into one form for both genders, and test items have been updated to remove
sexist language and items (Hansen, 1992a). Research, however, does not always support
the gender neutrality o f the SII. A study by Lapan, McGrath, and Kaplan (1990) found
gender differences in the way the BIS scales are assigned to the General Occupational
Themes. Spokane (1979) using preferred occupational choice measured during the senior
year as a criterion, found that the SII had lower predictive and concurrent validity for
college females than it did for males.
Many other studies have attempted to determine the validity o f the SII for females,
with mixed results. Most o f the studies occurred after the new 1974 form and supported
its validity for females. There may, however, be concerns that gender differences on the
SII are based on gender differences in career-decision making. Crowley (1979) noted that
the SII, which is based on Holland's classification scheme, uses intrinsic factors related to
job activities as a basis for selection. Many women, however, consider extrinsic factors
more heavily, such as working conditions and the effect o f the occupation on current or
future relationships in career choice (Raphael & Gorman, 1986; Lunneborg, 1978).
Nevertheless, it is true that men and women respond differently on the SII, as evidenced
by score differences on the scales (Hansen & Campbell, 1985). Reasons for these
differences are not fully understood, and consequently the interpretations o f a woman's
profile should be made with caution and consideration o f differences in ability,
environmental presses, career development and decision making for women and men
(Schneider & Overton, 1983).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
34
Another disadvantage o f the SII is that individual response style affects the scale
scores, and thus the usefulness and interpretation o f the profile (Isaacson, 1985). This has
been an on-going frustration for me as a career counselor because the instrument is to
help people in career-decision making, but is often unhelpful for those who are especially
undecided. In particular, individuals with high amounts o f "Indifferent" or "Dislike"
responses end up with profiles that are virtually useless. This is because the resulting
profiles have very few career matches and the RLASEC scores are essentially
undifferentiated.
Regardless o f the various disadvantages, the SII remains one o f the most widely used
tests to aid in the process o f career decision making. For the purpose o f this study, it is
necessary to consider its use with subjects from college counseling centers and who are o f
and who are classified as minorities.
Use of the SII for College Students
Because the subjects in this study were students at the University o f North Dakota, it is
important to discuss the use o f the SII with a college population. Furthermore, much o f
the research conducted on the SII has involved samples o f college students. For instance.
Lunneborg (1977) found that the SII correlated well with the Vocational Interest
Inventory which has a similar theoretical basis. This evidence o f construct validity was
gathered using a college counselee sample and supported the validity o f the GOT's.
Wallace and Walker ( 1990) also used a college sample for their research and found
that the level o f congruence between students' SII profile and their current college major
was moderated by the students' self-concept. Specifically, students with high self
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
35
concepts had better congruence scores (matches) between their interests on the SII and
their choice o f academic major. And conversely, students with low self concepts had
lower congruence scores between their SII interests and academic major choice.
Furthermore, the congruence between self concept and interests was not affected by
gender and ethnicity. Therefore, it might be wise for a college career counselor to
consider issues o f self-concept in addition to interests as part o f the client's career
decision-making process.
Holcomb and Anderson (1978) studied whether or not congruence between a student's
interests and academic major choice would predict college graduation. Among their
sample o f 195 college agriculture students, 54% had SVIB interest patterns congruent
with their current major (labeled Congruent). 46% o f the students (labeled Discrepant)
did not have SVIB interests congruent with major. There were no differences between
the groups on the rate o f college graduation. Differences were found, however, on
changing academic majors. Discrepant subjects were more likely to change majors than
congruent subjects.
Wigington (1985) examined occupational choice and various aspects o f client's
choices on the SCII in a college population. His study is based on the idea that, as
Campbell (1971) reported, there are correlations between the Academic Achievement
scale and the General Occupational Themes that suggest ordering o f the themes.
Campbell suggested that the relationship leads to a theme order o f I,A,S,C,R,E. This
occurs because students with high Academic Achievement Scale scores will likely choose
items which correspond to professional careers or college subjects most o f which
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
36
are represented in the GOT's o f Investigative, Artist, and Social. Holland (1973)
suggested that the relationship is I,S,A,C,E,R. Wigington (1985) found a different
relationship between Academic Comfort and GOT for college women than for men. For
women, he found an order like Campbell (1971) did o f I,A,S,C,R,E. But for men, he
found an order of I,A,R,S,C,E. Even though the order is somewhat different, it appears
that Academic Comfort scores correlate highly with Investigative and Artistic themes,
ones which have a high correspondence o f college majors leading to professional careers.
He also found that students with a high percentage o f "Like" scores will have elevated AC
and GOT scores, and low I/E scores. This suggests that the SCII is indeed, affected by
response style, something college counselors need to closely consider.
Based on subjects in a college sample, Smart (1989) developed a causal model to
investigate the influence o f various aspects o f life history on the development o f
Investigative, Social and Enterprising personality types proposed by Holland. These three
types were selected because most college students typically enter jobs that would be
classified in one o f these areas following graduation.
Smart (1989) found that gender had a direct effect on the development o f personality
type. The results indicated that more men were likely to become Investigative or
Enterprising types and more women were likely to become Social types (Smart, 1989).
Smart (1989) also found that family socioeconomic status as well as parental occupation
influenced the development o f the three personality types in both the positive and
negative directions. Specifically, those who develop into Social types tend to be women
from less affluent families whose parents are likely working in Social occupations.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
37
whereas Investigative and Enterprising types tend to be males from more affluent families
in which parents are typically not employed in Investigative and Enterprising jobs (Smart.
1989). Furthermore, Smart (1989) found that a subject’s orientation to being either
Social, Investigative, or Enterprising in 1971 was directly positively related
to their vocational type in 1980.
With respect to the college experiences o f the subject, it appears that undergraduate
GPA does not have a significant direct effect with reference to any o f the personality
types (Smart, 1989). It does, however, affect Investigatives in a positive, indirect fashion,
and Socials in a negative, indirect manner. Smart (1989) stated that the congruence
between choice o f college major and current vocational type has a much stronger
influence than GPA in his causal model.
Another aspect to consider regarding the use o f the SII for college populations is
whether there are differences between counseled and non-counseled students' responses
on the instrument. Tryon (1983) found that counseled students who took the SCII had
higher AC scores as well as higher scores on Investigative and Artistic themes than did
non-counseled students. She also found that non-counseled students had higher
Enterprising theme scores than did counseled students. She concluded that students who
are low on academic orientation do poorly in school. This study is interesting to me
because based on my experience, the results would be the opposite. That is, counseled
students tend to have lower AC scores and themes o f I and A than non-counseled
students.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
38
Use o f the SII with Minority Populations
Because this study is investigating the use o f the SII for a rural population, a group
often considered to be a minority, it is necessary to review the literature on this topic. It
should be noted that all studies reviewed here used college students as subjects, thus
supplementing the literature in that area as well. To date, several researchers have
investigated the use o f the Strong Interest Inventory with various minority populations.
Results have been mixed, with some studies supporting the use o f the SII for ethnic
minorities, and others not (Hansen, 1992b).
