See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/324776171

Development and Validation of the Subjective Occupational

Aspiration Scale Using a Rasch Model Approach

Article in Journal of Career Development · April 2018

DOI: 10.1177/0894845318771210

CITATIONS READS

2 1,130

3 authors, including:

Jay Rojewski Minho Kwak

University of Georgia University of Georgia
89 PUBLICATIONS 2,158 CITATIONS 15 PUBLICATIONS 78 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Minho Kwak on 10 June 2019.

The user has requested enhancement of the downloaded file.

 Journal of Career Development
1-22
ª Curators of the University
Development and Validation of of Missouri 2018
Reprints and permission:
the Subjective Occupational sagepub.com/journalsPermissions.nav
DOI: 10.1177/0894845318771210

Aspiration Scale Using a Rasch journals.sagepub.com/home/jcd

Model Approach

Hyojung Han1, Jay W. Rojewski2, and Minho Kwak2

Abstract
This study proposes and measures a new construct, subjective occupational aspiration, that refers to
individuals’ preferences for selected aspects of occupations—economic rewards, authority, freedom,
education required, and ability required—that, collectively, support the expression of those occu-
pations. Using a four-step approach to constructing psychometric instruments based on the Rasch
model, this study developed and validated the Subjective Occupational Aspiration Scale (SOAS). The
SOAS consists of 24 4-point Likert-type items. Both infit and outfit statistics indicated that the Rasch
model predicted both items and persons within a reasonable error bound, validating all initial 24 items
of the SOAS. However, two sets of items showed reversed difficulties. The final instrument that
reflects the corrected difficulty order of items is provided. Suggestions and implications for using the
SOAS in practice are also discussed.

Keywords
subjective occupational aspiration, instrument development, item response theory, Rasch model, test
reliability/validity

Introduction
Expressed career-related goals or choices under ideal circumstances reflect occupational aspirations
(Rojewski, 2005). Occupational aspirations are often contrasted with realistic aspirations, or occupa-
tional expectations, which refer to occupations people realistically expect to enter accounting for per-
ceived barriers or other influences. Occupational aspirations, on the other hand, reflect one’s
vocational desire or preference without accounting for limiting factors such as education, money, abil-
ity, or talent.

1
Korean Educational Development Institute, Chungcheongbuk-do, South Korea
2
University of Georgia, Athens, GA, USA

Corresponding Author:
Minho Kwak, University of Georgia, 2110 Carlton St., Aderhold hall 126C, Athens, GA 30602, USA.
Email: [email protected]
2 Journal of Career Development XX(X)

While aspirations are not necessarily indicators of eventual career attainment, they hold consider-
able psychological meaning and have some predictive value in identifying future career behavior
(Holland & Gottfredson, 1975). Occupational aspirations may reflect elements of an individual’s
self-concept, his or her perceptions about the opportunities available for education/training and work,
and personal interests and hopes (Rojewski, 2005). Thus, aspirations are important to career develop-
ment and occupational attainment in that they can effectively inform individuals and counselors about
possible career paths, prompt (or inhibit) planning, guide learning, help organize life choices, and con-
tribute to the preparation for adult life.

Theory
The development, nature, and expression of occupational aspirations have been explained from several
perspectives including developmental theory (Super, Savickas, & Super, 1996), Gottfredson’s (1996)
theory of occupational aspirations development, social cognitive career theory (SCCT; Lent, Brown, &
Hackett, 1994), and status attainment theory (Hotchkiss & Borow, 1996). To varying degrees, each
theory acknowledges that aspirations represent a combination of psychological development, external
influences, and learning that reflects an individual’s assessments of personal capabilities and available
opportunities.
Super’s (1990; Super et al., 1996) theory of career development provides a general understanding of
how occupational aspirations are formed and their importance or role in defining a career path. In chil-
dren, aspirations often reflect fantasies about glamourous or exciting careers but also represent an
emerging awareness of the world of work. Adolescence, the exploration stage, is characterized by a
progressive narrowing of career options from fantasy to tentative options to final career decisions.
Super (1990) posited that self-concept plays a critical role in career development. In fact, career choice
is seen as the development and implementation of occupational self-concept, which is reflected in
expressed occupational aspirations.
Gottfredson (1996) provided a more detailed theory about the development and role of occupational
aspirations in describing prominent influences on the process of narrowing and refining career options.
This longitudinal process is shaped by experiences defined by size and power (3–5 years of age), sex
roles (6–8 years of age), social valuation (9–13 years of age), and unique self (14 years of age and
older). These influences inform individuals about appropriate or desirable work options. As aspirations
are formed, individuals engage in two decision-making processes. The first, circumscription, involves
the progressive elimination of occupations deemed unacceptable. The second, compromise, occurs as
individuals begin to relinquish their most preferred occupations for less compatible, but more acces-
sible, ones. In the end, this iterative process results in career attainment that represents the joint product
of an individual’s assessment of job compatibility and accessibility.
SCCT (Lent, 2005; Lent et al., 1994; Lent, Hackett, & Brown, 1996) is also useful in understanding
the nature and role of aspirations in career behavior and attainment. SCCT emphasizes four key con-
cepts: self-efficacy, outcome expectations, goals, and contextual support and barriers. Aspirations, or
goals, are viewed as one’s determination to engage in particular activities or to produce particular out-
comes. Two elements influence these goals. The first, self-efficacy, reflects one’s beliefs about his
ability to attain a specific outcome. The second, outcome expectations, refers to one’s beliefs about
the consequences of given actions. Contextual supports or barriers refer to identified supports or bar-
riers that accompany the pursuit of any goal. SCCT proposes that educational and career goals are
influenced by beliefs that one develops and refines over time including personal performance, vicar-
ious learning, and special persuasion.
Status attainment theory (Hotchkiss & Borow, 1996) explains occupational aspirations from the
perspective of social stratification. Aspirations are formed at an early age by the opportunities or
barriers presented to individuals through external factors such as bias, discrimination, cultural
Han et al. 3

expectations, societal attitudes, and stereotypes based on gender, race/ethnicity, and social class. A
prominent difference between status attainment theory and psychological career theories is that the
former assigns greater weight to institutional and market forces on constraining career decisions and
attainment (Johnson & Mortimer, 2002).

Measurement of Aspirations
From a conceptual perspective, occupational aspirations are typically viewed as a unidimensional con-
struct that is qualitative in nature. They are determined by asking individuals to identify the particular
occupation that is desired if free of barriers or limiting factors. Despite the qualitative nature of the
concept, most researchers measure and analyze occupational aspirations from a quantitative perspec-
tive. In doing so, they employ existing occupational categorization schemes or other scales such as
occupational prestige hierarchies or socioeconomic indexes (SEIs) to organize and analyze them
(Rojewski, 2005). SEIs (e.g., Blau & Duncan, 1967; Ganzeboom, DeGraaf, & Treiman, 1992; Hauser
& Warren, 1996; Nakao & Treas, 1992; Stevens & Cho, 1985) are most frequently used to calculate
scores for occupations that reflect a combination of income, educational attributes, and social prestige.
Despite the popularity of occupational prestige hierarchies and SEIs in many social science fields,
this measurement approach is limited in several ways. First, expressed career goals are inevitably
restricted by an individual’s environment and accessible information about the world of work.
Researchers (e.g., Furlong & Cartmel, 1995; Schoon & Parsons, 2002) have argued that the formation
of occupational aspirations are significantly affected and constrained by an external opportunity struc-
ture that includes local work opportunities. As a result, measuring occupational aspirations using exist-
ing schemes runs the risk of having aspirations reflect environmental backgrounds and previous
experiences rather than true occupational preference. Second, using existing schemes casts a sociolo-
gical perspective on research even when it is unnecessary or inappropriate, since these schemes were
originally developed using socioeconomic indicators. When occupational aspirations are quantified
using an SEI, the aspiration is not so much a personal reflection of how the job is viewed but of how
society views the job’s desirability and status. Given this fact, it can be problematic for researchers
who are unaware of this discrepancy and associate occupational aspirations with educational and psy-
chological variables without needed caveats. Third, existing schemes can be inexact because they do
not always properly specify differences within occupations. For instance, an occupational aspiration of
movie star would be assigned to a category like “actors and directors,” which has a moderate prestige
score on the SEI (Hauser & Warren, 1996). This score is based on income and social prestige levels of
average actors and directors. However, income or societal recognition attributed to a highly successful
actor (i.e., a movie star) is starkly different than for average actors. Fourth, existing schemes cannot
always be updated in a timely way to reflect the current or emerging picture of a fast-changing society.
Indeed, revising and updating SEIs generally involve extensive amounts of work that can take years to
complete (Ganzeboom, 2010). The most recent U.S.-based SEI available at the time of this study was
developed by Hauser and Warren (1996). Jobs that have appeared since the late 1990s (e.g., social
media manager) are absent from that SEI. Fifth, using existing schemes is, in practical terms, ineffi-
cient because it can be time-consuming and/or less precise when coding aspirations to match estab-
lished occupational categorizations.
While beyond the scope of this study, we note that using these measurement approaches can also be
unreliable for occupational expectations. Occupational expectations and aspiration are often measured
together in the same survey by asking respondents to express particular occupations they expect to rea-
listically enter given limiting factors such as education, money, and ability and occupations they hope
to ideally enter without considering such limiting factors. Thus, the difference between the two con-
structs is supposed to reflect the limiting factors that individuals perceive as keeping them from
achieving their aspired jobs. However, in the current measurement scheme, the difference often
4 Journal of Career Development XX(X)

