Peer-Reviewed Article

© Journal of Underrepresented and Minority Progress

Volume 7, Issue 1 (2023), pp. 1-26
ISSN: 2574-3465 Print/ ISSN: 2574-3481 Online
http://ojed.org/jump

Differences in Career Outcome Expectations of College

Students by Race/Ethnicity and Gender
Jacqueline Doyle, Elyse Postlewaite, Philip M. Sadler, and
Gerhard Sonnert
Harvard University, USA

ABSTRACT
This study examined how U.S. college students’ career outcome
expectations—what they hope to get out of their careers—vary by intended
career path, racial/ethnic groups, gender, and other individual difference
factors. The data were drawn from the Persistence Research in Science and
Engineering (PRiSE) survey, a national study of U.S. college students
enrolled in college English courses (n = 7505). An exploratory factor analysis
revealed four foci of career outcome expectations, which we labeled as
follows: extrinsic (rewards are external, such as money or status), work-life
balance (work does not consume all of a person’s time/energy), pioneering
(work is intellectually stimulating and cutting edge), and people-related
(work involves working with and helping others). While controlling for career
interest, our findings indicate that students’ gender and race/ethnicity
influence their career outcome expectations in a wide variety of ways. Due to
the differences in career outcome expectations associated with student
backgrounds and demographics beyond career interest, recruiters and
program directors looking to attract more diverse populations may benefit
from matching the career outcomes they present and offer with those
populations’ outcome expectations.

Keywords: career interest, gender, race/ethnicity, socialization,

STEM careers

-1-
 INTRODUCTION
Work is a central aspect of an individual’s life (Csikszentmihalyi &
Schneider, 2000). It contributes to one’s positive self-concept, self-
satisfaction, happiness, identity, and fulfillment, but also to misery,
unhappiness, and boredom (Fogg, 2012; Gagné & Bhave, 2011;
Csikszentmihalyi & LeFevre, 1989; Csikszentmihalyi & Schneider, 2000).
For most of human history, work has been fairly immutable and divided along
gender and class lines. By contrast, young adults in the U.S. today are faced
with almost infinite choices in their career paths. The most recent generation
to enter the workforce, termed “millennials,” faces a unique labor milieu,
where they have been raised to believe they can pursue any career they desire.
As the freedom to choose an occupation expands, one’s job becomes less
about inevitability and more about individual preferences, desires, and the
fulfillment of individual potential. And while “men and women say they
would keep working even if they did not have to,” the motivations and values
behind why we work have changed over time from an existential necessity to
a calling, to social status, to material gain, to personal fulfillment
(Csikszentmihalyi & Schneider, 2000, p. 33). Studying which careers youth
end up selecting, and for what reasons, has become critically important.
The study of career development began in the early 1900s and
continues to be an important field today (McMahon, 2014a). Researchers
have long understood the importance of examining the choice and
development of careers. In today’s diverse work force, implicit and explicit
racism and the gender glass ceiling still persist. Career development
researchers have begun to shift attention to disadvantaged and
underrepresented groups (Hazari et al., 2013b; Riegle-Crumb et al., 2011;
Reyes, Kobus, & Gillock, 1999; Sandberg, Ehrhardt, Mellins, Ince, & Meyer-
Bahlburg, 1987), and this focus is bound to gain in importance. Some work
that particularly targets underrepresented groups in the labor force has
hypothesized that the observed gender and race/ethnicity gaps may not result
from a lack of opportunities, but a lack of desire (Reyes et al., 1999; Sandberg
et al., 1987). This finding highlights the importance of understanding the
factors that contribute to an individual’s career expectations, and whether they
differ by gender or race/ethnicity.
Today, young people are forming certain expectations of what
benefits they will get out of their careers (e.g., money, fame, time for family,
flexible hours, etc.) and how those outcomes will affect their personal
fulfillment. Career outcome expectations (COEs) are what students desire
from their future occupations (Fouad & Smith, 1996; Lent, Brown, & Hackett,
1994; Wigfield & Eccles, 1992). We believe that people may vary in what
-2-
they want to get out of their careers. This may be different for people who
select different careers, or are of different gender, or different racial and ethnic
background. For example, COEs may be connected with stereotypes about
certain careers (e.g., a health career is regarded as being people related, or a
massage therapist is believed to have a more flexible work-life balance).
Empirically, COEs with “people” or “thing” (object) orientations are
associated with various college majors. Yang and Barth (2015) found that
chemistry, engineering, mathematics, and physical science majors had the
lowest ratings for people-oriented careers; jobs affording family and social
impact goals mapped to people-orientation; and jobs affording status goals
mapped to thing-orientation. Extant research on the importance of finding a
work-life balance has mixed results as to whether work-life balance is
universally appealing or only desirable to those who can fully take advantage
of it (Casper & Buffardi, 2004). For example, one study found specific work
benefits, like telecommuting and flexible work time, to have varying
importance depending on the individual (Rau & Hyland, 2002), while another
study found flexible career paths to be universally attractive (Honeycutt &
Rosen, 1997). Additionally, COEs have been empirically associated with
gender (Hazari, Sonnert, Sadler, & Shanahan, 2010), but how COEs vary by
race/ethnicity has been less studied.
With their increased job possibilities, millennials may be guided more
by COEs in their career choice than previous generations. Therefore, it is
important to better understand COEs as they become a stronger motivating
force for career selection. While this study is limited in making causal claims,
it examines the association of individual factors, such as race/ethnicity,
gender, and career interests, with COEs. The goal of this study was twofold:
To identify the COEs college students have.
To determine to what extent gender, race/ethnicity, and career interest
predict college students’ COEs.
We hypothesize that all three factors are associated with students’
COEs, while focusing on the effects of race/ethnicity and gender. This study
is one of a kind in its large-scale investigation of this hypothesis.