Haviland and Hansen (1987) found that the SCII had reasonably good concurrent
validity when examining college major and Occupational Scale scores for a group o f
American Indian college students. Their research sample used 49 American Indian
students at a Rocky Mountain region college (28 women, 21 men). They assessed
criterion validity in two ways. First, they used a modified McArthur method (assigning
point values for prediction correctness for an "excellent hit, moderate hit, or poor h it," ) to
determine the concurrent validity o f the SVIB-SCII Occupational Scales for declared
academic major. Second, they evaluated congruence between corresponding GOT scores
and major (Haviland & Hansen, 1987).
Their results showed that for sixty-four percent o f the women and forty-four percent o f
the men, there was excellent and moderate concurrent validity for relationships between
major and OS. For congruence between major and GOT, their results indicated high and
moderate congruence for eighty-nine percent o f the women and one hundred percent o f
the men (Haviland & Hansen. 1987). Their discussion o f the results suggested that this
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
39
instrument is at least as valid for Native American women as it is for White women,
given that the hit rate for the Native American women resembles that o f White women.
However, the forty-four percent rate for Native American men is markedly lower than the
sixty percent reported for white men by Hansen and Swanson (1983). The authors
believed this may be due to the lower graduation rate for Native American men, the fact
that the men select majors that do not lead to a terminal degree, and the evidence that
Native American women seek career counseling services more often than the men,
leading to more well-developed career interests. Finally, the authors suggested that the
results o f the study are limited in their generalizability to other American Indian college
students because the sample was small in number and represented only a few tribes.
Another study by Whetstone and Hayles (1975) found that the Strong Vocational
Interest Blank (SVIB) predicted career group membership for a sample o f Blacks as well
as it did for Whites. In their study, they wished to know if the SVIB norms were usable
for Black men, if the SVIB was adequate for detecting primary and secondary interest
patterns for Black college men, and whether Black and White college men differed in
their interest patterns. Their sample included 69 Black college men and 75 White college
men from the University o f Colorado. The researchers classified a subject's major as
either consistent, questionable, or inconsistent with his SVIB results. The results were
positive, indicating that approximately two-thirds o f the Blacks and three-fourths o f the
whites had SVIB primary and secondary interests consistent with their major (Whetstone
& Hayles, 1975). There were differences, however, on the BIS, and Black men also had
lower Academic Achievement scores (precursor to AC). The authors suggest that the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
40
results should be applied with caution to Blacks from rural populations, as they were not
included in this sample (Whetstone & Hayles, 1975).
Swanson (1992) also found that the SII appears applicable for African-American
college students. Her study sought to ascertain whether two aspects o f Holland's model,
the underlying hexagonal structure and the ordering o f the GOT's, was the same for Black
students as for Whites. The subjects consisted o f 1826 Black Midwestern college
students. The results showed that Holland's model, on which the SVIB-SCII is based has
the potential for applicability for Black students, but that more research is needed. The
underlying hexagonal structure, while still somewhat hexagonal in shape, deviated from
its equilateral shape because o f significant differences in the intercorrelations among the
GOT scales. Swanson (1992) also found significant differences between Black men and
women's GOT scores, as well as between the GOT scores for the Black subjects and the
reference sample indicating both gender and race effects. Again, Swanson (1992)
suggests that these findings indicate a need for more research with the SVIB-SCII for
Black Americans.
Montoya and DeBlassie (1985) endeavored to determine if differences existed on the
SCII responses between a group o f Anglo and Hispanic college students. Their subjects
were 88 Anglo and 88 Hispanic students at New Mexico State University, who were
administered the SCII as part o f career counseling. Results showed no differences
between Anglo and Hispanic students on the Introversion/Extroversion scale or on the
GOT's. However, on the Realistic theme, although there were no race effects found, there
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
41
were significant differences between males and females. These results suggest that the
SCII may be applicable to the Hispanic population.
Tomlinson and Evans-Hughs (1991) researched gender and ethnic differences on the
SCn for a group o f White, Black, and Hispanic college students at an eastern university.
They found an effect for gender on the Realistic theme with males scoring higher than
females. They found no differences between ethnic groups on any GOT scales. The
Academic Comfort (AC) score mean for the group was 27.9, lower than would be
expected for a typical university sample. No significant effects for ethnicity or gender
were found on the AC scale. Another observation was that 49% o f the profiles were "flat
and undifferentiated", which the authors attribute to the low AC scores. A question which
the authors did not address, is whether the sample AC score differs significantly from the
general reference sample, and if so, why?
The results o f validity studies for the use o f the SII with other minority populations
have yielded mixed results. Therefore, it is appropriate to examine the use o f the SII for
rural individuals, who are often considered a minority group.
Career Concerns for Rural Populations
Although there are many within group differences, there are some commonalities
known about the quality o f life for rural individuals. Murray and Keller (1991) reported
that rural areas contain a disproportionate number o f poor, and that the economic
development o f rural areas generally lags behind that o f metropolitan areas. They further
stated that rural areas have smaller proportions o f workers employed in white-collar
occupations, rural residents attain lower educational levels, and are less frequently
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
42
members o f the work force. If this is true, how might this impact the development o f
interests germane to the career selection process? Additionally, many rural Americans
have, in the past, been farmers. However, due to recent agricultural crises, many have
had to leave this career. How has this factor affected students' career decision-making?
Apostal and Bilden (1991) suggested that rural students as compared to urban students
have unique considerations affecting the process o f educational and career decision
making. They stated that rural students are hampered because o f "reduced accessibility to
higher education, narrow rural school curricula, limited exposure to the world o f
occupations, and few role models (p. 153.) They further suggested that the 1980
agricultural crisis may have had an effect on the educational aspirations o f rural students
such that increasing numbers are moving towards higher education. Indeed, their study
showed that there is a great disparity between rural and urban students' educational
aspirations. Their sample o f North Dakota high school students showed that 72.4%
intended to pursue a four-year college education, compared with 29% in a national
sample. Apostal and Bilden (1991) stated that North Dakota career counselors should
note these high aspirations and should provide students with guidance which also
considers their abilities and interests.
Trice (1990) concluded that the career interests o f rural children are relatively stable
over time, and are very much related to those o f their parents and other community
members. Thus, if rural children are exposed to smaller numbers o f careers, how might
this factor affect the range o f their career interests and responses on the SII?
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
43
Poole, Langan-Fox, and Omodei (1991) discussed several examples o f research in
which geographic location is taken into consideration for determining career choice.
They refer to Schiamberg and Chong-Hee Chin's study (1987, cited in Poole, et al, 1991)
which concluded that the strongest influence on a child's later choice o f education and
career is their family background. They further referenced several studies which suggest
that rural families are more traditional than urban families. This may be due to strong
kinship ties and limited social interaction which in turn allows already held conventional
values to be reinforced and strengthened. I f these conventional values influence
subsequent career choice, how then might this affect responses on the SII?