represents one’s perceived psychological proximity to occupational opportunities, not limiting factors.
For instance, one may express an occupational aspiration of photographer because of interest in photo-
graphy and an occupational expectation of surgeon because of enrollment in medical school. In such
cases, aspirations may only indicate an individual’s vocational interest (which may be instantaneous or
fleeting) or psychological remoteness to a particular job, while expectations reflect familiarity or prox-
imity. This type of response/interpretation is frequently observed in real cases.
Gray and O’Brien (2007) criticized common approaches to understanding (and measuring) occupa-
tional aspirations for not distinguishing among varying levels of aspirations toward positions of lead-
ership within specified occupations. To address this issue, they developed the Career Aspiration Scale
(CAS) to measure aspirations toward leadership and further education within occupations rather than
merely aspirations toward the occupations themselves. The CAS provides a useful conceptualization
and measurement of aspirations related to work with occupational field controlled (i.e., varying posi-
tions within the same basic occupation). However, measurement approaches using occupational cate-
gorization (prestige) schemes still dominate research. Furthermore, while the CAS provides an
alternative way to focus on aspirations, our understanding of aspirations related to occupations is still
relatively crude. Therefore, a need exists to improve the measurement of aspirations related to occu-
pations and to the elements that support the selection of specific occupations.
This study reports on our efforts to develop a scale that assesses individuals’ subjective occupa-
tional aspirations. Subjective occupational aspirations refer to individuals’ preferences for selected
aspects of an occupation that, collectively, support the expression of that occupation. People’s generic
perceptions of a specific occupation are typically affected by five aspects represented by that occupa-
tion—(1) economic rewards, (2) degree of authority, (3) freedom of action, (4) education required, and
(5) academic ability required (Gottfredson, 1996; Rojewski, 2005). Occupational aspirations reflect
actual jobs; subjective occupational aspirations reflect the strength of individuals’ desires to achieve
on each of these five significant aspects.
Compared to existing approaches to measuring occupational aspirations, the Subjective Occupa-
tional Aspiration Scale (SOAS) was designed to offer several advantages. Primarily, the results of the
SOAS reflect individuals’ aspirations toward their general preferences for work and life without being
restricted by existing occupational schemes that accompany the aforementioned limitations. Second,
the SOAS does not reflect constrained or stereotypical knowledge of the world of work. Rather, knowl-
edge gleaned from the SOAS may provide respondents and counselors with a clearer idea of actual
preferences and orientations because of its focus on the foundational elements that ultimately define
occupations. Third, in practical terms, the SOAS may be more inclusive and useful for counseling and
exploration activities than the identification of a single occupation, as results can be connected to a
number of different occupational possibilities.

Developing the SOAS Using a Rasch Model Approach

Construction of the SOAS was accomplished using item response theory (IRT), viz., the Rasch model.
Although factor analytic methods are predominant in the development and analysis of career assess-
ment measures, IRT models represent several advantages in scale development over traditional factor
analytic models. First, factor analytic models have been criticized because their methods may be based
on an erroneous assumption. Researchers (e.g., Hambleton, Jaeger, Plake, & Mills, 2000; Miguel,
Silva, & Prieto, 2013) have argued that one of the fundamental assumptions of factor analytic mod-
els—that a linear relationship exists between construct and sub-factors—may not be accurate for a
majority of constructs in psychosocial investigations. Second, traditional models are mainly concerned
with the linearly limited structural and test-level features of a test (i.e., reliability) and offer little or
limited item-level information. IRT approaches, on the other hand, provide extensive item-level infor-
mation, yielding better descriptions of item performance such as item difficulty1 (Hambleton et al.,
Han et al. 5

2000; Miguel et al., 2013). Therefore, IRT approaches make it possible to construct a more precise
instrument by developing a series of individual items with varying levels of difficulty to better fit
developers’ intentions. Third, IRT models enable more concrete interpretation of the test result. Par-
ticularly, Rasch analyses produce a Wright map where, using a logit scale, items and people are located
together on the same continuum of the measurement construct (Engelhard, 2013). Items are aligned on
a continuum based on difficulty and people with their ability. Thus, scores can be interpreted in a more
tangible way by comparing participants’ ability with specific capabilities or traits manifested through
individual items.
Engelhard (2013) suggested a four-step approach when applying a Rasch model. This scheme iden-
tifies domains to consider when developing psychometric instruments including (1) the latent variable,
(2) an observational design (items used to measure the latent variable), (3) the scoring rubric, and (4)
the Rasch measurement model. In the following sections, we describe each domain and explain how
they were applied during development of the SOAS.

Step 1: The latent variable. The first domain, the latent variable, involves developing a prototypical
variable map in which the construct being measured is defined and conceptualized on a unidimen-
sional continuum. This continuum represents the qualitative order (e.g., low to high) of the construct
being measured. On the continuum, different levels of the construct (trait) are specified, and each
level is defined with typical responses (manifestations) that may be expected from persons occupy-
ing that level.
We defined subjective occupational aspiration as a unidimensional construct that reflects individ-
uals’ preferences for selected aspects that, collectively, support the expression of an occupation. Then,
a prototypical variable map was developed. In the map, three levels (high, medium, and low) of
subjective occupational aspirations were assumed. In each level, typical responses (psychological or
behavioral) of persons possessing that level of subjective occupational aspiration were described. For
high-level responses, an orientation of “one aspires to very high/high achievement in selected dimen-
sions of work” was expected. For the medium level, responses of “one aspires to average achievement
in selected dimensions of work” were expected. For the low level, responses of “one aspires to achieve
in selected dimensions of work is low/very low” were expected. These descriptions were used as stan-
dards when developing individual items with different levels of difficulties.

Step 2: Observational design. The second domain generates an observational design, a set of items
that can collect responses or draw observations from respondents. Two components are typically
constructed, content and levels. Content is related to what respondents are expected to know or
show regarding the measurement construct. Levels reflect the vertical dimension of the construct
(e.g., low to high) and are applied to each content of the measurement construct. Then, research-
ers develop individual items that correspond to each intersection between the layers of contents
and levels.
For the SOAS, content was represented by the five elements that formed the definition for subjec-
tive occupational aspirations. The decision on number and content of elements was based on a review
of the literature on occupational aspirations. We first established that subjective occupational aspira-
tions represented a construct with a vertical dimension, so a possible range of aspirations from low to
high existed. Next, we focused on literature that addressed the common elements constituting levels or
perceptions of occupations.
Rojewski (2005) explained that the level of an occupational aspiration has historically been decided
by “a vertical dimension of classification based on occupational prestige” (p. 135) and that typical indi-
cators of occupational prestige include economic rewards, degree of authority, freedom of action,
amount of education required, and intelligence and task complexity. Similarly, Gottfredson (1996)
asserted that:
6 Journal of Career Development XX(X)

prestigious occupations require more education, pay more money, are more complex, exercise authority,
and require higher levels of ability, whereas occupations low in prestige require little education, pay poorly,
involve simple tasks, are directed by others, and require little cognitive ability. (p. 68)