Theoretical Framework
Two main themes emerge in the research about COEs and,
consequently, career choice: career theories that emphasize dispositional
traits to explain individual differences and career theories that emphasize the
contextual and environmental, alternatively societal factors, that influence
individual differences. There are two major dispositional models:

-3-
 In Holland’s (1997) RIASEC career interest model, individuals orient
themselves in varying degrees toward realistic, investigative, artistic, social,
enterprising, or conventional (RIASEC) goals for COEs.
People-thing orientation (PTO) looks at the analysis of gender
differences through either people-orientations or thing-orientations (Yang &
Barth, 2015).
On the other hand, there are also two influential contextual and
environmental orientations: Social cognitive career theory (SCCT) accounts
for a large portion of the literature about career development and career
choice. SCCT was developed in reference to, and shares a similar structure
with, Bandura’s Social Cognitive Theory, which suggests “people act on their
judgments of what they can do, as well as on their beliefs about the likely
effects of various actions” (Bandura, 1986, p. 231). SCCT posits self-efficacy,
outcome expectations, interests and goals, mediated by environmental and
contextual influences as the main predictors of career choice (Lent et al.,
1994).
Role conformity theory focuses on contextual and environmental
factors (Brown, Darden, Shelton, & Dipoto, 1999; Conkel Ziebell, 2010; Su
et al., 2009) and factors that are largely out of an individual’s control (Arbona,
1990; Conkel Ziebell, 2010; Constantine et al., 1998; Hanson, 1994).
According to this theory, for instance, differences between men and women’s
career choices result from the socialization of gender norms (Yang & Barth,
2015).
The value of considering individuals and their environment (e.g.,
society, time) in conjunction is now widely recognized (Betz, Fitzgerald, &
Hill, 1989; Conkel Ziebell, 2010; McMahon et al., 2014b), because individual
(dispositional) and contextual factors often interact. For example, at the
individual level, Blustein’s (2011) relational theory of working builds on the
constructivist approach to career theory (McMahon, 2014) and puts a greater
emphasis on personal agency, meaning making between the individual and
their broader context, and narrative discourse. However, it is also recognized
that many people have no “choice” in their career development; for these
individuals, environmental and societal barriers play a larger role than does
personal agency (Brown et al., 1999; Conkel Ziebell, 2010). This difference
may be particularly salient for minority groups. Unfortunately, there is a
limited amount of research on different racial/ethnic groups’ career
development, a lacuna that our study was intended to address.
Because societally reinforced gender norms in occupations change
slowly, a strong gender norm still suggests that women should be interested
primarily in helping-type occupations and men should be primarily interested
-4-
in mechanical-type occupations (Arnold, 1993; Riegle-Crumb, Moore, &
Ramos-Wada, 2011; Reyes, Kobus, & Gillock, 1999; Su et al., 2009). For
example, Yang and Barth (2005) found that on the occupational thing-
orientation (a desire to work with objects), men scored higher than women.
Conversely, on the occupational people-orientation (a desire to work with
people), women scored higher than men. It is important to continue to assess
differences between women’s and men’s interests and attitudes as they select
their vocations. Our study contributes to the body of evidence in this area.
Finally, time plays a significant role in career theory. It dictates the
emergence of career interests and accounts for the fact that people are
interested in different careers at different times in their lives (Arthur, Hall, &
Lawrence, 1989). At the individual level, attitudes about careers develop
rapidly, mainly in the first couple of decades of life, where interventions may
be implemented (Su et al., 2009). A critical time in career interest
development is the late-adolescent to early-adult period (Su et al., 2009; Lent
et al., 1994). In particular, college age is when students have the greatest
opportunity to convert career interests to actual career choices, and career
interests at that stage become good predictors of career choice (Sadler et al.,
2012; Riegle-Crumb et al., 2010). Hence, our study focuses on college
students and their career interests.