Lauver and Jones (1991) explored a self-efficacy model for career choice among rural
American Indian, White, and Hispanic high school students. They found gender
differences in career self-efficacy and perception o f career options. In particular, their
findings suggested that rural females’ efficacy and aspirations equal or exceed those o f
rural males. But, what happens when these rural students enter college? Does efficacy
remain high given the abrupt change from a close-knit rural community to the larger
urban campus?
Summary o f the Literature
Throughout the review o f the literature on vocational interest development, there exist
some common threads. Whether the theory is developmental, based on personality or
social learning principles, or focused on valence instrumentality for career choice, the
basic assumption is that we all have interests or preferences for different activities or
occupations. Furthermore, the theories assume that we all have a basic fund o f
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
44
knowledge about different careers. Additionally, many o f the theories have suggested
that interests are very much related to aspects o f the environment as well as individual
characteristics.
All o f these assumptions underlie the structure o f the Strong Interest Inventory, which
continues to be one o f the most widely used and researched career interest inventories.
When researching the SII, many scientists have endeavored to ascertain its usefulness
with a variety o f populations. A number o f studies have been based on university
populations and have found the SII to be useful for career decision making. Research
studies based on college samples have helped establish the SITs reputation for excellent
reliability and validity. Additional research has generally concluded that males and
females do tend to respond differentially on the SII, but that this is not necessarily an
indication o f gender bias. Furthermore, many researchers have concluded that, for the
most part, the SII appears to be applicable to many minority populations including
African-American, Hispanic, and Native American.
Statement o f the Problem
This study is an attempt to build on the already extensive body o f literature on the SII.
To date, relatively little has been written which specifically addresses the vocational
interests o f rural students. Since rural students are considered to be o f minority status and
often an under-served population, it seems important to consider their needs. While
serving as a career counselor at the University o f North Dakota, I had the opportunity to
assist rural students in the process o f career decision making. In many instances, the SII
was used to assist in this process, and the SII results obtained were not very helpful due to
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
45
the low GOT scores and few corresponding careers on the OS. I began to wonder if the
aspect o f being from a rural environment had an effect on the SII responses. Based on my
clinical observations, as well as on the lack o f literature support for the use o f the SII with
rural populations, this study was designed to investigate whether differences exist
between rural and urban students on selected SII scales. And, because previous research
suggests that gender differences often exist on the scales, it is necessary to consider this
aspect, as well, in some o f the analysis. Also, since previous research suggests that
gender, AC, and ACT scores may be used to some extent to predict college GPA, I wish
to investigate whether they predict in the same way for rural students. This factor is
important to consider for purposes o f retention, a common concern for universities with
respect to minority students. Furthermore, given that students from rural and urban
backgrounds may have different interests and levels o f knowledge and exposure the world
o f work, and that these interests are likely reflected in their SII GOT scores, I would like
to investigate whether differences exist between the two groups on measures o f
congruence between the RIASEC scores and college major. A related question is whether
or not differences exist between rural and urban groups in the way they respond on the SII
with respect to the number o f “indifferent” responses chosen. A final factor for
examination is whether the two groups differ in the amount o f differentiation in their
RIASEC scores, since a lack o f differentiation results in low, undifferentiated, and
unuseful profiles. Therefore, I have formulated the following questions that I would like
to investigate in this study:
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
46
a. Are there mean differences between rural and urban college students on the SII GOT
scales, AC scale, and the I/E scale?
b. Does being from a rural/urban environment moderate the relationship o f AC and ACT
to GPA?
c. Does being from a rural or urban environment affect the lachan congruence score
between choice o f college major and RIASEC scores?
d. Does being from a rural or urban environment affect the subject’s level o f indifference
in responding on the SII?
e. Does being from a rural or urban environment affect the subject’s profile
differentiation on RIASEC scales?
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 3
Method
Statement o f the Problem
The purpose o f this study was to examine whether differences existed between rural
and urban subjects’ scores on various scales o f the SII, their level o f “indifferent”
responses, predictive validity o f the GOT and the subject’s major, and the degree o f
differentiation in a subject’s GOT scores. An additional focus was whether the
relationship between AC and cumulative GPA was the same for subjects from rural and
urban environments. The questions for study, stated in null hypothesis form are as
follows:
a. There are no mean differences between rural and urban subjects on the
SII GOT scales, the AC scale, and the I/E scale.
b. There is no difference in relationship between GPA and AC for
rural and urban subjects.
c. There are no mean differences between rural and urban subjects’ Iachan
scores (measuring the congruence between the GOT scores and major).
d. There are no mean differences between rural and urban subjects on their
level o f indifference in responding on the SII.
47
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
48
e. There are no mean differences between rural and urban subjects on their
profile differentiation scores for the RIASEC scales.
Subjects
The subjects for this study were 665 students who received services at the University
o f North Dakota Counseling Center over six academic years (1988-89 through 1993-94).
O f the total sample, 36.2% were males (n = 241) and 63.8% were females (n = 425). The
ages o f these participants ranged from 18 to 25. The majority o f subjects were age 18 to
20. (Subjects older than age 25 were excluded from the sample because the impact o f
geographic location may be different for them than for subjects in the traditional student
age range o f 18 to 25. Another reason for exclusion is that the bulk o f career literature is
based on this age group.) Table 1 summarizes the demographic characteristics o f the
subjects. The sample included enrolling students as well as students ranging from
Freshmen through Graduate/ Professional school. A few subjects were listed as Other
because they were neither enrolling nor currently enrolled. Over half the sample were
either Freshman or Sophomores. With regard to ethnicity, the vast majority (96.7%) o f
the subjects participating in the study were Caucasian. The majority o f students were
enrolled in University College which is assigned when students are undecided about their
major. Cumulative grade point average ranges and the percentage o f subjects falling into
each range are also noted in Table 1.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
49
Table 1
Demographic Characteristics of the Sample
Variable Category %
Age
Class
Ethnicity
College
Cumulative GPA
18 128 19.2 19 231 34.7 20 148 22.3 21 61 9.2 22 46 6.9 23 21 3.2 24 18 2.7 25 12 1.8
Enrolling 3 0.5 Freshman 212 33.8 Sophomore 255 40.6 Junior 113 18.0 Senior 38 6.1 GradVProfessional 2 0.3 Other 5 0.8
Caucasian 549 96.7 African-American 2 0.4 Asian/Pacific Island 3 0.5 Intemat’l Student 4 0.7 Hispanic 2 0.4 Native American 3 1.1 Other 5 0.9
Arts & Sciences 132 21.3 Business & Professional 67 10.8 Aerospace Sciences 28 4.5 Teaching and Learning 13 2.1 Engineering & Mines 23 3.7 Fine Arts 6 1.0 Graduate School 1 0.2 Medicine 13 2.1 Nursing 13 2.1 University College 303 48.9 Non-students 9 1.5 Human Resources/Dev. 11 1.8
2.00 & Below 55 8.7 2.01-2.49 108 16.8 2.50-2.99 192 30.2 3.00-3.49 180 28.4 3.50-4.00 130 15.0
Note. Total n=665.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
50
Measure and Variables
The principle measure used in this study was the Strong Interest Inventory (SII). The
SII is a 325 item self-report career interest inventory containing questions about various
occupations, school subjects, leisure activities, types o f people and characteristics o f the
test-taker. Internal consistency reliabilities as well as validity coefficients are reported for
all scales in Chapter 2 o f this dissertation.