Crawley and Cardinale (2008) developed the Perception of Occupational Status Survey (POSS) to
measure individuals’ perceptions of prestige or status of occupations. The POSS consists of items that
reflect various aspects of a chosen occupation that typically affect one’s perception of the prestige or
status of an occupation (Crawley, 2014). Occupational aspects in the POSS include social recognition,
required education, required commitment, difficulty of job tasks, required intelligence and skills, mon-
etary rewards, selectivity of entering, degree of authority, and degree of dishonesty.
Among the elements identified by the above studies that constitute the level of occupations, five
specific elements were commonly found: (1) economic rewards, (2) degree of authority, (3) freedom
of action, (4) education required, and (5) academic ability required. Researchers have consistently sug-
gested that occupational levels possess these five common defining aspects. Thus, we selected these
five factors to constitute the main components of the subjective occupational aspirations, that is, sub-
jective occupational aspirations are constituted by the combination of one’s preferences (or achieve-
ment aspirations) toward the five selected dimensions.
The SOAS is comparable with the Minnesota Importance Questionnaire (MIQ; Rounds, Henly,
Dawis, Lofquist, & Weiss, 1981) in that both measures specify elements of work. The MIQ identifies
20 types of individual needs that must be met to gain satisfaction in work. The needs can also be con-
sidered as conditions of work or individuals’ work values. Some elements in the MIQ (e.g., compen-
sation, authority, independence, responsibility, and ability utilization) parallel the elements of the
SOAS because both are concerned with what people seek in work. However, the two constructs—sub-
jective occupational aspirations and work values—essentially differ because subjective occupational
aspiration focuses only on aspects that define the levels of occupations. Values included in the MIQ
such as coworkers, creativity, and variety are not generally thought of as defining occupational level
and are not the focus of subjective occupational aspirations. Furthermore, the SOAS seeks to measure
the degree to which one aspires to achieve in the five occupational domains, while the MIQ is con-
cerned with ranking one’s priorities among 20 basic work needs.
Next, the levels of subjective aspirations that constituted the vertical dimension were addressed.
Three levels were selected for the prototypical variable map: high, medium, and low. Given the five
elements and three levels, 15 possible combinations of each element and level were possible. We
developed 2–5 items for each combination, resulting in 45 initial items (see Table 1). A subsequent
review by the authors and external consultants for validity resulted in the removal of 21 items. Item
removal was guided by seven criteria that were developed by the authors and consultants prior to and
during the process: (1) The item statement is unclear, vague, misleading, or difficult to understand; (2)
the topic of the item does not properly represent the target content; (3) the topic of the item does not
properly represent the target level; (4) the item is not in line with other items within the same content
and does not represent a gradual increment of the content; (5) the item is not useful because its topic is
too far from reality or does not reflect the actual world of work; (6) the item does not truly measure
subjective occupational aspirations, but some other construct; and (7) the topic of the item, per se, is
fine, but the item is unable to elicit responses in a way that can appropriately reflect the target level. In
the end, 24 items were selected for the final scale. Each item was designed to reflect a specific element-
level component of subjective occupational aspirations.

Step 3: Scoring rubric. The third step in IRT Scale development concerns categorizing responses and
designing a scoring rubric. Respondents’ answers are categorized to represent meaningful increments
in the level of the construct reflected in each item. Then, scoring rules, such as dichotomous or poly-
tomous schemes, are determined. For the SOAS, it was assumed that respondents’ degree of agreement
Han et al. 7

Table 1. Initial Items of the Subjective Occupation Aspirations Scale.

Final Item
Reason of Code of
Content Level Initial Items Removala Selected Items

Economic Low Money is the last thing that I seek in my future job or career 6 —
rewords I want my future job to pay enough for me to purchase basic life — ER1
(ER) necessities
I want my future job to pay enough for me to live at least an — ER2
average life
Medium I want my future job to pay enough for me to buy nice things 1 —
I aspire to a job that pays fairly well, allowing me to purchase — ER3
high-quality goods and services and live a comfortable life
In the future, I would like to have a job that pays as well as jobs 1, 4 —
of people around me
High I want my future job to pay higher than average wages 3, 4 —
I aspire to a job that pays very well, allowing me to purchase — ER4
highly luxurious houses, cars, and goods and live an
immensely comfortable life
I aspire to a job that pays extremely well and provides the — ER5
greatest comforts and luxury in life but only few people can
obtain
Among other factors, monetary rewards are most important in 6 —
choosing my future job and career
Degree of Low I will be satisfied even if my future work is not something 1, 6, 7 —
authority influential in society as long as the work pleases me and gives
(DA) meaning to my life
In my future job, I do not want to be controlled by other people — DA6
for every little thing I do
I aspire to a job where I have control of my daily work — DA7
Medium I aspire to a job where I can make decisions that influence the — DA8
future of my organization and can manage many people in the
organization
In my future job, I want to be respected by colleagues and other 2 —
members of the organization
In my future job, I want to have authority to manage, train, and 1 —
evaluate many people in the organization
High I aspire to a job where I have an influence on a state-size area or — DA9
a nationwide large group of people
In the future, I want to have a job that is looked up to by most 2 —
people in society
I aspire to a job where my decisions affect the whole country or — DA10
most people in the country
Freedom of Low I aspire to a job that allows me to take a day off if I am sick or — FA11
action injured
(FA) I aspire to a job that allows me to easily take vacations on official — FA12
leave days or national holidays
I want my future job to have a strict work schedule and detailed 6, 7 —
task instructions
Medium I aspire to a job that allows me to have enough time for my — FA13
personal, nonwork life
In the future, I want a job that offers a condition in which I can 1, 2 —
pursue not only work but also family successfully

(continued)
8 Journal of Career Development XX(X)

Table 1. (continued)

Final Item
Reason of Code of
Content Level Initial Items Removala Selected Items

High
I want my future job to offer an independent work environment 2, 4
such as a private office
In my future job, I want to be able to adjust my working hours — FA14
flexibly depending on my personal schedule
In my future job, I do not want to belong to an organization but 4 —
wish to work independently as a freelancer
I want to be able to take long leaves (years) from my future job 6
occasionally, even if doing so may have negative effects on my
status or salary
I aspire to a job where I can freely choose what to perform as — FA15
daily tasks
Education Low I do not want my future work to be something that anyone with 1 —
(EDU) no training or education can do
required I want my future job to require at least elementary education 7 —
I want my future job to require education equal or higher than — EDU16
compulsory education (K–12).
Medium I aspire to a job that requires at least a 2-year college degree 4 —
I aspire to a job that requires at least a 4-year college education — EDU17
High I aspire to a job that requires an international education such as 3, 4 —
a foreign degree
I aspire to a job that requires either graduate degrees or — EDU18
professional certificates such as lawyer or doctor
I aspire to a job that requires an elite education that only few — EDU19
select people possess
Academic Low I want my future job to involve tasks that require at least basic — AA20
ability literacy skills
(AA) I want to have a job that involves at least a medium level of — AA21
required academic difficulty such as basic reading and math skills
I want a job that a person of average intelligence can perform 1, 7 —
without difficulties
Medium I aspire to a job that involves tasks that require intelligence of at 1, 2 —
least a college-educated person
I aspire to a job that involves fairly complex tasks that only a — AA22
person of average intelligence or higher can complete
High I aspire to a job that involves very complex tasks that can be — AA23
performed by only a very intelligent person
I aspire to a job that involves extremely complex tasks that can — AA24
be performed by only a few people (less than 5% of people)
I aspire to a job that only a few people with highly specialized 1, 2, 4 —
skills can enter
a
(1) The item statement is unclear, vague, misleading, or difficult to understand, (2) the topic of the item does not properly
represent the target content, (3) the topic of the item does not properly represent the target level, (4) the item is not in line with
other items within the same content and does not represent a gradual increment of the content, (5) the item is not useful
because its topic is too far from reality or does not reflect the actual world of work, (6) the item does not truly measure
subjective occupational aspirations, but some other construct, and (7) the topic of the item, per se, is fine but the item is unable
to elicit responses in a way that can appropriately reflect the target level.
Han et al. 9

with each item statement would reflect their levels of subjective occupational aspiration. To capture
different levels of agreement, a 4-point, Likert-type scale was used that directed respondents to select
their degree of agreement with each item statement.

Step 4: The Rasch measurement model. In the last step, the Rasch measurement model was used to exam-
ine the psychometric properties of the scale. The quality of the instrument was examined by evaluating
the fit of persons and items to the requirements of the Rasch model. The locations of persons and items
fitted by the Rasch model (i.e., Wright map) were also evaluated to determine the fit with the hypothe-
sized prototypical variable map developed in previous stages.

Method
Participants
A particular target population was not assumed for the SOAS in terms of age, gender, or race ethnicity.
However, as an exploratory tool, the SOAS may be particularly beneficial for adolescents and young
adults making important decisions about their future careers. Thus, an initial validation of the SOAS
was conducted using data from college students.
From a large 4-year public university in the Southeast United States, 245 undergraduate students
from 10 elective courses were recruited. A sample size of 245 satisfies the minimum sample size
requirements for Rasch analyses (Embretson & Reise, 2000). Diversity in majors and grade years were
considered by recruiting from both lower and upper level elective courses where students with diverse
backgrounds could be accessed. Of 245 participants, 100 (40.8%) were men and 145 (59.2%) were
women. The average age and GPA of participants were 20.02 and 3.49, respectively. Participants were
composed of 65 (26.5%) freshmen, 55 (22.4%) sophomores, 61 (24.9%) juniors, 50 (20.4%) seniors,
and 12 (4.9%) fifth-year students.