LITERATURE REVIEW
The voluminous research on careers has mainly focused on self-
efficacy (Adachi, 2004; Lent et al., 1994), identity (Hazari et al., 2010), role
models (Hazari et al., 2013a), exposure (Hazari et al., 2010), interest (Hazari
et al., 2013a; Hazari, Sadler, & Sonnert, 2013b; Sadler, Sonnert, Hazari, &
Tai, 2012), goals (Lent et al., 1994), career choice (Adachi, 2004; Hazari et
al., 2010; Lent et al., 1994), much under the general umbrella of social
cognitive career theory (SCCT) (Lent et al., 1994). There has been some
attention to COEs (e.g., Adachi, 2004; Csikszentmihalyi & Schneider, 2000;
Yang & Barth, 2015), along with gender (Arnold, 1993; Sadler et al., 2012;
Su, Rounds, & Armstrong, 2009) and race (Hazari et al., 2013b; Riegle-
Crumb et al., 2011; Reyes et al., 1999), but it seems fair to say that they have
received comparatively little quantitative empirical consideration (Domene,
Socholotiuk, & Woitowicz, 2011). Notably, most of the research on gender,
race/ethnicity, and careers has been about science, technology, engineering,
and mathematics (STEM) careers (see Hazari et al., 2010; Sadler et al., 2012;
Riegle-Crumb et al. 2011). In a meta-analysis of gender differences in career
development, Su et al. (2009) recognized the importance of future research
focusing on variability between racial and ethnic groups.
-5-
 In a majority of the studies that address COEs, for example, in studies
using social cognitive career theory (Lent et al., 1994), the COEs have been
predictors of career choice. Not surprisingly, it was found that students who
have certain COEs were more likely to persist in a career that they believed
would fulfill those expectations (Adachi, 2004; Csikszentmihalyi &
Schneider, 2000; Hazari et al., 2010; Lent et al., 1994). For instance, Casper
and Buffardi (2004) found that COEs related to schedule flexibility,
dependent care assistance, and salary predicted career choice.
The reverse pathway—that is, how career interests predict students’
COEs, which is the subject of this paper—has been less studied. This pathway
allowed us to place career interests at the same level as race/ethnicity and
gender and to conjointly assess the comparative strength of impact of those
three predictors on COEs (while controlling for several background
variables).
We are aware of no studies of careers that have specifically looked at
gender, race/ethnicity, and career interests as predictors for college students’
COEs. However, a few noteworthy studies have examined how students’
COEs predicted their occupations. Yang and Barth (2015) investigated
students’ COEs predicting interest in different occupations using both PTO
and role congruity theory to examine gender differences in CEMP (computer
science, engineering, mathematics, physical sciences) majors of college
students (n = 1848). They found no significant differences by major in thing
orientation (a preference to work with objects rather than with people).
However, they did find a significant gender interaction: men in CEMP majors
had less interest in working with people than did men in biology or health
majors. Interestingly, women interested in CEMP majors had a similar level
of people-orientation (preference to work with people rather than objects) to
women in the health fields. Yang and Barth similarly found that role congruity
theory significantly predicted interest in people-related jobs and thing-related
jobs, but that gender only explained less than 1% of the variance in interest.
Our study extends their research by using a large national sample that looks
at racial/ethnic and gender differences in COEs in addition to differences in
COEs by career interest.
Another study examined students’ physics identity (whether someone
sees themselves as a “physics person”) in relation to their COEs, while
controlling for gender and previous physics experiences. With a robust
sample size of 3,829 students from 34 randomly selected colleges, Hazari and
colleagues (2010) found that college students’ physics identity not only
predicted their career interests, but also correlated positively with a desire for
an intrinsically fulfilling career and negatively with a desire for
-6-
personal/family time and working with people. However, the study was
limited to only students interested in a career in physics. Our study expands
this scope to include health, medicine, and non-science careers in addition to
STEM careers.
In a representative cross-sectional study of 3,602 middle and high
school students’ expectations for the future, Csikszentmihalyi and Schneider
(2000) found that male and female students had similar expectations for their
academic pathways, occupations, and lifestyles. However, female students
expected “their future jobs to be more enjoyable” (p. 77) than did male
students. Interestingly, in contrast to the current reality of gender-based pay
gaps, female students’ salary expectations were similar to male students.
There were no significant differences between racial groups and their
expectations for well-paying and enjoyable jobs.
In the same study, students were questioned about the occupation
they would like to have and the occupation they expected to have. Students
became more pragmatic and realistic with age, which supported our rationale
of examining students’ career interests and COEs at college age.
Csikszentmihalyi and Schneider (2000) also found clear gender differences
aligned with gender career stereotypes (e.g., boys tended to prefer
occupations such as athlete, engineer, or police officer, while girls tended to
prefer occupations such as nurse, teacher, or secretary). They also found clear
racial and ethnic group differences. For example, Black students more
frequently mentioned athlete and lawyer as occupations, Hispanic students
mentioned police officer and nurse more frequently, and Asian students
mentioned architect, businessperson, doctor, and engineer more frequently
than did other student groups.
As a follow-up, students were asked to indicate how important certain
COEs were to the job they expected to have in the future. Csikszentmihalyi
and Schneider (2000) found students to have distinct COEs that correlated
with the occupation they expected to have. This suggests that students may
already have some pre-conceived associations of COEs with specific
occupations, which also seem to follow occupational stereotypes.
Furthermore, this finding appears to indicate that students understand that
different jobs afford different opportunities (e.g., future teachers do not expect
to be famous or make a lot of money; future doctors do not expect to have a
lot of free time). While this finding supports our hypothesis, it has yet to be
studied with college students.
Within the same study, Csikszentmihalyi and Schneider (2000)
investigated students’ motivations (intrinsic, extrinsic, social) for their work,
which we define as part of students’ COEs. “Intrinsic” motivation for work
-7-
included enjoyment, interest, learning something new, taking on a challenge,
and inherent talent. “Extrinsic” motivation for work included job security,
making money, meeting parental expectations, not falling behind, and
learning something useful. A third “social” motivation included impressing
friends, doing better than others, and getting respect. These motivations are
not necessarily mutually exclusive within a specific occupation. The
researchers found 52% of students rated intrinsic motivations the highest in
importance to them, 40% rated extrinsic motivation the highest, and only 8%
rated social motivations the highest. Surprisingly, female students were more
likely than male students to be motivated by extrinsic rewards. Also,
Caucasian students were more likely than other racial/ethnic groups to be
motivated by intrinsic factors (as opposed to extrinsic factors). Again, these
findings support our hypothesis regarding differences in COEs by gender and
race; however, these differences have yet to be studied at the college age.

RESEARCH METHOD
Data
The data used in this article were collected as part of the Persistence
Research in Science and Engineering (PRiSE) project, a study of students in
college English classes that was funded by the National Science Foundation.
The project included a 50-question survey developed to examine in-school
and out-of-school factors, as well as demographics, which may predict college
students’ persistence in science- and engineering-related careers as well as
non-science careers. In the fall semester of 2007, responses were obtained
from a large national sample of college students (n = 7505) from 40 two- and
four-year U.S. colleges and universities selected from a stratified random
sample that accounted for institution size and type. Additionally, six of the
schools were oversampled to ensure adequate representation of students from
underrepresented populations (one historically Black college, one Hispanic-
serving college, and four women’s colleges). Whereas this dataset was
collected a while ago, it remains relevant owing to its unique large size and
national scope. Furthermore, the pandemic may have drastically redefined
students’ career outcome expectations. This urgently calls for a repetition of
this research, for which the results of the present study would constitute a
valuable baseline measurement. Students were surveyed in a mandatory
college English course in order to generate a sample that included both
students who were interested in STEM careers and those who were
uninterested. Of the participants, 53% were female. In terms of race/ethnicity,
14% of the respondents were Hispanic, 62% non-Hispanic White, 7% non-

-8-
Hispanic Black, 5% non-Hispanic Asian, and 6% non-Hispanic Other, with
the remainder providing no information.1
The PRiSE survey contained 50 items that were validated through
focus group discussions with college students and experts in science
education. Moreover, to obtain good content validity, the development of the
questionnaire had been guided by open-ended responses from 412 science
teachers and scientists to a preliminary survey that had the purpose of
identifying and incorporating a breadth of hypotheses and views. The survey
was pilot tested with 49 students to ensure items, vocabulary, and scaling
could be adjusted to reflect the natural variation in experiences. Test-retest
reliability of the survey was established by administering the survey to 96
students twice, in an interval of about two to three weeks. Correlation
coefficients and Cohen’s kappas indicated an overall mean test-retest
reliability of .67.