General Occupational Theme Scales. The SII GOT scales reflect each o f the six
Holland types. The scales are: Realistic (R), Investigative (I), Artistic (A), Social (S).
Enterprising (E), and Conventional (C). For the General Occupational Themes, test-retest
reliabilities range from .85 to .93 for a two week interval and from .78 to .87 for a three
year interval (Hansen & Campbell, 1985). Scores are normed to have means o f 50 and
standard deviations o f 10 for a mixed gender norm sample.
Academic Comfort Scale (AC). This scale reflects the level o f comfort in academic
settings. Scores are normed to have means o f 60 and standard deviations o f 10 for mixed
gender norming samples (Hansen & Campbell, 1985). Scores in the 50's and 60's tend to
indicate comfort in higher academic settings such as graduate school. Scores below 40
tend to denote individuals who are more comfortable in applied settings.
Introversion/Extroversion (I/E). This scale is considered an index o f a subject’s
interest in working with things or ideas versus working with people. Scores are normed
with means ranging from 48 to 50 for men and women, respectively. Standard deviations
range from 10.4 for women to 11.2 for men. Test-retest statistics show that the scale is
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
51
relatively stable over time with correlations ranging from .91 to .82 for intervals o f two
weeks to three years respectively (Hansen & Campbell, 1985).
GPA. This term refers to the cumulative grade point average o f the subjects, based on
a 4.0 scale. This information was obtained from the student’s university files.
ACT. This term refers to the American College Test (ACT) composite score. Like
GPA, this information was obtained from the subject’s university files.
Rural/Urban Status CRur/Urb or R/UV The subjects were classified according to
hometown population as either rural or urban. The U.S. Bureau o f Census definition o f
rural and urban populations was used to determine this classification (Murray & Keller.
1991). Based on this definition, an urban subject is one from an area in which the city
and closely settled surrounding area has 50,000 or more inhabitants. A rural subject,
then, is one from an area in which the people live outside o f the city and closely inhabited
areas and contains less than 50,000 people.
Congruence flachanf. In order to examine the congruence between a subject’s selected
college major and his/her three-letter sequentially ordered Holland code (Holland, 1973,
1985) the measure proposed by Iachan (1984, 1990). The Holland code for each subject’s
major were assigned by the Career Counseling Services at the University o f North
Dakota’s Counseling Center. Then, a computer program was used to calculate lachan’s
Congruence Index (Iachan, 1984, 1990) for each subject. Table 2 depicts an adapted
version o f the scores assigned to the SII GOT and major three-letter Holland codes using
the Iachan method (Iachan, 1990, p. 177).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
52
Table 2
Iachan Congruence Index Calculation Codes
Major or Degree Code
SII GOT Code
1st Letter 2nd Letter 3rd Letter
1st Letter 22 10 4
2nd Letter 10 5 2
3rd Letter 4 2 1
The resulting Iachan Congruence Index score is a measure o f the predictive validity for
a subject’s SII GOT for college major. Thus, higher Iachan scores indicate higher
predictive validity o f the SII.
Indifference Z-scores IZIND:A.B.C.D: ZINDTOT). To examine whether the number
o f “Indifferent” responses a subject made on selected scales is less than, greater than, or
as expected given their scale score, an index called Z-Indifference (coded ZIND) was
computed based on the item response theory modeling o f SII responses by Henly (1995).
This information is o f interest because it could be that both rural and urban subjects have
similar scale scores, but one group may have a higher or lower number o f “Indifferent”
responses. For instance, if a rural subject and an urban subject have similar scores on the
Artistic theme scale, it would appear that their interests in this area are similar, when in
fact the urban subject may have selected more “ Dislike” responses and the rural subject
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
53
may have chosen more ‘‘Indifferent” responses. This index was formulated by first
calculating a subject’s score on a particular scale. There were four scales for female
subjects and three scales for males used to provide information about the number of
“Indifferent” responses selected. Once an individual’s score on the scale is calculated,
then the number o f “Indifferent” (or “I”) responses that the subject made is compared to
the expected number o f “I” responses for a subject with that score. A difference score is
calculated and then divided by the expected variance o f the number o f “I” responses at
that score level. The score that results is a z-score with a mean o f 0.0 and a standard
deviation o f 1.0. In practical terms, this means that subjects with negative z-scores
obtained their scale score by selecting fewer Indifferent responses and therefore, more
Like and Dislike responses than would have been expected. The reverse, then, is true for
positive z-scores.
Profile Differentiation fProDifl. This score is a measure o f differentiation between the
highest and lowest GOT scores. It was calculated by computing the variance o f the
respondent’s GOT scores.
Procedure
The SII data for this study was archival and obtained from the files at the University o f
North Dakota’s Career Counseling Services. In addition to the SII information, other
demographic information was obtained from the subjects’ files at the Counseling Center.
This included: race/ethnicity, and age. The following information was obtained from the
subjects' general university records located in the Office o f Student Affairs: hometown
and its population. ACT scores, cumulative GPA and college major.
Reproduced with permission of the copyright owner. Further reproduction prohibited w ithout permission.
54
The data from the University Counseling Center and Student Affairs was linked and
statistical procedures were applied to analyze the data.
Analysis
To examine the RIASEC scales, the AC scale, and the I/E scale for mean differences
between rural and urban groups, a multivariate analysis o f variance was conducted
separately for each gender because the assumption o f equivalent variance-covariance
matrices was violated, likely due to restricted variance for females on the R scale. In this
analysis, rural/urban was used as the independent variable, and the six GOT scores, AC
and I/E as dependent variables.
Multiple regression was used to examine whether AC, ACT, Gender or being from a
rural or urban environment had an effect on GPA. Two separate equations were formed
in which GPA was used as the dependent variable. In the first equation, AC, Gender, and
Rural/Urban were used as the independent variables. Also, the relationships were
examined for interaction effects between the variables including Gen by AC, R/U by AC.
and Gen by R/U. In the second equation, AC, Gender, Rural/Urban and ACT served as
independent variables. Also, the relationships were examined for interaction effects
between the variables including Gen by AC, Gen by ACT, AC by ACT, R/U by AC. Gen
by R/U and R/U by ACT.
A 2 (gender) x 2 (rural/urban) hierarchical analysis o f variance was conducted to
examine differences between rural and urban subjects on predictor fit as measured by the
Iachan congruence index controlling for effects o f gender. In this analysis, the Iachan
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
55
score served as the dependent variable and gender and rural/urban status were the
independent variables.
To examine the differences between rural and urban groups on the Z-indifference
scores, independent sample t-tests were used. A total z-indifference score was calculated
by summing the z-indifference scores across the scales (ZINTOT). Four scales for
women (ZINDA, B, C, and D) and three scales for men (ZINDA, B, and C) were tested.
The analyses were run separately for males and females due to the differences between
genders in the models for the scales.