Measures
Subjective occupational aspirations. Subjective occupational aspirations were measured using the SOAS.
The SOAS comprises five content components—economic rewards, degree of authority, freedom of
action, education required, and academic ability required—that contribute to the formation of desirable
aspects of occupations. The instrument included 24 items that asked respondents to indicate their level
of agreement with each item statement. Items were scored using a 4-point Likert-type scale (1 ¼
strongly disagree, 2 ¼ mildly disagree, 3 ¼ mildly agree, and 4 ¼ strongly agree). A composite score
of responses to all 24 items indicates respondents’ preferences for and desires to achieve the composite
of the five occupational aspects.

Socioeconomic occupational aspirations. Participants were asked to indicate an occupation most desirable
to them without considering limiting factors such as money, education, or talent. Answers were later
coded using the SEI developed by Hauser and Warren (1996), the most recent U.S.-based SEI avail-
able. Hauser and Warren’s (1996) SEI provides socioeconomic scores for 501 occupations based on
average earnings and education levels of the occupation as reported in the 1990 Census (U.S. Bureau
of the Census, 1992) as well as prestige ratings calculated by Nakao and Treas (1994). Scores in the
Hauser and Warren’s (1996) SEI range from shoe machine operator (7.13) to physician (80.53). Par-
ticipants’ specified occupations were placed into categories defined by the SEI and then scored using
SEI ratings. Occupations not specifically identified by the SEI were assigned a score of the most
comparable occupational category.
10 Journal of Career Development XX(X)

Data Analysis
SPSS (Version 20.0; IBM Corporation, 2012) was used to analyze descriptive statistics. Prior to Rasch
analyses, we examined data for meeting assumptions of unidimensionality and local independence by
looking at the portion of variance of the SOAS explained by the Rasch model. In general, if the Rasch
model explains more than 20% of the total variance of the data, it is assumed that both assumptions are
satisfied and subsequent analyses can be performed within the Rasch measurement perspective
(Walker, Engelhard, & Thompson, 2012). For our sample, the Rasch model was able to explain
49.46% of the total variance of the data, suggesting the presence of unidimensionality and local inde-
pendence in the SOAS.
Rasch analyses were then performed using the computer program FACET (Version 3.71.3; Linacre
& Wright, 1988). We used the rating scale model (RSM), a model generally used when all items within
an instrument have the same number of polytomous response choices (Engelhard, 2013). The RSM is
an extension of the Rasch model for polytomous items and is provided by:
 
Pnik
log ¼ yn  di  t k ;
Pniðk1Þ
where Pnik is the probability of participant n endorsing answer choice k on an item i; Pniðk1Þ is the
probability of participant n endorsing answer choice k  1 on an item i, yn is the level of career aspira-
tion of nth participant, di is the difficulty of endorsing item i, and tk is a threshold of the k rating com-
pared to k  1 rating. Also, for the estimation process, the joint maximum likelihood estimation
approach was applied.
Rasch analyses estimated the person separation reliability to evaluate the reliability of the SOAS.
This statistic, analogous to coefficient a (Cronbach, 1951) in factor analytic models, indicates how
efficiently the developed set of items separates respondents according to their level of subjective occu-
pational aspiration. A person separation reliability value of over .7 is considered acceptable (Walker
et al., 2012).
Then, infit and outfit statistics for both items and persons were examined. Fit values for items assess
how well the Rasch model predicted individual items’ functioning. The outfit statistic, also known as
an unweighted mean square, reflects outlier observations that are far off from what is expected by
imposing penalties on those outliers. Conversely, the infit statistic, or weighted mean square, is sen-
sitive to observations that were taken into account (or predicted) by the Rasch model. The infit statistic
usually goes unreported because it approximates the outfit when data are stratified by category
(Linacre, 2014).
Typically, a reasonable range for outfit (and infit) values is from 0.6 to 1.4. Values above 1.4 indi-
cate that more unexpected variance exists (40% or more) than predicted by the model. On the other
hand, values under 0.6 indicate that the data have less variance (40% or less) than the model predicted
(Walker et al., 2012; Wright, Linacre, Gustafson, & Martin-Löf, 1994). From a conservative perspec-
tive, outfit statistics below 0.6 or above 1.4 are considered to be model misfits, which indicate that the
Rasch model does not fit the data collected from the constructed instrument well. However, in applied
research, it is known that an outfit value below 0.6 or between 1.4 and 2.0 causes no real problems
(Linacre, 2004; Smith, 1996).
Infit and outfit statistics can also be estimated for persons. Fit values for persons are used to exam-
ine the quality of the measurement system. Essentially, the interpretations and mathematical properties
of the values are the same as those for items. However, Wright, Linacre, Gustafson, and Martin-Löf
(1994) suggested a more lenient criterion for the person parameters because unlike item parameters,
which are estimated using data from a large number of participants, the person parameters are esti-
mated from a limited number of items. Although the literature lacks any extensively agreed-upon
Han et al. 11

criterion for the person parameter, prior studies (e.g., Miguel et al., 2013) have considered it to be a low
misfit rate when the portion of their sample showing misfits was less than 20%.
Next, item discrimination of the SOAS was calculated with a point-biserial correlation of items.
Item discrimination refers to the ability for items to distinguish between participants with different lev-
els of a latent trait. The point-biserial correlation represents the association between an item score and
the total instrument score. An item with a positive point-biserial correlation means that participants
who answered high on that item are likely to have higher total scores on the instrument, while parti-
cipants who answered low on the item are likely to have lower total scores. In other words, items with
positive, and ideally high, point-biserial correlations are assumed to discriminate participants’ abilities
well. On the other hand, a negative point-biserial correlation means poor item discrimination. In gen-
eral, positive point-biserial correlations are desirable. Specifically, point-biserial correlations values
larger than.15 are recommended (Varma, 2006). Lower or negative point-biserial correlation of an
item suggests further examination is needed on that item in terms of its wording, presentation, or
contents.
Item difficulty was also examined. Item difficulty indicates the level of the latent trait represented
by an item. Simultaneously, the RSM model estimates participants’ level of the latent trait (or ability)
based on their responses. Both item difficulties and person abilities are estimated using a logit scale,
enabling researchers to directly compare participants with items. Item difficulties and participants’
abilities can be displayed together on a Wright map. On the Wright map, items and respondents are
located on a continuum of the latent trait.
Throughout the analyses, a listwise deletion method was applied to handle missing data. This was
based on Schafer’s (1999) assumption that a missing rate of 5% or less is inconsequential. Our data
showed a missing rate of 2% (only 4 of the 249 participants did not complete the entire survey).

Results
Descriptive Statistics
In Table 2, means and standard deviations (SDs) of the SOAS and its content components are pre-
sented. Possible SOAS scores ranged from 24 to 96, which were transformed into 4-point scale scores
for analyses. The overall data exhibited a tendency toward negative skewness with a mean of 3.19.
Among the five content components, freedom of action showed the highest mean, while degree of
authority showed the lowest.

Rasch Analyses
In estimation of the Rasch model, the expected value of item difficulty parameters was fixed at 0. The
person separation reliability was .81, indicating good reliability.

Infit and outfit statistics and item discrimination. Fit values for persons were examined first. Of 245 parti-
cipants, 42 students showed outfit values higher than 1.4. These participants with misfits constituted
15.9% of the whole sample, which is considered low (Miguel et al., 2013). Next, fit values for items
were explored (see Table 2). The means of infit and outfit values were 1.07 and 1.06, respectively. All
items, except for Items 1, 2, and 20, showed outfit values that fell into the acceptable range of 0.6–1.4.
Outfit values of Items 1, 2, and 20 were 1.80, 1.44, and 1.55, respectively, slightly exceeding the
acceptable range. However, they were lower than 2.0, suggesting they were neither problematic nor
distorted the measurement system. Table 2 also displays point-biserial correlations of items of the
SOAS. Two items, Items 1 and 11, exhibited point-biserial correlations that were smaller than the rec-
ommended criterion value of .15.
12 Journal of Career Development XX(X)

Table 2. Means, Standard Deviations, Infit and Outfit Statics, Point-Biserial Correlations, and Item Difficulties of
the Subjective Occupational Aspiration Scale.