Measures
This study focuses on college students’ COEs in relation to their
career interest, gender, and race/ethnicity, while controlling for other
influences (e.g., socio-economic status).

Dependent variables
The dependent variables in this study are the COEs of college
students. These variables were created considering students’ answers to 15
items regarding their specific COEs. The 15 items include most of the COEs
on the list of important career values in The Career Decision-Making System
Revised (excluding only “physical activity,” “outdoor work,” “risk,”
“variety,” and “work with hands”) (Fogg, 2012). Our COE variables were
created through an exploratory factor analysis by identifying clusters of
original items that could be grouped together. Participants rated the
importance of items relating to their future career satisfaction including
making money, becoming famous, helping other people, having a leadership
role, having job security, working with people (rather than objects), inventing
new things, developing new knowledge, having time for family, having time
for myself, making my own decisions, having an easy job, having an exciting
job, making use of my talents, and having lots of job opportunities. The
question read “Rate the following factors in terms of their importance for your
future career satisfaction.” Participants rated each item on a six-point scale
from 1 = “not at all important” to 6 = “very important.” A factor analysis
using a Varimax rotation grouped most of the items into four factors with two
items standing alone (see Table 1).
-9-
Table 1
Factor Analysis of Career Outcome Expectations

Finally, the items constituting a factor were standardized and added

and the resulting composite was standardized again to facilitate interpretation.
We named the composites extrinsic outcomes, pioneering outcomes, work-
life balance outcomes, and people-related outcomes, and excluded two items
which did not strongly load onto any of the four factors (“having an easy job”
and “having an exciting job”).

Independent variables of interest

Gender was coded as a dummy variable (female = 0, male = 1).
Race/ethnicity was coded as one categorical variable (or, equivalently, as
separate dummy variables). On the survey, students could identify their
ethnicity as Hispanic or non-Hispanic. Students could also identify their racial
identity as White, Black, Asian, Native American, Pacific Islander, Other, or

- 10 -
more than one race (indicated by multiple selections). For the purpose of our
analyses, all students who indicated they were Hispanic were categorized as
Hispanic, regardless of their further racial identification.

Table 2
Career Interest Fields Presented to Students and Their Composited
Groupings
Field Grouping
Biologist
Earth/Environmental scientist
Astronomer
Chemist
Physicist
STEM
Engineer
Computer scientist
Mathematician
Science teacher
Math teacher
Medical professional Medicine
Health professional Health
Other teacher
Social scientist (e.g., psychologist, sociologist)
Businessperson
Non-science
Lawyer
English/Language arts specialist
Other non-science related career
The other racial/ethnic categories thus included non-Hispanic White,
non-Hispanic Black, non-Hispanic Asian, and Other. Too few students
indicated their racial identity as Native American or Pacific Islander to be
included independently in our analysis. Therefore, the racial category labeled
“Other” includes students who indicated they were non-Hispanic Native
American, non-Hispanic Pacific Islander, Other, and more than one race.
On the survey, students had 19 options of career interests or career
interest combinations (see Table 2). We collapsed students’ responses into
four composite variables of broader fields: medicine, health, STEM, and non-
science (i.e., not medicine, health, or STEM). The breakdown of career
- 11 -
interests was: 11% medicine, 13% health, 26% STEM, and 50% non-science
careers. We kept medicine and health as separate variables because on certain
items (especially those belonging to the pioneering factor), students interested
in medicine responded markedly differently from students interested in health
careers.

Control variables
In addition to our variables of interest, we included several other
variables to control for differences in students’ backgrounds and personalities.
First, we controlled for parental education as an indicator of socioeconomic-
related factors, which have previously been found to have an effect on career
interest (Csikszentmihalyi & Schneider, 2000). Students indicated their
parents’ highest level of education on a scale: 0 = did not finish high school
to 4 = completed a master’s degree or higher. Parental education was
calculated by averaging the education level of both parents, taking into
account students who had only one parent. These scores were normalized to
have a mean of zero and a standard deviation of one before regression.
On the survey, students could indicate whether they or their parents
were born in the U.S. or not. From these responses, we grouped students into
two categories: immigrant or non-immigrant. We described students as
immigrants if students indicated both themselves and their parents as non-
U.S. born. We described students as non-immigrants if they indicated that
they and/or at least one parent were born in the U.S. These definitions align
with the common definition of “immigrant” by the U.S. Citizenship and
Immigration Services (2015).
As an indicator of overall ability in school, we used students’ overall
SAT or ACT scores. ACT scores were mapped onto the SAT scale according
to College Board (1999). Scores were divided by 100 prior to modelling, so
any coefficients are “per 100 points.”
We included a personality trait as a control variable because we also
expected that the extent to which students are introverted or extroverted
might, for example, influence the extent to which they desire a career working
with people as opposed to objects. On the survey, students rated their
personality on a scale of 1 = introverted to 6 = extroverted. These scores were
normalized to have a mean of zero and a standard deviation of one before
regression. Of course, this variable is not a psychometrically valid and reliable
way of determining this trait; however, if it makes a difference, it may indicate
an avenue of further research.