A 2 (gender) x 2 (rural/urban) hierarchical analysis o f variance was conducted to
examine differences and potential interaction effects between gender and rural/urban
status on profile differentiation among the RIASEC scales. In this analysis, ProDif
served as the dependent variable and gender and rural/urban were the independent
variables.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 4
Results
The means and standard deviations for GOT scales, AC, IE, ACT, GPA, Iachan
congruence index, and Profile Differentiation for rural and urban subjects by gender are
summarized in Table 3. In general, the RIASEC scores for rural, urban, male and female
subjects would be categorized in the moderately low to average range when compared to
those o f the general sample upon which the SII was normed (Hansen and Campbell.
1985). The ordering o f the GOT scores from highest to lowest is as follows: for rural
males, RECSLA.; for urban males, ERCASI; for rural females SCEAIR, and for urban
females, SECAIR.
The mean AC scores were as follows: rural males, M=28.4; urban males, M=27.57;
rural females, M=32.6; and urban females, M =31.54. These scores are all substantially
lower than those for the norming sample whose mean was equal to 39 (Hansen and
Campbell, 1985). The I/E scores for all four groups are similar to the mean o f the
norming sample which was M=53 (Hansen and Campbell, 1985). The I/E mean scores
are as follows: rural male, M =55.64; urban male, M=54.37; rural female, M =50.59, and
urban female M =54.37.
The results o f the multivariate analysis o f variance run on the female sample indicated
that there were no significant differences between rural and urban groups on any o f the
56
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
57
GOT scales for R, I, A, S, E, C, or on the AC and I/E scales (F=.891, p<.50). The results
o f the multivariate analysis o f variance for the sample o f males indicated that there were
Table 3
Sample Means and Standard Deviations on SII Scales and for ACT. GPA. Iachan
Congruence Index Scores, and Profile Differentiation Scores
Males______________________Females Total Sample Rural Urban Rural Urban
(n=665) (n=l 16) (n=125) (n=231) (n=193) Variable M, SD M SD M S P M S P M S P
R 41.69 9.61 48.88 9.85 47.31 9.73 38.44 7.73 37.61 6.88
I 41.12 9.18 42.42 9.94 42.47 9.36 40.30 8.93 40.44 8.76
A 42.96 10.65 39.60 11.01 40.68 10.68 44.62 10.54 44.47 9.87
S 49.38 10.12 45.15 10.33 46.07 9.45 51.37 10.02 51.69 9.17
E 48.07 10.21 47.59 10.37 49.78 10.91 47.77 10.24 47.62 9.54
C 47.34 10.29 45.62 9.49 46.69 10.09 48.82 10.61 47.02 10.33
AC 30.61 14.64 28.40 13.99 27.57 16.13 32.60 13.92 31.54 14.45
IE 52.13 11.65 55.64 12.41 54.37 11.28 50.59 11.55 50.40 10.88
ACT 22.12 4.13 22.57 4.28 20.54 4.09 22.91 3.92 21.72 4.06
GPA 2.85 .64 2.80 .64 2.67 .59 3.04 .59 2.77 .66
Iachan 14.88 8.13 13.77 8.11 13.84 7.89 15.05 8.12 16.06 8.19
ProDif 90.58 56.86 86.32 53.62 85.39 59.54 93.73 56.04 92.71 57.97
no significant differences between the rural and urban groups on any o f the scales
(F=.885, p<50). So, in this case, the null hypothesis was retained. The means and
standard deviations are reported in Table 3.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
58
Multiple regression was used to examine whether GPA could be predicted by AC,
gender or by geographical category o f rural/urban. Test results indicate that gender and
AC contribute to the prediction o f GPA (F=l 7.11, p < .01). Further, addition o f rural-
urban status significantly improved prediction o f GPA. None o f the interaction terms in
the model was significant, indicating that the relationship between AC and GPA does not
differ for gender or rural-urban subgroups. The results are reported in Table 4.
For those subjects for whom ACT scores were available, multiple regression was also
used to examine whether GPA could be predicted by AC, gender, rural/urban and ACT
score. Results o f the analysis indicate that only AC, Gen, and ACT contribute to the
prediction o f GPA as evidenced by F = 25.81, p<.QQQ. It should be noted that when ACT
scores are added to the equation, the effect o f rural/urban disappears, either as an additive
or interactive effect. The potent impact o f rural/urban in the first regression equation may
be impacted in the second equation by the sizeable differences in ACT scores between
rural and urban students (rural students scored a h a lf standard deviation higher). The
results are reported in Table 5.
The results o f the 2x2 hierarchical analysis o f variance which measured predictor fit
between selected college major and the RIASEC scales (the Iachan index) for each
subject indicated that there were no significant differences between rural and urban
subjects. However, significant differences did exist between genders, with females
tending to have greater congruence between their SII type and their college major
(F ( 1 ,665)=5.758, p < .017). There was no interaction effect between rural/urban status
and gender. The results are reported in Table 6.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
59
Table 4
Summary o f Hierarchical Regression Analysis for Variables Predicting GPA-Equation 1
Variable B SEB P AR* df Step 1 .051 17.11 <.001 2,632
AC .008 .002 .180*
Gender .152 .052 .115*
Step 2 .000 .01 ns 3,631
AC .007 .006 .167
Gender .142 .115 .108
GenXAC .000 .004 .016
Step 3 .026 17.62 < 0 1 4,630
AC .007 .006 .164
Gender .132 .113 .100
GenXAC .000 .004 .014
Rur/Urb -.205 .049 -.161
Step 4 .005 1.05 ns 6,628
AC .011 .008 .245
Gender .330 .191 .250
GenXAC .000 .004 -.003
Rur/Urb .061 .190 .048
R/UXAC -.002 .003 -.086
GenXR/U -.123 .102 -.215
Note. Fa is for change due to addition o f variables in step.
a * e<.05.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
60
Table 5
Summary o f Hierarchical Regression Analysis for Variables Predicting GPA-Equation 2
Variable B SEB. @ & R a Fa B df
Step I .312 25.81 <.001 3,171
AC -.003 .003 -.069
Gender .235 .069 .219*
ACT .052 .007 .514*
Step2 .005 .43 ns 6,168
AC -.016 .014 -.421
Gender .167 .289 .156
ACT .040 .243 .402
GenXAC .003 .005 .160
GenXACT -.000 .015 -.023
ACXACT .000 .000 .306
Step 3 .000 .02 ns 7,167
AC -.017 .014 -.424
Gender .168 .290 .156
ACT .040 .025 .400
GenXAC .003 .005 .169
GenXACT -.000 .015 -.024
ACXACT .000 .000 .313
Rur/Urb -.011 .070 -.010
Reproduced with permission of the copyright owner. Further reproduction prohibited w ithout permission.
61
B § E B £ A Ra I 0' E df
Step 4 XH9 L52 ns 10,164
AC -.038 .018 -.994*
Gender .360 .403 .335
ACT .059 .036 .591
GenXAC .005 .006 .268
GenXACT -.004 .016 -.113
ACXACT .000 .000 .378
Rur/Urb .093 .370 .089
R/UXAC .011 .006 .536*
GenXR/U -.114 .143 -.230
R/UXACT -.012 .015 -.265
Note. F^is for change in R due to variables added at step.
a *E<-05
Reproduced with permission of the copyright owner. Further reproduction prohibited w ithout permission.