Fit
Descriptive Statistics Statistics Item Difficultya

Average
Content Point-
Content M M SD Item Component Biserial
Components Items (Item) (Component) (Component) Infit Outfit Difficulty Difficulty Correlation

Degree of 10 (DA) 2.05 2.89 .526 1.20 1.22 2.00 .67 .33
authority 9 (DA) 2.51 0.99 1.03 1.33 .38
(DA) 8 (DA) 3.19 0.78 0.81 0.24 .39
6 (DA) 3.26 0.79 0.84 0.11 .29
7 (DA) 3.47 0.67 0.68 0.34 .38
Academic 24 (AA) 2.11 3.12 .526 0.93 0.97 1.92 .42 .46
ability 23 (AA) 2.87 0.94 0.94 0.79 .41
(AA) 22 (AA) 3.43 0.92 0.93 0.25 .41
required 21 (AA) 3.59 1.18 1.13 0.67 .33
20 (AA) 3.63 1.56 1.50 0.80 .21
Education 19 (EDU) 2.15 3.12 .555 0.99 1.02 1.86 .07 .37
(EDU) 18 (EDU) 2.88 1.19 1.18 0.76 .28
required 17 (EDU) 3.72 1.34 1.32 1.13 .30
16 (EDU) 3.73 1.49 1.27 1.20 .27
Economic 5 (ER) 2.40 3.31 0.421 1.02 1.03 1.48 .35 .37
rewards 4 (ER) 2.97 0.79 0.80 0.63 .43
(ER) 3 (ER) 3.61 0.70 0.71 0.73 .39
2 (ER) 3.80 1.52 1.44 1.50 .16
1 (ER) 3.82 2.01 1.80 1.64 .13
Freedom of 15 (FA) 3.02 3.48 .447 0.84 0.86 0.55 .57 .39
action 14 (FA) 3.28 0.81 0.83 0.08 .39
(FA) 12 (FA) 3.60 0.98 1.00 0.70 .33
13 (FA) 3.69 0.88 0.88 1.02 .31
11 (FA) 3.84 1.09 1.40 1.77 .08
Total 3.19 .310 1.07 1.06
a
The unit of the values is logit.

Item difficulty distribution and Wright map. The difficulties of each item of the SOAS on a logit scale are
also presented in Table 2. The most difficult item was Item 10, which asked respondents their degree of
agreement with the statement, “I aspire to a job where my decisions affect the whole country or most
people in the country.” To this item, only 10.2% (n ¼ 25) of participants answered 1 (strongly agree),
while 35.9% (n ¼ 88) answered 4 (strongly disagree). In contrast, the easiest item was Item 11,
“I aspire to a job that allows me to take a day off if I am sick or injured.” To this item, 85.3%
(n ¼ 209) of participants answered 4, whereas no one answered 1. We also examined difficulties of
individual content components. Difficulties of components ranged from 0.57 to 0.67. The easiest
component was freedom of action, followed by economic rewards, education, academic ability, and
authority, which is consistent with the order of the means of the content components.
The Wright map is shown in Figure 1. The first column of the map represents the logit scale from
2 to 5. The second and third columns show the vertical location of items and participants in order of
their difficulty and ability levels. Participants’ levels of subjective occupational aspirations ranged
from 0.47 to 6.12. The range of the distribution of items, compared to that of participants, was
Han et al. 13

Logits Items Participants Scoring

+5 Difficult item ▪ (Maximum) 4

▪
+4

+3
▪▪▪▪▪
▪▪▪▪▪▪▪▪
▪▪▪▪▪
▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪
+2 DA10 AA24 ▪▪▪▪▪▪▪▪
EDU19 ▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪
▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪
▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪
ER5 DA9 ▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪
▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪
+1 ▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪
EDU18 AA23 ▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪
ER4 DF15 ▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪ 3
▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪
DA8 DA6 ▪▪▪▪▪▪▪▪
0 DF14 ▪▪▪▪
AA22 ▪
DA7 ▪
AA21 DF12
ER3 AA20 2
-1 DF13
EDU17 EDU16
ER2
ER1
DF11
-2 Easy item 1

Figure 1. Wright map of the Subjective Occupational Aspiration Scale.

located in the slightly lower part of the Wright map. Most participants showed subjective occupational
aspirations that were higher than the medium level (the logit score of 0). Scarcely observed were par-
ticipants with low levels of subjective occupational aspirations.
With regard to the order of difficulty of individual items, almost all items were shown to reflect
correct difficulties; items under each of the five content components properly represented the gradually
increasing levels of the component in an expected order. However, two sets of items turned out to have
reversed difficulties. Under degree of authority, the difficulty of Item 6 (In my future job, I do not want
to be controlled by other people for every little thing I do) was higher than that of Item 7 (I aspire to a
job where I have control of my daily work”). Similarly, under freedom of action, the difficulty of Item
12 (I aspire to a job that allows me to easily take vacations on official leave days or national holidays)
was greater than that of Item 13 (I aspire to a job that allows me to have enough time for my personal,
nonwork life). This indicated that each set of items was perceived by participants differently from what
was expected. Participants found Item 6 to be more difficult than Item 7 and Item 12 more difficult
than Item 13.
14 Journal of Career Development XX(X)

Revising and Finalizing the SOAS

A series of Rasch analyses showed that the overall functioning of the instrument was good, and all 24
individual items were functionally sound. Even so, we attempted to find possible modifications that
would further improve the SOAS by removing Items 1 and 6. Item 1 presented misfits—although the
degree of misfits was not severe—in more than one item-functioning indicator (i.e., in/outfit and biser-
ial correlation). Item 6 was chosen for removal because we speculated that its high difficulty might be
due to being a negatively worded statement, which could affect validity. Thus, three additional Rasch
analyses were performed—that is, model without Item 1, model without Item 6, and model without
both Items 1 and 6. However, none of the three models produced any noticeable difference from the
initial model. To be specific, the person separation reliability value did not change in any of the three
modified models (.81). Also, in none of the three models was there any significant improvement in fit
statistics of the other remaining items. This suggests that the 2 items did not significantly distort the
measurement system of the initial instrument. Therefore, all 24 items were retained. In addition, we
transposed Items 6 and 7 as well as Items 12 and 13 to reflect the correct order of difficulty among
items. The finalized instrument is provided in the Appendix.

Correlations With Selected Variables

None of the demographic or academic variables, including age, grade year, and grade point average
(GPA), were significantly associated with subjective occupational aspirations. However, socioeco-
nomic occupational aspiration was significantly correlated with subjective occupational aspiration.
Specifically, the Pearson correlation value was .30 (p < .001), indicating that the two types of occu-
pational aspirations were correlated to a weak degree. To see the relationship between gender and sub-
jective occupational aspirations, we compared the levels of male and female participants’ subjective
occupational aspirations. Although the average of men’s subjective occupational aspiration (M ¼
3.23, SD ¼ .28) was slightly higher than that of women (M ¼ 3.16, SD ¼ .32), there was no statistically
significant difference between gender, t(243) ¼ 1.658, p ¼ .099.

Discussion
SOAS Findings
This study developed and validated a new psychometric instrument, the SOAS, using a Rasch model-
based approach. The results of Rasch analyses supported the overall reliability and validity of the
SOAS, suggesting that the instrument was appropriate for measuring individuals’ subjective occupa-
tional aspirations. The instrument satisfied the unidimensionality and local independence assumptions
and exhibited a good person separation reliability. Also, all items of the SOAS produced acceptable
outfit values. Most items were distributed in the correct order of difficulty as designed. This implies
that the logic used to develop individual items was reasonable and that items were appropriate for
reflecting incremental levels of respondents’ subjective occupational aspirations.
Our sample’s average score on the SOAS was 3.19, showing a tendency toward negative skewness.
Among the five content components, freedom of action had the highest mean, followed by economic
rewards, education, academic ability, and degree of authority. It appeared that our 4-year college stu-
dent sample was more concerned with what their future jobs can offer to enhance the well-being of
their overall lives rather than how much they can be challenged or exert influence on others through
work. This result could be used as a reference or a standard of comparison when counselors use the
SOAS with college student clients. The results, however, might be different for different populations,
which is an interesting topic for future research. Knowing how the relative importance of the five
Han et al. 15