- 12 -
Main Analysis
We carried out a linear regression with each of the four COEs—
pioneering, extrinsic, work-life balance, and people-related outcomes—as
dependent variables. We included our variables of interest and control
variables in the models to analyze any main effects. We furthermore tested
for interaction effects. P-values were adjusted with a Bonferroni correction to
account for multiple comparisons.

Missing data
Our data contained several variables with missing data. The variable
with the most missing data was the overall SAT/ACT score (2,073 missing).
To avoid the compounded data loss through listwise deleting that would have
summed to 4,040 missing observations, we performed a multiple imputation
(Rubin, 1996) that resulted in five complete datasets, each with 7505
complete responses that were pooled for analysis. In our imputation, we
included each of the variables considered in our regression model, with the
addition of several related variables (e.g., intended career at middle school,
beginning of high school, end of high school, and in college).

Normalized data
After completing our primary analyses, we found particular
racial/ethnic groups--black and Hispanic students--to display consistently
significant elevated results across each COE. While the COEs are not
mutually exclusive and it is possible for students to show a high (or low)
interest in all COEs, normalizing the COE scores for race/ethnicity allowed
us to determine the effects of race/ethnicity, net of differences in the overall
levels of COEs.

RESULTS
Main Effects
A summary of the main effects from a linear regression analysis of
each of the four COEs—pioneering outcomes, extrinsic outcomes, work-life
balance outcomes, and people-related outcomes—can be found in Table 3.
To compare across variables, the standardized coefficient (β) is reported.
We found that Asian students placed greater importance on extrinsic
outcomes compared with White students (β = 0.393, p < .001) but showed no
differences on the other three outcome expectations. Black students tended to
place a higher value on every COE than did White students (pioneering β =
0.478, extrinsic β = 0.618, work-life balance β = 0.280, people-related β =
0.193, all p < .001). Likewise, Hispanic students rated every COE as more
- 13 -
important than White students did (pioneering β = 0.273, extrinsic β = 0.388,
work-life balance β = 0.191, people-related β = 0.173, all p < .001). We found
no statistically significant differences between students we categorized as
“Other” (non-Hispanic Native American or Pacific Islander students, another
racial group, and those who indicated more than one racial group) and White
students.

Table 3
Main Effects Regression of Career Outcome Expectations

Compared with students who were primarily interested in a non-

science career, students who were most interested in a career in medicine
showed higher ratings, on average, for pioneering (β = 0.137, p < .01) and
people-related (β = 0.287, p < .001) outcomes. Students interested in a career
in health similarly had higher ratings for people-related outcomes (β = 0.307,
p < .001), but also showed a higher importance of work-life balance outcomes
(β = 0.132, p < .01). By contrast, their ratings for pioneering outcomes were
not elevated. Students interested in a STEM career placed higher importance
on pioneering outcomes (β = 0.226, p < .001), but lower importance on
people-related outcomes (β = -0.153, p < .001), compared with students
interested in a non-STEM career.
- 14 -
 Male students placed a higher importance on pioneering (β = 0.169,
p < .001) and extrinsic (β = 0.279, p < .001) outcomes compared with female
students, and a lower importance on work-life balance (β = -0.074, p < .05)
and people-related (β = -0.397, p < .001) outcomes. Students who were
classified as “immigrant” placed a higher importance on pioneering (β =
0.221, p < .001) and extrinsic (β = 0.154, p < .01) outcomes compared with
students who were either themselves or had at least one parent born in the
U.S. Students with higher SAT/ACT scores placed lower importance on
extrinsic (β = -0.047, p < .001) and work-life balance (β = -0.020, p < .05)
outcomes. Students who reported higher average parental education placed
lower importance on work-life balance outcomes (β = -0.043, p < .01).

Table 4
Interaction Effects Regression of Career Outcome Expectations

- 15 -
Interaction Effects
A second set of linear regression analyses included interaction effects
(see Table 4). Five interactions were significant. Firstly, we found three
significant interactions involving career interest variables associated with
pioneering outcomes.
There was an interaction between career interest in STEM and gender
(β = .223, p < .001), such that pioneering outcomes appeared to be more
important for male students interested in a career in STEM than for female
students interested in a STEM career (see Figure 1), although there was little
difference in the importance of pioneering factors between male and female
students who had no interest in a STEM career.

Figure 1
Differences in the importance of pioneering outcomes for male and female
participants primarily interested in a career in a STEM vs. Non-STEM field.
Shaded bands show standard error in the mean estimate.

The other two interactions were between race/ethnicity and career

interest, specifically between career interest in medicine and Asian students
(compared with White students) and between career interest in STEM and
students classified as Other (compared with White students). The effect of
these interactions is shown in Figure 2. Among those who were intending a
career in medicine, students who identified as White valued pioneering
outcomes more than students who identified as Asian, but among those who
did not want to go into medicine, we found no difference. For students
- 16 -
intending a career in STEM, students classified as Other placed lower
importance on pioneering COEs than did students who identified as White.

Figure 2
Differences in the importance of pioneering factors for White students, Asian
students, and students of other racial/ethnic groups who are primarily
interested in careers in STEM, Medicine, and Non-STEM. Vertical bars show
standard error of the mean estimate.

Secondly, we found a statistically significant interaction between

gender and extraversion for extrinsic outcomes (β = .090, p ≤ .001) and work-
life balance outcomes (β = .065, p ≤ .05). Figure 3 shows that more introverted
students do not appear to particularly care about extrinsic outcomes,
regardless of gender. Furthermore, while extraverted students tended to find
extrinsic outcomes more important than introverted students, extroverted
male students rated extrinsic outcomes as even more important than
extroverted female students. For work-life balance outcomes, we find the
opposite effect: highly extroverted students, regardless of gender, valued
these outcomes equally highly, while more introverted male students placed
much less importance on this outcome than similarly introverted female
students. There were no interaction effects for people related COEs.