Table 6
62
Analysis o f Variance for the Iachan Index o f Congruence by Gender and Rural/Urban Status
Source o f Variation SS d f MS F c
Main Effects 440.03 2 220.01 3.354 .036
Gender 377.73 1 377.73 5.758 .017
Rural/Urban 62.30 1 62.30 .950 ns
2-wav Interactions
Gender x Rur/Urb 28.85 1 28.85 .440 ns
Explained 468.88 3 156.29 2.383 ns
Residual 37324.86 569 65.60
Total 37793.73 572 66.07
Note. 92 cases had missing data.
The results for the analysis o f Z-indifference are listed in Table 7. The analysis using
t-tests for independent samples indicated that there were no significant differences
between rural and urban groups on any o f the ZIND measures for women or men.
When examining mean differences between groups for profile differentiation in the
RIASEC scores, the results o f the 2x2 hierarchical analysis o f variance indicated that
there were no significant differences between rural and urban subjects or between
genders. There also was no interaction effect between rural/urban and gender on this
measure.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Table 7
63
Results o f Independent t-tests for z-indifference Scores
Variable Female Male Rural Urban Rural Urban
________ n M/(SD) n M /tSm t p n M/fSDt n M/fSDt t n ZIN D A 212 -.03 180 -.07 .24 .81 105 -.19 116 -.14 -26 .80
ZINDB 211 .01 180 -.17 1.10 .27 105 .08 114 -.06 .62 .53
ZINDC 155 -.03 133 -.09 .53 .60 100 -.15 107 .04 -.99 .32
-.03 180 -.07 (1.73) (1.83)
.01 180 -.17 (1.53) (1.62)
-.03 133 -.09 (0.94) (0.87)
.10 167 .10 (1.34) (1.45) .16 133 -.28
(3.45) (3.35)
ZI NDD 203 .10 167 .10 -.05 .96 (1.34) (1.45)
ZINTOT 155 .16 133 -.28 1.11 .27 100 -.21 105 .05 -.52 .61
-.19 116 -.14 (1.53) (1.50)
.08 114 -.06 (1.60) (1.50)
-.15 107 .04 (1.32) (1.50)
-.21 105 .05 (3.65) (3.53)
Reproduced with permission of the copyright owner. Further reproduction prohibited w ithout permission.
CHAPTER 5
Discussion
The purpose o f this study was to explore whether differences existed between rural
and urban geographical groups on the GOT scales, AC, and I/E o f the SII for a sample o f
college counseling center students at the University o f North Dakota. An additional focus
of the study was determining whether cumulative GPA could be predicted by variables
such as gender, AC, rural/urban, ACT, or interactions among them. Additional
rural/urban differences were explored on Iachan measures o f congruence between college
major and GOT scores, for scales measuring the level o f indifference in responding on the
SII, and on the degree o f differentiation among a subject’s GOT scores. Also, because o f
literature support for differences between genders on different facets o f the SII, this was
examined in several analyses as well.
Overall, the findings in this study seem to indicate that there are no significant
differences between rural and urban subjects on AC, I/E, any o f the RIASEC scales, on
the Iachan measure o f predictor fit between the RIASEC scales and selected college
major, on measures o f z-indifference or on the levels o f profile differentiation. In light o f
the researcher’s practical experience, these results are surprising. Reflection upon these
findings leads to several hypotheses.
64
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
65
First, it could be that there are, in fact, no real differences between rural and urban
subjects career interests as measured by the SO. While there is no published literature
to date comparing these two cultural groups, there is research to support the efficacy o f
the SII for various other cultural groups. The results o f the multivariate analysis o f
variance indicated there were no mean differences between rural and urban subjects on
any o f the RIASEC scales, AC or I/E. The findings o f this study are similar to the data
obtained by Montoya and DeBlassie (1985), who found no significant differences on the
GOT’s and the I/E scale between a sample o f Anglo and Hispanic students. Tomlinson
and Evans-Hughs (1991) also found no differences between groups o f White, Black and
Hispanic college students on any GOT scales.
Another hypothesis about the lack o f differences between the groups is that the results
were affected because this sample was strictly drawn from students seeking career
counseling services at the UND Counseling Center. Indeed, Tryon (1983) found
differences in certain SII scores between students seeking counseling services and
students in the general college population. In reality, perhaps differences do exist
between rural and urban students on the SII for a general college sample. However, at the
time o f this research, only SII’s from a counseled population were available. It would be
important in future research to obtain data from both counseled and non-counseled
populations to see if this factor has an effect on the mean scores o f rural and urban
students on the GOT’s, AC and I/E.
The results o f the first multiple regression analysis showed that AC, gender, and
rural/urban status all made significant contributions to the prediction of GPA. However.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
66
it should be noted that Fp^.08 indicating that these variables only account for 8% o f the
explained variance when predicting GPA. In the second multiple regression analysis,
when ACT was added into the equation along with AC, gender and the interaction
variables, R*was equal to .32. This, then, accounted for 32% o f the explained variance in
predicting GPA. The impact o f rural/urban is reduced by the addition o f ACT in the
second equation. This may be due to sizeable differences between rural and urban
students on ACT composite scores. Rural students’ scores are half a standard deviation
higher. These findings are similar to studies by other researchers who found very modest
correlations between AC and GPA (Hansen & Campbell, 1985; Johnson, 1969; Wagman.
1971; Swanson & Hansen, 1985). One could conclude from the results that while AC.
gender, ACT and to some extent rural/urban all contribute to the prediction o f cumulative
GPA, there are many other factors which have an impact. Universities which are
investigating such factors for purposes o f student selection and retention should consider
this.
The data from this study’s examination o f concurrent validity using the Iachan model
to ascertain the level o f predictor fit between a subject’s RLASEC scores and their college
major suggest that the RIASEC scores predict college major equally well for both rural
and urban students. These results are similar to those o f Whetstone and Hayles (1975)
who found that the SVIB predicted career group membership as well for Black males as
for White males. The results o f this study also support those found by Haviland and
Hansen (1987). Their study included both American Indian and Caucasian subjects and
concluded that the GOT's were good predictors o f college major for both groups.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
67
Also, it should be noted that there were significant differences between males and
females when using the Iachan model to measure the fit between the GOT scores and
college major. This supports the research o f Spokane (1979) who found that concurrent
validity for the SII and preferred occupation differed for men and women.
The results o f the Independent t-tests examining differences between rural and urban
students on selected measures o f z-indifference showed no differences between the two
groups. The purpose o f this analysis rested on the theory that perhaps rural and urban
students had different response patterns when responding to the SII which requires
selection o f “Like,” “Indifferent,” or “Dislike” to answer each question. This part o f the
study is relevant to ideas about occupational stereotypes. For instance, Trice (1990)
suggested that rural children’s career interests are directly related to their parent’s careers.