components differs across people with different backgrounds could help us understand the character-
istics of those people as well as the nature of the subjective occupational aspiration.
With regard to item-level aspects of the SOAS, the items of the SOAS, in general, properly discri-
minated among participants according to their levels of subjective occupational aspirations. The nega-
tively skewed pattern of data, as well as the item-difficulty distribution in the Wright map, suggested
that for our sample, there were many easy items in the SOAS. This tendency makes sense given the
nature of the measurement construct and the demographic characteristics of our sample. First and fore-
most, subjective occupational aspirations reflect future-oriented cognition regarding one’s career with-
out being defined by limiting factors such as perceived abilities, societal norms, and others’
expectations (Eccles, Barber, Stone, & Hunt, 2003; Nurmi, 2004). People tend to evaluate such
future-oriented cognitive traits as high in general or as higher than realistic expectations, which are
formed after being refined by acknowledging limiting factors (e.g., Beal & Crockett, 2010). Second,
high subjective occupational aspirations may partially be attributed to the academic characteristic of
our sample. Our sample represented U.S. college students who possessed relatively high academic
goals. Previous research also reported a positive association between educational achievement and
occupational aspirations (Rojewski, 2005).
Although the overall functioning of the instrument was good, several items appeared to require fur-
ther consideration. First, 3 items (Items 1, 2, and 20) showed outfit values higher than the criterion of
1.4 but, for applied research, fell within an acceptable range (<2.0). Item 1, though, caught our atten-
tion because it showed a low point-biserial correlation (.13). While the degree of misfit that these indi-
cators (infit and point-biserial correlation) signify was not critical, we speculated about the cause.
Possibly, Item 1 (I want my future job to pay enough for me to purchase basic life necessities) was
not distinct enough from other items to distinguish between respondents’ different levels of subjective
occupational aspirations. In order for an item to have proper discriminating function, it must be able to
elicit low answers from respondents with low levels of the measurement construct and high answers
from respondents with higher levels of the construct. In this sense, if an item is so easy that all respon-
dents give the highest answer to it, then the item is unable to function to discriminate different levels of
measurement construct. Indeed, approximately 95% (n ¼ 233) of the sample scored Item 1 with either
3 (mildly agree) or 4 (strongly agree).
Second, Item 11 showed a low point-biserial correlation (.08) but had a proper outfit statistic. Thus,
it was likely that the low point-biserial correlation resulted only from the low difficulty of the item as
explained with regard to Item 1 and would not distort the overall measurement system. Actually, Item
11 was the easiest item in the SOAS with nearly 99% (n ¼ 242) of the sample answering either a 3 or 4
on the item.
Third, two sets of items showed reversed difficulties. Items 6 and 7 measured the degree of author-
ity inherent in a particular occupation. Our analyses showed that participants responded in a way that
Item 6 (In my future job, I do not want to be controlled by other people for every little thing I do) rep-
resented a higher level of aspiration than Item 7 (“I aspire to a job where I have control of my daily
work”), which was inconsistent with the initial item design. We originally constructed the 2 items
based on an assumption that the wording “every little thing” would convey a weaker tone of aspiring
to authority than “daily work.” As we reexamined the 2 items, however, we noticed that Item 6 could
be understood in a different way by participants as a result of its negative wording. Many researchers
have warned of the possibility that negatively worded items may not essentially measure a target con-
struct but measure a different construct (DiStefano & Motl, 2006). For example, for a few participants,
Item 6 may have represented a reluctance toward the idea of “being controlled by other people,” not
aspirations toward the degree of authority present in an occupation. If this discrepancy is real, the item
cannot be aligned with other degree-of-authority items (Items 7–10). This can also account for the fact
that Item 6 did not represent a correct difficulty in relation to the other 4 items.
16 Journal of Career Development XX(X)

Items 12 and 13, which measured the freedom of action, also showed reversed difficulties; Item 12
(I aspire to a job that allows me to easily take vacations on official leave days or national holidays),
which was set to be easier, turned out to be more difficult than Item 13 (I aspire to a job that allows me
to have enough time for my personal, nonwork life). After thoroughly examining the 5 items under the
freedom of action component, we posited that this unexpected result was likely to be a wording issue.
The wording of Item 13—“have enough time for my personal, nonwork life”—could actually be inter-
preted extensively. Originally, we supposed that what Item 13 signified (work–life balance) would be
perceived by respondents as harder to achieve than what Item 12 signified (freedom to take lawfully
designated holidays). However, if respondents interpreted “personal, nonwork life” into something
very basic and fundamental, such as marriage or childbirth, they might have thought that having such
a fundamental personal part of life aside from work would be, or also should be, easier to achieve than
taking holidays from work.
Subsequent Rasch analyses were performed to examine the validity of the revised SOAS that
removed either Item 1 or Item 6, or both, from the initial SOAS. In the IRT scheme, a common
approach to improving an instrument is to remove problematic items. Since the literature offers no par-
ticular standard for item removal, researchers must clearly state the criteria they employed to deter-
mine which items were removed from the instrument (Weinhardt, Morse, Chimeli, & Fisher, 2012).
We considered removing 2 items because we thought it possible they were contaminated. That is,
we suspected that the 2 items, due to their wording, might have measured constructs other than, or
in addition to, subjective occupational aspiration. Results of analyses, however, indicated no notice-
able improvement in any of the revised versions. Given that the misfits shown by both items were
rather negligible and that their functioning had no critical problems, we retained all 24 items in the
SOAS. Thus, the finalized SOAS contains 24 items while the orders of two sets of items (Items 6 and
7 and Items 12 and 13), which showed reversed difficulties, were switched.
Last, we found a significant positive association between subjective and socioeconomic occupa-
tional aspirations. The fact that the two aspirations were associated, but weakly, appears to reflect two
aspects of measurement. First, the two variables share certain aspects in common since some compo-
nents of subjective occupational aspirations (i.e., economic rewards, education required, and academic
ability required) are also reflected by the indicators of the socioeconomic measure of occupational
aspiration (i.e., wage, education required, and prestige score). Participants may have responded to the
SOAS thinking of a specific occupation (possibly the one they specified for the socioeconomic occu-
pational aspiration question) or they may have just expressed their general preferences toward the five
components. Whichever the case, they appeared to be fairly consistent in specifying their occupational
aspirations. Second, despite such consistency, the correlation between subjective and socioeconomic
occupational aspirations was not large, possibly because the two variables were differentiated by other
factors. These factors could include the following: (1) the difference between specifying a single sim-
ple career goal and detailing preferences regarding work and subsequent lifestyles, (2) the other com-
ponents of the SOAS (i.e., degree of authority and freedom of action) that are not directly related to the
SEI indicators, (3) ignorance of the work world, (4) misspecified expressed career goals, or (5) noise
caused by the inaccuracy of the SEI.

Implications for Practice

Knowledge of occupational aspirations offers insight into an individual’s self-concept, sense of self-
efficacy and outcome expectation, perceptions of abilities or limitations, beliefs about the availability
of opportunity, and degree of desire or motivation to achieve stated goals (Rojewski, 2005). Even so,
occupational aspirations are guided, to a large degree, by an individual’s knowledge of and exposure to
the world of work. Knowledge of work is often directly influenced by factors such as gender, race/eth-
nicity, and socioeconomic status. Subjective occupational aspirations stand in contrast to the general
Han et al. 17

notion of occupational aspirations in that they focus on inner longings or preferences represented by
various occupations rather than the exclusive focus on societally defined (often stereotypical) job
titles. Subjective occupational aspirations represent “what individuals want their future jobs to be like”
rather than “what specific jobs they want to have.” This is increasingly important in a world where the
nature of work is changing at an ever-growing rate.
In counseling, the SOAS could be used in two ways. First, the SOAS could serve as an initial
screening tool that detects individuals who necessitate special attention and guidance. Normally,
individuals’ subjective occupational aspirations are expected to be rather high, given the fact that
the construct is a future-oriented cognitive trait (e.g., Beal & Crockett, 2010). Also, individuals’
aspirations toward the five elements should be congruent with the SOAS’ psychometric (i.e., unidi-
mensionality) and conceptual (i.e., the five elements “collectively” express certain occupations)
assumptions. Thus, abnormal responses, such as a significantly low level of aspirations or too much
variation in aspirations between the five elements, can be signs of distorted self-concept or self-
efficacy, inaccurate perceptions of personal abilities and limitations, or wrong beliefs or lack of
knowledge about the world of work.
Second, we view the SOAS as a complement to our understanding and use of occupational aspira-
tions. The scale identifies work-related preferences (aspirations) by reducing the importance of know-
ing that a particular occupation exists. Instead, aspirations for the underlying elements that define
occupations—economic rewards, degree of authority, freedom of action, education required, and aca-
demic ability required—are emphasized. This information could be used in several ways to support
existing career exploration and guidance efforts. For example, SOAS results could be used in more
traditional trait-factor approaches that attempt to match work preferences with work characteristics.
However, when individuals are uncertain about the aspirations or have limited knowledge about the
world of work, SOAS results—such as overall level of the aspiration and relative importance of the
five elements—could be used as a first step to examine the characteristics of personal preferences
toward work. This information could then be used to help individuals explore potential occupations
that reflect these preferences. Finally, since the SOAS work elements are common to most occupa-
tions, this knowledge can be used to expand, rather than restrict, an individual’s range of potential
occupational choices.