- 17 -
Figure 3
Differences in extrinsic and work-life balance factors for male and female
students according to their self-reported extraversion. Shaded bands show
standard error of the mean estimate. Mean extraversions for all participants
was 3.91, with a standard deviation of 1.17.

Main Effects Renormalized for Race/Ethnicity

As mentioned above, we noticed that both Hispanic and Black
students rated all four COEs significantly higher than did other students. To
determine if Hispanic or Black students simply rated everything more highly,
or if there were, in fact, differences by race/ethnicity for individual COEs, we
normalized each of the COEs scores by race. To normalize the scores, we
subtracted the mean score of all COEs for each racial/ethnic group from an
individual student’s response, respectively. Linear regressions with the
normalized data reduced but did not completely eliminate the associations
with racial or ethnic identity (see Table 5).
We no longer found any significant differences in pioneering
outcomes associated with student race/ethnicity. The effects on the
importance of extrinsic outcomes for Asian, Black, and Hispanic students
- 18 -
were all reduced, compared with the unnormalized model. Before
normalizing, Black and Hispanic students had placed higher importance on
both work-life balance and people-related outcomes than did White students,
but after normalizing, Black and Hispanic students showed lower average
ratings for these outcomes. All interaction effects persisted unchanged after
normalizing the data, as the interaction effects in linear models are invariant
under linear transformations of the data.

Table 5
Race/Ethnicity Renormalized Regression of Career Outcome Expectations

DISCUSSION
Consistent with the work of Csikszentmihalyi and Schneider (2004),
we found that college students vary in what they hope to get out of their
careers, which we termed their COEs. Specifically, we found that college
students’ COEs differed depending on their gender, racial or ethnic
background, and desired occupation, even after controlling for other

- 19 -
background and personality factors. Additionally, some differences by race
and ethnicity persisted after normalizing the data in this respect.

Career Interests
Because we inverted the traditional relationship between COE and
career interest in our analysis, this dimension is less interesting than the others
because these associations have already been well studied in prior literature.
We included them in our analysis primarily as controls to more properly
isolate the effects of the other demographic variables.

Race and Ethnicity

One overall effect we see in comparing the results of our normalized
and unnormalized regressions is that Black and Hispanic students tend to
place a higher importance on various COEs than White students do. These
students are invested in their careers and have stronger ideas about what they
want from their careers.
After normalizing, we found differences in the relative importance of
several factors. Asian, Black, and Hispanic students placed greater
importance on extrinsic outcomes than did white students, while White
students rated work-life balance and people-related outcomes higher than did
Black and Hispanic students. It is not clear why that may be the case and
future research may want to further examine this finding. One possible
explanation is that there may be more importance placed on the traditional
(i.e., extrinsic) outcomes of success for minority students when those
outcomes have traditionally been more difficult to achieve, compared with
their White counterparts. In other words, a long-standing condition of
systemic disadvantages and even oppression may have heighted the focus
among the Black and Hispanic students on extrinsic outcomes as avenues of
advancement for themselves and their communities.

Gender
We found differences in the importance male and female students
placed on all four COEs and additional differences associated with their levels
of extraversion. These results indicate differences in values and priorities of
young men and women, and these differences exist beyond their preferences
for one kind of career. For example, in alignment with a longitudinal study of
female valedictorian high-school students who increasingly reduced their
workload to make preparations for family time as they advanced in their
careers (Arnold, 1993), we found that female students value a work-life
balance more strongly compared with male students.
- 20 -
 Why do these differences exist? Why were male students more
interested in pioneering career outcomes than female students, separately
from any differences in their preferences for careers typically associated with
those outcomes? We suspect that these differences are a result of differences
in childhood socialization, such as boys being more frequently encouraged
than girls to experiment and tinker, with toys to support such behaviors, than
girls. Moreover, male students’ strong desire for extrinsic outcomes, in
comparison with female students, aligns with the idea that money, fame, and
leadership are the keys to success for men. These aspects resonate with deep-
seated cultural definitions of what it means to “succeed” as a man.
A person’s extraversion and their gender interacted to predict interest
in extrinsic and work-life balance COEs. The differences in extrinsic COEs
exist primarily between the more extroverted men and women, while the
introverted students expressed more similar preference for these outcomes.
Meanwhile, the differences in work-life balance existed primarily between
the introverted students, with more extroverted male and female students
indicating similar preferences. These findings may have implications for
gender differences in income and leadership positions, which are extrinsic
career outcomes. However, further research is needed to understand why there
are greater differences for extroverted individuals compared to introverted
individuals.
Again, we should note that our simple measure of
introversion/extroversion was not psychometrically valid so future research
should explore whether these results hold up when measures of higher quality
are used.

Other Demographic Factors

Other demographic factors were used as control variables, but they
do tell an interesting story. Parental education level is traditionally used as an
indicator of socio-economic status (SES), as a proxy for their careers and
income levels. Our study found that students with parents who had less total
amount of education were more interested in a work-life balance. This finding
might be owed to these students experiencing difficult work-life balances in
their families while growing up. Another interesting finding is that the better
students performed on the SAT/ACT, the weaker was their desire for extrinsic
rewards in their career. It may be the case that ‘good’ students, or those who
perform better on standardized tests, are more concerned with intellectual
rewards. On the other hand, poor performing students may feel more pressure
to find a more traditional form of success (i.e., extrinsic rewards) when their
academic ability does not guarantee attainment. Additionally, immigrant
- 21 -
college students were particularly interested in a pioneering career, possibly
indicating that the idea of an American dream is still alive and well among
newcomers to the country.