Given the restriction o f career ranges in the rural environment, it is then likely that
interests may also be restricted causing subjects’ responses on the SII to also be affected.
However, this portion o f the study did not support this notion. It is possible that rural and
urban students do respond in a differential fashion with respect to “Like” and “Dislike”
answers. This aspect would be important to examine in future studies.
The hierarchical analysis o f variance used to examine mean differences between rural
and urban subjects on profile differentiation between the RIASEC scores yielded results
indicating no differences between the two groups or between genders. Again, it should be
noted that this sample was taken from students at a college counseling center. Further
research should include both counseled and non-counseled students.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
68
The orderings for the GOT scales found in this study were RECSIA for rural males.
ERCASI for urban males, SCEAIR for rural females, and SECAIR for urban females.
These findings are dissimilar to the orderings o f IASCRE and ISACER found by
Campbell (1971) and Holland (1973), respectively. Again, the orderings for this study
are likely influenced by low AC score which directly affect the scores on I and A.
The mean AC scores for the subjects in this sample are lower than those reported for
the norming sample (Hansen & Campbell, 1985). This is consistent with the observations
o f the researcher during actual career counseling sessions. The mean AC scores for this
sample were 27.6 and 32.6 for males and females respectively, and 28.4 to 31.5 for rural
and urban subjects, respectively. As noted by Hansen and Campbell, (1985) scores less
than 40 on AC are consistent with individuals who typically have only high school
educations or who would be more suited for two-year vocational or technical training.
Low scores are also indicators o f individuals who find intellectual pursuits boring, and
who are likely to drop out o f college. Given that this sample consisted o f career
counseling subjects who are in the midst o f career indecision, these results are not
surprising. It would be interesting to have tracked the educational path o f these subjects
to see if, indeed, there was any correlation between AC and whether or not the students
graduated from college. However, at the time o f this study, the latter information was not
available.
Also, a finding o f this study was the relatively low RIASEC scores, on the average, for
the subjects in the study. When RIASEC and AC scores are low, this often results in a
profile that is undifferentiated. This finding is consistent with the researcher’s
Reproduced with permission of the copyright owner. Further reproduction prohibited w ithout permission.
69
observations that the profiles she observed for the counseled students were, for the most
part, low and undifferentiated. Again, this could be due to the fact that all students were
coming to the counseling center in the midst o f career indecision. If so, this is
problematic, because the SII is designed to assist students in this process.
Even so, what accounts for the fact that this sample has relatively low scores? Is it
that the pattern o f interests for the subjects in this sample is somehow different from the
groups on which the test was normed? North Dakotans, upon which this sample is
largely based, have long argued that the whole state should be considered rural because o f
its small population. If that is the case, then perhaps the variables affecting career interest
development such as SES, parental occupation, and levels o f self-efficacy may have had
an effect here. It would be important in future research to include these variables. The
possibility also exists that studies conducted in other geographical regions would yield
different results. This suggests that the study results were potentially confounded by this
factor.
Yet another hypothesis about the lack o f mean differences between the rural and urban
groups may be related to the population number o f 50,000 which was selected, based on
the literature, to delineate between them. Possibly, this is not the most appropriate
dividing line between the groups. Since the researcher had coded the data for smaller
population numbers, it was possible to conduct some post hoc analysis to examine this
hypothesis further. Means and standard deviations for rural and urban subjects with
different breaks in population levels were examined, but, again, no apparent differences
were found.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
70
Obviously, this study was limited because the subjects were strictly from a college
counseling center population. An additional problem is that most subjects were from the
upper Midwestern region o f the United States. A further limitation is that the greatest
percentage o f participants where Caucasian. All factors affect the generalizability o f the
results to non-counseled college students, individuals from different geographic regions
and from different ethnicities or cultures. However, the purpose o f this study was to
explore whether there were differences on the SII for rural and urban populations.
Although no differences were found for this sample, and given the limitations and
exploratory nature o f this study, it is an indication that further research into potential
rural/urban differences should be conducted in other regions, with more ethnically diverse
subjects and for counseled and non-counseled subjects alike.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
REFERENCES
Anastasi, A. (1988). Psychological testing (6th ed.). New York: Macmillan.
Apostal, R. & Bilden J. (1991). Educational and occupational aspirations o f rural high
school students. Journal o f Career Development. 18(21. 153-159.
Banducci, R. (1970). Accuracy o f occupational stereotypes o f grade-twelve boys.
Journal o f Counseling Psychology. 17(61. 534-539.
Bandura, A. (1977). Social learning theory. Englewood Cliffs, N.J.: Prentice-Hall.
Bandura, A. (1986). Social foundations o f thought and action: A social cognitive theory.
Englewood Cliffs, N.J.: Prentice-Hall.
Bloch, T. & Rim, Y. (1979). The latent structure o f occupations. Journal o f Vocational
Behavior. 14. 145-168.
Broday, S. & Braswell, L. (1990). The relationship between academic comfort and other
Strong-Campbell Interest Inventory scales. Journal o f College Student Development.
31 (51.454-459.
Brown, D. & Brooks, L. (1984). Career choice and development. San Francisco, CA.:
Jossey-Bass, Inc.
Campbell, D. (1971). Handbook for the Strong Vocational Interest Blank. Stanford,
CA: Stanford University Press.
71
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
72
Campbell, D. (1974). Manual for the SVIB-SCII. Stanford, CA: Stanford University
Press.
Campbell, D. (1987). Strong-Campbell Interest Inventory, 4th ed. Journal o f Counseling
and Development. 66(1). 53-56.
Campbell, D., Crichton, L., Hansen, J., & Webber, P. (1974). A new edition o f the SVIB:
the Strong-Campbell Interest Inventory, Measurement and Evaluation in Guidance.
7(21 92-95.
Campbell, D., & Holland, J. (1972). A merger in vocational interest research: Applying
Holland's theory to Strong's data. Journal o f Vocational Behavior. 2, 353-376.
Crowley, A. (1979). Work environment preference and self-concepts. An investigation
o f Holland's theory. British Journal o f Guidance and Counseling. 7. 57-63.
Dawis, R., & Loftquist, L. (1984). A psychological theory o f work adjustment.
Minneapolis, MN.: University o f Minnesota Press.
Gati, I. (1979). A hierarchical model for the structure o f vocational interests. Journal of
Vocational Behavior. 15, 90-106.
Gottfredson, L. (1981). Circumscription and compromise: A developmental theory o f
occupational aspirations. Journal o f Counseling Psychology. 28(6). 545-579.
Hansen, J. (1987). Edward Kellog Strong, Jr.: first author o f the Strong Interest
Inventory. Journal o f Counseling and Development. 66(3). 119-125.
Hansen, J. (1992a). A note o f thanks to the women's movement. Journal o f Counseling
and Development. 70(4). 520-521.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
73
Hansen, J. (1992b). User’s guide for the Strong Interest Inventory (rev, ed.I. Palo Alto.
CA: Consulting Psychologists Press, Inc.
Hansen, J. & Campbell, D. (1985). Manual for the Strong Interest Inventory. Palo Alto.