Limitations and Future Research

We note several suggestions for future research given the limitations of our study. First, this study was
an initial instrument validation study that used a rather homogeneous sample in terms of educational
level: 4-year university students. Although the chosen sample constitutes an essential portion of the
target population of the SOAS, it is strongly encouraged for future research to make further validation
efforts with other than college-aged subjects in order to establish generalized validity. Good future tar-
gets would be middle and high school students who are facing important career choices (i.e., college
and major). An inclusion of people with diverse levels of education is important also because it would
further reveal the nature of the subjective occupational aspiration. For example, examining whether
our sample’s high level and negatively skewed pattern of subjective occupational aspirations are also
the case for people with lower levels of education (i.e., high school graduates and 2-year college grad-
uates) will illuminate the relationship between educational achievement (or aspiration) and subjective
occupational aspiration. On one hand, the literature on occupational aspirations generally supports a
positive relationship between the two variables (e.g., Rojewski, 2005); thus, people with less education
may show lower subjective occupational aspirations. On the other hand, as discussed in the Introduc-
tion section, individuals’ subjective occupational aspirations are less likely to be influenced by their
surrounding environment or backgrounds than occupational aspiration; therefore, most people might
just possess high hopes for their future career regardless of their educational condition. All in all, the
18 Journal of Career Development XX(X)

relationship between education and subjective occupational aspirations is currently unknown and
requires investigation to gain a deeper understanding of how subjective occupational aspirations
develop in individuals.
Second, as with the first suggestion, examining other demographic characteristics such as race, eth-
nicity, socioeconomic status, and so on in relation to subjective occupational aspirations would also be
important in further understanding the new construct. Comparing what has been known about the rela-
tionships between socioeconomic occupational aspiration and demographic characteristics with the
subjective occupational aspiration will help illuminate distinct attributes of the subjective occupational
aspiration.
Third, this study, as an initial development and validation attempt, showed that the SOAS is an
internally sound psychometric instrument that can properly measure and calibrate individuals’ levels
of subjective occupational aspirations. Thus, we regard examining external aspects of the SOAS as a
next good step to further establish the SOAS’s validity. Particularly, future research can strengthen
criterion and construct validity of the SOAS by examining the association between subjective occupa-
tional aspirations and other relevant constructs. For example, the SOAS and CAS (Gray & O’Brien,
2007) can be comparable in that they both concern one’s tendency to achieve in a career but can also
be different because the CAS focuses on one’s achievement regardless of professions while the SOAS
emphasizes professions per se. Thus, a relationship such as a mild correlation may support the two
measure’s conceptual association. In addition, existing literature on occupational aspirations provides
a good starting point for such attempts, suggesting possible covariates of the new construct such as
self-concept, perceived barriers and opportunities, career self-efficacy and expectations, motivations,
and so on (e.g., Rojewski, 2005). If future research finds any meaningful relationships between the
subjective occupational aspiration and the suggested variables or others, it will help us to further
understand subjective occupational aspirations’ role in individuals’ career development, thereby
strengthening the validity of the SOAS.
Fourth, in future validation efforts, researchers ought to reexamine several items that showed a
slight possibility of misfit (Items 1, 2, and 20). Although our results show that those items caused
no major problems in the SOAS’s measurement system, we recommend that researchers ensure that
this also holds for other samples or explore possible modifications.
Fifth, we found two sets of items that indicated reversed difficulty levels (Items 6 and 7 and Items
12 and 13) and provided our interpretation and explanation of the phenomenon. Two things must be
noted regarding this. One is that future studies need to reexamine these sets of items to ascertain that
the revised difficulty order is supported by different samples. The other is, if the revised difficulty is
supported, users of the SOAS need to fully understand the difficulty structure of the instrument with
the guidance of this study in applying the SOAS.

Appendix
Subjective Occupational Aspiration Scale
Directions. One’s work is characterized by many things including money, authority, freedom, educa-
tion, and intellectual ability. The items below represent these issues and ask you to think about how
important they are to you. Unlike the previous page, as you respond to each of these issues, you do
not have to have a particular occupation in mind (e.g., lawyer), instead focus on each issue and its
importance to you. Even if you do not know what type of work you want to do, you may still want
a job that pays extremely well but requires minimal education.
Each of the five issues are measured using 4 or 5 items, which represent gradually increasing levels
of aspiration. Your response to each item can range from 1 to 4 (1 ¼ strongly disagree and 4 ¼ strongly
agree). For each issue, please circle a number that most closely represents your thoughts.
Han et al. 19

Strongly Mildly Mildly Strongly

Disagree Disagree Agree Agree

Economic rewards
1. I want my future job to pay enough for me to purchase basic life 1 2 3 4
necessities
2. I want my future job to pay enough for me to enjoy at least an average 1 2 3 4
life style
3. I aspire to a job that pays fairly well, allowing me to purchase high- 1 2 3 4
quality goods and services and live a comfortable life
4. I aspire to a job that pays very well, allowing me to purchase highly 1 2 3 4
luxurious houses, cars, and goods and live an immensely
comfortable life
5. I aspire to a job that pays extremely well and provides the greatest 1 2 3 4
comforts and luxury in life but only few people can obtain
Degree of authority
6. I aspire to a job where I have control of my daily work 1 2 3 4
7. In my future job, I do not want to be controlled by other people for 1 2 3 4
every little thing I do
8. I aspire to a job where I can make decisions that influence the future 1 2 3 4
of my organization and can manage many people in the
organization
9. I aspire to a job where I have an influence on a state-size area or a 1 2 3 4
nationwide large group of people
10. I aspire to a job where my decisions affect the whole country or most 1 2 3 4
people in the country
Degree of freedom
11. I aspire to a job that allows me to take a day off if I am sick or injured 1 2 3 4
12. I aspire to a job that allows me to have enough time for my personal, 1 2 3 4
nonwork life
13. I aspire to a job that allows me to easily take vacations on official 1 2 3 4
leave days or national holidays
14. In my future job, I want to be able to adjust my working hours flexibly 1 2 3 4
depending on my personal schedule
15. I aspire to a job where I can freely choose what to perform as daily 1 2 3 4
tasks
Education required
16. I want my future job to require education equal or higher than 1 2 3 4
compulsory education (K–12)
17. I aspire to a job that requires at least a 4-year college education 1 2 3 4
18. I aspire to a job that requires either graduate degrees or professional 1 2 3 4
certificates such as lawyer or doctor
19. I aspire to a job that requires an elite education that only few select 1 2 3 4
people possess
Academic ability required
20. I want my future job to involve tasks that require at least basic 1 2 3 4
literacy skills
21. I want to have a job that involves at least a medium level of academic 1 2 3 4
difficulty, such as basic reading and math skills
22. I aspire to a job that involves fairly complex tasks that only a person 1 2 3 4
of average intelligence or higher can complete
23. I aspire to a job that involves very complex tasks that can be 1 2 3 4
performed by only a very intelligent person
24. I aspire to a job that involves extremely complex tasks that can be 1 2 3 4
performed by only a few people (less than 5% of people)
20 Journal of Career Development XX(X)

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or
publication of this article.

Funding
The author(s) received no financial support for the research, authorship, and/or publication of this
article.

Note
1. In item response theory, difficulty of an item can be interpreted as a degree of the item’s threshold. To give a
right answer to an item, a test-taker’s ability must be higher than the item’s threshold (thresholds and abilities
are manifested in a logit scale). In a test setting or a binary scoring scheme, a difficult item indicates that the
item has a high threshold; thus, it is relatively difficult for respondents (or their abilities) to surpass the item’s
threshold and provide a right answer or a “yes” to the item. In this sense, in a polytomous scheme, a difficult
item indicates that respondents would rarely give the item a high score. On the other hand, an easy item indi-
cates a low threshold, making it more likely that respondents (or their abilities) exceed that threshold and pro-
vide a correct answer or higher scores.

References
Beal, S. J., & Crockett, L. J. (2010). Adolescents’ occupational and educational aspirations and expectations:
Links to high school activities and adult educational attainment. Developmental Psychology, 46, 258–265.
doi:10.1037/a0017416
Blau, P. M., & Duncan, O. D. (1967). The American occupational structure. New York, NY: Wiley.
Crawley, D. (2014). Gender and perceptions of occupational prestige. SAGE Open, 4, 1–11. doi:10.1177/
2158244013518923
Crawley, D., & Cardinale, T. (2008, May). Perceptions of occupational status: Survey development. Poster pre-
sented at the meeting of the association for psychological science, Chicago, IL.
Cronbach, L. J. (1951). Coefficient alpha and the internal structure of tests. Psychometrika, 16, 297–334. doi:10.
1007/BF02310555
DiStefano, C., & Motl, R. W. (2006). Further investigating method effects associated with negatively worded
items on self-report surveys. Structural Equation Modeling, 13, 440–464. doi:10.1207/s15328007sem1303_6
Eccles, J. S., Barber, B. L., Stone, M., & Hunt, J. (2003). Extracurricular activities and adolescent development.
Journal of Social Issues, 59, 865–889. doi:10.1046/j.0022-4537.2003.00095.x
Embretson, S. E., & Reise, S. P. (2000). Item response theory for psychologists. Mahwah, NJ: Lawrence Erlbaum.
Engelhard, G. (2013). Invariant measurement: Using Rasch models in the social, behavioral, and health sciences.
New York, NY: Routledge.
Furlong, A., & Cartmel, F. (1995). Aspirations and opportunity structures: 13-year-olds in areas with restricted
opportunities. British Journal of Guidance and Counselling, 23, 361–375. doi:10.1080/03069889508253694
Ganzeboom, H. B. G. (2010, May). A new international socio-economic index (ISEI) of occupational status for
the international standard classification of occupations 2008 (ISCO-08) constructed with data from the ISSP
2002–2007. Paper presented at the annual meeting of the international social survey programme. Lisbon,
Portugal.
Ganzeboom, H. B. G., De Graaf, P. M., & Treiman, D. J. (1992). A standard international socio-economic index of
occupational status. Social Science Research, 21, 1–56. doi:10.1016/0049-089X(92)90017-B
Gottfredson, G. D. (1996). Prestige in vocational interests. Journal of Vocational Behavior, 48, 68–72. doi:10.
1006/jvbe.1996.0006
Gray, M. P., & O’Brien, K. M. (2007). Advancing the assessment of women’s career choices: The Career Aspira-
tion Scale. Journal of Career Assessment, 15, 317–337. doi:10.1177/1069072707301211
Han et al. 21