CONCLUSIONS
Our study demonstrates that students vary in their COEs. Students’
race and ethnicity, followed by gender, are the strongest and most consistent
predictors of COEs, while career interests have varying degrees of influence
on college students’ COEs. Our findings regarding career interest and COEs
aligned with certain job stereotypes and previous research associating career
outcomes with career interests. Further studies could look at how shifting
career stereotypes influence career choice. There could be two mechanisms
by which COEs interact with career interests. In the first, students may have
a particular set of values (such as those gained through socialization) and may
then search out a career that matches those goals. Alternatively, students may
primarily be motivated by interest in a field or career, and their values are
subsequently shaped by exposure to the communities of practice that already
exist in that field.
If it is a goal to attract students with more varied desires for COEs,
fields like medicine, health, or STEM could work to break the stereotypical
COEs by emphasizing that other COEs can also be achieved by working in
these careers. This change may be a real change (if the outcomes associated
with the field in fact match the stereotypes), or it may be a change in
messaging and branding (if the associated outcomes reflect a distorted view
of the field). For example, Hazari and colleagues (2011) hypothesized that
promoting more balanced motivations for a career in physics and countering
stereotypes—such as this career catering mainly to intrinsic rewards—may
help attracting students from underrepresented groups who may need to focus
on external rewards like monetary compensation in consideration of their
career.
As usual with correlational research, this study can offer no causal
explanation. However, our findings do indicate a relationship between career
interests and COEs. In light of these results, as well as those of other research
on COEs and career interest, it appears likely that there is mutual feedback
within this relationship. Longitudinal research would be valuable in helping
disentangle this interdependent relationship and determine how a career
interest influences COEs, and vice versa. Furthermore, follow-up
investigations of student COEs could investigate whether or how these values
shift with generational changes over time.

- 22 -
 Research on career development has rarely examined the impact of
career interest, gender, and race/ethnicity on what students hope and expect
to get out of their careers. This study contributes to the field by identifying
differences in career outcome expectations by gender and race/ethnicity. In
an increasingly diverse workforce where racism and gender discrimination
persist, more studies are needed to understand the underlying factors that
impact career choice in minority and marginalized groups of individuals.

Acknowledgments
Funding for Project PRiSE was provided by the National Science
Foundation (grant #062444). Any opinions, findings, conclusions, or
recommendations expressed in this material are the authors' and do not
necessarily reflect the views of the National Science Foundation. The authors
would like to thank Jaimie Miller, Annette Trenga, Hal Coyle, Freeman
Deutsch, and other members of the PRiSE team for their dedicated work. We
also thank all the participating English professors and their students for
making this study possible.

REFERENCES
Adachi, T. (2004). Career self-efficacy, career outcome expectations and vocational
interests among Japanese university students. Psychological Reports,
95(1), 89-100.
Arnold, K. D. (1993). Undergraduate aspirations and career outcomes of
academically talented women: A discriminant analysis. Roeper Review,
15(3), 169-175.
Arthur, M. B., Hall, D. T., & Lawrence, B. S. (1989). Generating new directions in
career theory: The case for a transdisciplinary approach. In M. B. Arthur,
D. T. Hall, & B. S. Lawrence (Eds.), Handbook of career theory (pp. 7-
25). New York, NY: Cambridge University Press.
Bandura, A. (1986). Social foundations of thought and action: A social cognitive
theory. Englewood Cliffs, NJ: Prentice Hall.
Betz, N. E., Fitzgerald, L. F., & Hill, R. E. (1989). Trait-factor theories: traditional
cornerstone of career theory. In M. B. Arthur, D. T. Hall, & B. S.
Lawrence (Eds.), Handbook of career theory (pp. 26-40). New York, NY:
Cambridge University Press.
Blustein, D. L. (2011). A relational theory of working. Journal of Vocational
Behavior, 79, 1-17.
Brown, C., Darden, E. E., Shelton, M. L., & Dipoto, M. C. (1999). Career
exploration and self-efficacy of high school students: Are there
urban/suburban differences? Journal of Career Assessment, 7, 227-237.

- 23 -
Casper, W. J., & Buffardi, L. C. (2004) Work-life benefits and job pursuit
intentions: The role of anticipated organizational support. Journal of
Vocational Behavior, 65(3), 391-410.
College Board Office of Research and Development. (1999). Concordance between
SAT I and ACT scores for individual students. Report RN-07 (June 1999).
New York, NY: The College Board.
Conkel Ziebell, J. L. (2010). Promoting viable career choice goals through career
decision-making self-efficacy and career maturity in inner-city high school
students: A test of social cognitive career theory (Unpublished doctoral
dissertation). University of Minnesota, Minneapolis, MN.
Csikszentmihalyi, M., & LeFevre, J. (1989). Optimal experience in work and
leisure. Journal of Personality and Social Psychology, 56, 815-822.
Csikszentmihalyi, M. & Schneider, B. (2000). Becoming adult: How teenagers
prepare for the world of work. New York, NY: Basic Books.
Domene, J. F., Socholotiuk, K. D., & Woitowicz, L. A. (2011). Academic
motivation in post-secondary students: Effects of career outcome
expectations and type of aspiration. Canadian Journal of Education, 34(1),
99-127.
Fogg, N., Harrington, P., & Harrington, T. F. (2012). College majors handbook with
real career paths and payoffs: The actual jobs, earnings, and trends for
graduates of 60 college majors (3rd ed.). Indianapolis, IN: JIST
Publishing.
Fouad, N. A., & Smith, P. L. (1996). A test of a social cognitive model for middle
school students: Math and science. Journal of Counseling Psychology,
43(3), 338-346.
Gagné, M. & Bhave, D. (2011). Autonomy in the workplace: An essential
ingredient to employee engagement and well-being in every culture. In V.
I. Chirkov, R. M. Ryan, & K. M. Sheldon (Eds.), Human autonomy in
cross-cultural context: Perspectives on the psychology of agency, freedom,
and well-being (pp. 163-187). New York, NY: Springer.
Hanson, S. L. (1994). Lost talent: Unrealized educational aspirations and
expectations among U.S. youths. Sociology of Education, 67, 159-183.
Hazari, Z., Potvin, G., Lock, R. M., Lung, F., Sonnert, G., & Sadler, P. M. (2013a).
Factors that affect the physical science career interest of female students:
Testing five common hypotheses. Physical Review Special Topics –
Physics Education Research, 9(2), p.020115-1.
Hazari, Z., Sadler, P. M., & Sonnert, G. (2013b). The science identity of college
students: Exploring the intersection of gender, race, and ethnicity. Journal
of College Science Teaching, 42(5), 82-91.
Hazari, A., Sonnert, G., Sadler, P. M., & Shanahan, M. C. (2010). Connecting high
school physics experiences, outcome expectations, physics identity, and
physics career choice: A gender study. Journal of Research in Science
Teaching, 47(8), 978-1003.