CA: Consulting Psychologists Press, Inc.
Hansen, J. & Johansson. (1972). The application o f Holland's vocational model to the
Strong Vocational Interest Blank for Women. Journal o f Vocational Behavior. 2, 479-
493.
Hansen, J. & Stocco, J. (1980). Stability o f vocational interests o f adolescents and young
adults. Measurement and Evaluation in Guidance. 13. 173-178.
Hansen J. & Swanson, J. (1983). Stability o f interests and the predictive and concurrent
validity o f the 1981 Strong-Campbell Interest Inventory for college majors. Journal o f
Counseling Psychology. 30. 194-201.
Haviland, M. & Hansen, J. (1987). Criterion validity o f the Strong-Campbell Interest
Inventory for American Indian college students. Measurement and Evaluation
in Counseling and Development. 19(4). 196-201.
Henly, G. (1995). Modem test theory meets the venerable test: modeling responses to
the Strong Interest Inventory. Paper presented at the 1995 meeting o f the Psychometric
Society. Minneapolis, MN, June 17,1995.
Holland, J. (1959). A theory o f vocational choice. Journal o f Counseling Psychology. 6.
35-45.
Holland, J. (1973). Making vocational choices: a theory o f careers. Englewood Cliffs.
N.J.: Prentice-Hall.
Reproduced with permission of the copyright owner. Further reproduction prohibited w ithout permission.
74
Holland, J. (1985). Making vocational choices: a theory o f career. 2nd ed. Englewood
Cliffs, N.J.: Prentice-Hall.
Hollander, M., & Parker, H. (1972). Occupational stereotypes and self-descriptions:
their relationship to vocational choice. Journal o f Vocational Behavior. 2, 57-65.
lachan, R. (1984). A measure o f agreement for use with the Holland classification
system. Journal o f Vocational Behavior. 24. 133-141.
lachan, R. (1990). Some extensions o f the lachan Congruence Index. Journal o f
Vocational behavior. 36. 176-180.
Isaacson, L. (1985). Basics o f career counseling. Newton, M A.: Allyn and Bacon, Inc.
Johnson, R. (1969). Effectiveness o f SVIB academic interest scales in predicting college
achievement. Journal o f Applied Psychology. 53. 309-316.
Lapan, R . , McGrath, E., & Kaplan, D. (1990). Factor structure o f the Basic Interest
Scales by gender across time. Journal o f Counseling Psychology. 32(2), 216-222.
Lauver, P. & Jones, R. (1991). Factors associated with perceived career options in
American Indian, White, Hispanic rural high school students. Journal o f Counseling
Psychology. 3812). 159-166.
Lent, R., Brown, S., & Hackett, G. (1994). Toward a unifying social cognitive theory o f
career and academic interest, choice, and performance. Journal o f Vocational
Behavior. 45, 79-122.
Lowe, B. (1981). The relationship between vocational interest differentiation and career
undecidedness. Journal o f Vocational Behavior. 19.346-349.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
75
Lunneborg, P. (1977). Construct validity o f the Strong-Campbell Interest Inventory and
the Vocational Interest Inventory among college counseling students. Journal o f
Vocational Behavior. 10(2). 187-195.
Lunneborg, P. (1978). Sex and career decision-making styles. Journal o f Counseling
Psychology. 25. 299-305.
Mitchell, T., & Beach, L. (1976). A review o f occupational preference and choice
research using expectancy theory and decision theory. Journal o f Occupational
Psychology. 4 6 . 231-248.
Montoya, H. & DeBlassie, R. (1985). Strong-Campbell Interest Inventory comparisons
between Hispanic and Anglo college students: a research note. Hispanic Journal o f
Behavioral Sciences. 7(3). 285-289.
Murray, J. & Keller, P. (1991). Psychology and rural America: current status and future
directions. American Psychologist. 46(3). 220-231.
Poole, M. & Cooney, G. (1985). Careers: adolescent awareness and exploration o f
possibilities for self. Journal o f Vocational Behavior. 2 6 .251-263.
Poole, M., Langan-Fox, J., & Omodei, M. (1991). Career orientations in women from
rural and urban back-grounds. Human Relations. 44(9). 983-1005.
Raphael, K. & Gorman B. (1986). College women's Holland theme congruence: effects
o f self-knowledge and subjective occupational structure. Journal o f Counseling
Psychology. 33(2). 143-147.
Schneider, L. & Overton. T. (1983). Holland personality types and academic
achievement. Journal of Counseling Psychology. 30(2). 287-289.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
76
Smart, J. (1989). Life history influences on Holland vocational type development.
Journal o f Vocational Behavior. 3 4 .69-87.
Spokane, A. (1979). Occupational preference and the validity o f the Strong-Campbell
Interest Inventory for college women and men. Journal o f Counseling Psychology.
26(41. 312-318.
Super, D. (1953). Theory o f vocational development. American Psychologist. 8.
185-190.
Super, D. (1957). The psychology o f careers. New York: Harper & Row.
Super, D. (1980). A life-span, Iife-space approach to career development. Journal o f
Vocational Behavior. 16.282-298.
Swanson, J. (1992). The structure o f vocational interests for African-American college
students. Journal o f Vocational Behavior. 40(21. 144-157.
Swanson, J. & Hansen, J. (1985). The relationship o f the construct o f academic comfort
to educational level, performance, aspirations, and prediction o f college major choices.
Journal o f Vocational Behavior. 26( 11. 1-12.
Taylor, K. & Betz, N. (1983). Applications o f self-efficacy theory to the understanding
and treatment o f career indecision. Journal o f Vocational Behavior. 2 2 .63-81.
Tomlinson, S. & Evans-Hughes, G. (1991). Gender, ethnicity, and college students'
responses to the Strong-Campbell Interest Inventory. Journal o f Counseling and
Development. 70( 1). 151-155.
Trice, A. (1990). Stability o f children's career aspirations. Journal o f Genetic
Psychology. 152(11. 137-139.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
77
Tryon, G. (1983). Differentiation between counseled and non-counseled students on the
General Occupational Themes o f the Strong Campbell Interest Inventory. Journal o f
College Student Personnel. 24(11. 51 -54.
Wagman, M. (1971). Clinical and research use o f the SVIB Academic Achievement
scale. Journal o f Counseling Psychology. 1 8 .337-340.
Wallace G. & Walker, S. (1990). Self concept, vocational interests, and choice o f
academic major in college students. College Student Journal. 23(41.361-367.
Wheeler, K. (1983). Comparisons o f self-efficacy and expectancy models o f
occupational preferences for college males and females. Journal o f Occupational
Psychology. 56. 73-78.
Wheeler, K., & Mahoney, T. (1981). The expectancy model in the analysis o f
occupational preference and occupational choice. Journal o f Vocational Behavior. 19.
113-122.
Whetstone, R. & Hayles, R. (1975). The SVIB and Black college men. Measurement
and Evaluation in Guidance. 8(2). 105-109.
Wright, J. (1976). The SVIB academic achievement score and college attrition.
Measurement and Evaluation in Guidance. 8(4). 258-259.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