Hambleton, R. K., Jaeger, R. M., Plake, B. S., & Mills, C. (2000). Setting performance standards on complex edu-
cational assessments. Applied Psychological Measurement, 24, 355–366. doi:10.1177/01466210022031804
Hauser, R. M., & Warren, J. R. (1996). A socioeconomic index for occupations in the 1990 Census. (Working
Paper No. 96-01). Madison, WI: University of Wisconsin–Madison, Center for Demography and Ecology.
Retrieved from https://www.ssc.wisc.edu/cde/cdewp/96-01.pdf
Holland, J. L., & Gottfredson, G. D. (1975). Predictive value and psychological meaning of vocational aspirations.
Journal of Vocational Behavior, 6, 349–336. doi:10.1016/0001-8791(75)90007-x
Hotchkiss, L., & Borow, H. (1996). Sociological perspectives on work and career development. In D. Brown, L.
Brooks, & Associates. (Eds.), Career choice and development (2nd ed., pp. 262–307). San Francisco, CA:
Jossey-Bass.
IBM Corporation. (2012). IBM SPSS Statistics for Windows (ver. 21.0). Armonk, NY: Author.
Johnson, M. K., & Mortimer, J. T. (2002). Career choice and development from a sociological perspective. In D.
Brown & Associates (Eds.), Career choice and development (4th ed., pp. 37–81). San Francisco, CA: Jossey-
Bass.
Lent, R. W. (2005). A social cognitive view of career development and counseling. In S. D. Brown & R. W. Lent
(Eds.), Career development and counseling: Putting theory and research to work (pp. 101–127). New York,
NY: Wiley.
Lent, R. W., Brown, S. D., & Hackett, G. (1994). Toward a unifying cognitive theory of career and academic inter-
est, choice, and performance [Monograph]. Journal of Vocational Behavior, 45, 79–122. doi:10.1006/jvbe.
1994.1027
Lent, R. W., Hackett, G., & Brown, S. D. (1996). A social cognitive framework for studying career choice and
transition to work. Journal of Vocational Education Research, 2, 3–13.
Linacre, J. M. (2004). Rasch model estimation: Further topics. Journal of Applied Measurement, 5, 95–110.
Linacre, J. M. (2014). Reliability and separation of measures: A user’s guide to win steps ministep Rasch-model
computer programs (ver. 3.81.0). Retrieved from http://www.winsteps.com/winman/reliability.htm
Linacre, J. M., & Wright, B. D. (1988). Facets. Chicago, IL: MESA.
Miguel, J. P., Silva, J. T., & Prieto, G. (2013). Career decision self-efficacy scale-short form: A Rasch analysis of
the Portuguese version. Journal of Vocational Behavior, 82, 116–123. doi:10.1016/j.jvb.2012.12.001
Nakao, K., & Treas, J. (1992). The 1989 socioeconomic index of occupations: Construction from the 1989 occu-
pational prestige scores (GSS Methodological Report, No. 74). Chicago, IL: National Opinion Research
Center.
Nakao, K., & Treas, J. (1994). Updating occupational prestige and socioeconomic scores: How the new measures
measure up. Sociological Methodology, 24, 1–72. doi:10.2307/270978
Nurmi, J. E. (2004). Socialization and self-development: Channeling, selection, adjustment, and reflection. In R.
Lerner & L. Steinberg (Eds.). Handbook of adolescent psychology. (2nd ed., pp. 85–124). Hoboken, NJ: Wiley.
Rojewski, J. W. (2005). Occupational aspirations: Constructs, meanings, and application. In S.D. Brown & R. W.
Lent (Eds.), Career development and counseling: Putting theory and research to work (pp. 131–155). New
York, NY: Wiley.
Rounds, J. B. Jr., Henly, G. A., Dawis, R. V., Lofquist, L. H., & Weiss, D. J. (1981). Manual for the Minnesota
Importance Questionnaire: A measure of vocational needs and values. Minneapolis: University of Minnesota.
Schafer, J. L. (1999). Multiple imputation: A primer. Statistical methods in medical research, 8, 3–15. doi:10.
1177/096228029900800102
Schoon, I., & Parsons, S. (2002). Teenage aspirations for future careers and occupational outcomes. Journal of
Vocational Behavior, 60, 262–288. doi:10.1006/jvbe.2001.1867
Smith, R. M. (1996). A comparison of methods for determining dimensionality in Rasch measurement. Structural
Equation Modeling: A Multidisciplinary Journal, 3, 25–40. doi:10.1080/10705519609540027
Stevens, G., & Cho, J. H. (1985). Socioeconomic indexes and the new 1980 census occupational classification
scheme. Social Science Research, 14, 142–168. doi:10.1016/0049-089X(85)90008-0
22 Journal of Career Development XX(X)

Super, D. E. (1990). A life-span, life-space approach to career development. In D. Brown & L. Brooks (Eds.),
Career choice and development: Applying contemporary theories to practice (2nd ed., pp. 197–261). San Fran-
cisco, CA: Jossey-Bass.
Super, D. E., Savickas, M. L., & Super, C. M. (1996). The life-span, life-space approach to careers. In D. Brown &
L. Brooks (Eds.), Career choice and development (3rd ed., pp. 121–178). San Francisco, CA: Jossey-Bass.
U.S. Bureau of the Census. (1992). Classified index of industries and occupations: 1990 Census of population and
housing (1990 CPH-R-4). Washington, DC: U.S. Government Printing Office.
Varma, S. (2006). Preliminary item statistics using point biserial correlation and p value. Educational data sys-
tem. Retrieved from http://www.eddata.com/resources/publications/eds_point_biserial.pdf
Walker, E. R., Engelhard, G., & Thompson, N. J. (2012). Using Rasch measurement theory to assess three depres-
sion scales among adults with epilepsy. Seizure, 21, 437–443. doi:10.1016/j.seizure.2012.04.009
Weinhardt, J. M., Morse, B. J., Chimeli, J., & Fisher, J. (2012). An item response theory and factor analytic exam-
ination of two prominent maximizing tendency scales. Judgment and Decision Making, 7, 644–658.
Wright, B. D., Linacre, J. M., Gustafson, J. E., & Martin-Löf, P. (1994). Reasonable mean-square fit values. Rasch
Measurement Transactions, 8, 370.

Author Biographies
Hyojung Han is an associate research fellow at the Korean Educational Development Institute, a government-
funded research institute that conducts research on the educational system and policies in South Korea. Her
research interests center on individuals’ career development and lifelong learning; specifically, her most recent
works have explored career patterns of women, lifelong learning for women, occupational aspirations, and job
satisfaction. She also has a special interest in development and validation of psychometric instruments in career
research and practice. Personal interests include playing the piano and watching TV shows.

Jay W. Rojewski is a professor in the Department of Career and Information Studies at the University of Georgia.
His research emphasizes various aspects of adolescent career development and the transition from school to post-
secondary education and work. His most recent works have investigated the development of career aspirations,
connections between adolescents’ aspirations and postsecondary educational attainment, and the career outcomes
of individuals with high-incidence disabilities. Personal interests include spending time with his family, riding his
Harley-Davidson Electra Glide Classic, and building and playing electric guitars.

Minho Kwak is a doctoral student in the program of quantitative methodology at the University of Georgia. He
graduated with bachelor and master’s degrees in vocational education and workforce development. His research
focuses on analyzing constructed response items using text mining and items response theory (IRT). He is also
interested in instrument development based on an IRT approach. He spends his spare time playing the piano.
He also has a strong interest in art and likes to draw and appreciate paintings.

View publication stats