- 24 -
Honeycutt, T. L., & Rosen, B. (1997). Family friendly human resource policies,
salary levels, and salient identity as predictors of organizational
attraction. Journal of Vocational Behavior, 50(2), 271-290.
Lent, R. W., Brown, S. D., & Hackett, G. (1994). Toward a unifying social
cognitive theory of career and academic interest, choice, and performance.
Journal of Vocational Behavior, 45(1), 79-122.
Maslow, A. H. (1943). A theory of human motivation. Psychological
Review, 50, 370-396.
McMahon, M. (2014). New trends in theory development in career psychology. In
G. Arulmani, A. J. Backshi, F. T. L. Leong, & A. G. Watts (Eds.),
Handbook of career development (pp. 13-28). New York, NY: Springer
Science+Business Media, LLC.
McMahon, M., Watson, M., & Patton, W. (2014). Context-resonant systems
perspectives in career theory. In G. Arulmani, A. J. Backshi, F. T. L.
Leong, & A. G. Watts (Eds.), Handbook of career development (pp. 13-
28). New York, NY: Springer Science+Business Media, LLC.
Rau, B. L., & Hyland, M. A. M. (2002). Role conflict and flexible work
arrangements: The effects on applicant attraction. Personnel
Psychology, 55(1), 111-136.
Reyes, O., Kobus, K., & Gillock, K. (1999). Career aspirations of urban, Mexican
American adolescent females. Hispanic Journal of Behavioral Sciences,
21(3), 366-382.
Riegle-Crumb, C., Moore, C., & Ramos-Wada, A. (2011). Who wants to have a
career in science or math? Exploring adolescents’ future aspirations by
gender and race/ethnicity. Science Education, 95(3), 458-476.
Rubin, D. B. (1996). Multiple imputation after 18+ years. Journal of the American
Statistical Association, 91, 473-489.
Sadler, P. M., Sonnert, G., Hazari, Z., & Tai, R. (2012). Stability and volatility of
STEM career interest in high school: A gender study. Science Education,
96(3), 411-427.
Sandberg, D. E., Ehrhardt, A. A., Mellins, C. A., Ince, S. E., & Meyer-Bahlburg, H.
F. L. (1987). The influence of individual and family characteristics upon
career aspirations of girls during childhood and adolescence. Sex Roles,
16(11), 649-688.
Su, R., Rounds, J., & Armstrong, P. I. (2009). Men and things, women and people:
A meta-analysis of sex differences in interests. Psychological Bulletin,
135(6), 859-884.
Tracey, T. J. G., & Robbins, S. B. (2005). Stability of interests across ethnicity and
gender: A longitudinal examination of grades 8 through 12. Journal of
Vocational Behavior, 67, 335-364.
U.S. Citizenship and Immigration Services. (2015). Are you the foreign-born child
of a parent who becomes a U.S. citizen? Retrieved from
https://www.uscis.gov/us-citizenship/citizenship-through-parents

- 25 -
Wigfield, A., & Eccles, J. S. (1992). The development of achievement task values:
A theoretical analysis. Developmental Review, 12(3), 265-310.
Wöhrmann, A. M., Deller, J., & Wang, M. (2013). Outcome expectations and work
design characteristics in post-retirement work planning. Journal of
Vocational Behavior, 83(3), 219-228.
Yang, Y., & Barth, J. M. (2015). Gender differences in STEM undergraduates’
vocational interests: People-thing orientation and goal affordances. Journal
of Vocational Behavior, 91, 65-75.
Endnote
1
Demographic percentages were calculated with the raw data (prior to the
multiple imputation). Hence, the percentages do not add up to 100%.

JACQUELINE DOYLE, PhD, is a data scientist at The Hanover Insurance Group.

Her major research interests focus on diversity, equity, and inclusion in STEM and
science education, with a particular interest in physics education and how to
increase student interest and engagement. Email: [email protected]

ELYSE POSTLEWAITE, PhD, is a Postdoctoral Researcher for the Institute for

Research on Youth Thriving and Evaluation at Montclair State University. Her major
research interests lie in the area of youth thriving and optimal development in and out
of school settings. Email: [email protected]

PHILIP M. SADLER, EdD, is the Director of the Harvard-Smithsonian Center for

Astrophysics’ Science Education Department and F.W. Wright Senior Lecturer in
Harvard’s Department of Astronomy. His research program includes assessment of
students' scientific misconceptions and how they change with instruction, the
transition to college of students who wish to purse STEM careers, and the
enhancement of the skills of teachers. Email: [email protected]

GERHARD SONNERT, PhD, is a senior research scientist in the Harvard College

Observatory and a lecturer on astronomy at Harvard University. His major research
interests lie in the areas of science education, gender in science, and science policy.
Email: [email protected]

Manuscript submitted: November 22, 2022

Manuscript revised: February 23, 2023
Accepted for publication: March 17, 2023

- 26 -