ArticlePDF Available

Abstract and Figures

Drawing on Eccles' expectancy-value model of achievement-related choices, we examined how work values predict individual and gender differences in sciences, technology, engineering, and math (STEM) participations in early adulthood (ages of 25/27, 6 or 8 years after postsecondary school), controlling for subjective task values attached to academic subjects in late adolescence (11th grade, age 18). The study examined 1,259 Finnish participants using a person-oriented approach. Results showed that: (a) we could identify four profile groups based on five core work values (society, family, monetary, career prospects, and working with people); (b) work-value profiles predicted young adults actual STEM participation in two fields: math-intensive and life science occupations above and beyond academic task values (e.g., math/science) and background information; (c) work-value profiles also differentiate between those who entered support-vs. professional-level STEM jobs; and (d) gender differences in work value profiles partially explained the differential representation of women across STEM sub-disciplines and the overall underrepresentation of women in STEM fields.
Content may be subject to copyright.
published: 02 July 2018
doi: 10.3389/fpsyg.2018.01111
Frontiers in Psychology | 1July 2018 | Volume 9 | Article 1111
Edited by:
Rachel Thwaites,
University of Lincoln, United Kingdom
Reviewed by:
David Stuart Smith,
Robert Gordon University,
United Kingdom
Shannon N. Davis,
George Mason University,
United States
Jiesi Guo
Specialty section:
This article was submitted to
Gender, Sex and Sexuality Studies,
a section of the journal
Frontiers in Psychology
Received: 01 February 2018
Accepted: 11 June 2018
Published: 02 July 2018
Guo J, Eccles JS, Sortheix FM and
Salmela-Aro K (2018) Gendered
Pathways Toward STEM Careers: The
Incremental Roles of Work Value
Profiles Above Academic Task Values.
Front. Psychol. 9:1111.
doi: 10.3389/fpsyg.2018.01111
Gendered Pathways Toward STEM
Careers: The Incremental Roles of
Work Value Profiles Above Academic
Task Values
Jiesi Guo 1
*, Jacquelynne Sue Eccles 2, Florencia M. Sortheix 3,4 and Katariina Salmela-Aro 3
1Institute for Positive Psychology and Education, Australian Catholic University, Sydney, NSW, Australia, 2School of
Education, University of California, Irvine, Irvine, CA, United States, 3Department of Education, University of Helsinki, Helsinki,
Finland, 4Department of Psychology, University of Jyväskylä, Jyväskylä, Finland
Drawing on Eccles’ expectancy-value model of achievement-related choices, we
examined how work values predict individual and gender differences in sciences,
technology, engineering, and math (STEM) participations in early adulthood (ages of
25/27, 6 or 8 years after postsecondary school), controlling for subjective task values
attached to academic subjects in late adolescence (11th grade, age 18). The study
examined 1,259 Finnish participants using a person-oriented approach. Results showed
that: (a) we could identify four profile groups based on five core work values (society,
family, monetary, career prospects, and working with people); (b) work-value profiles
predicted young adults actual STEM participation in two fields: math-intensive and life
science occupations above and beyond academic task values (e.g., math/science) and
background information; (c) work-value profiles also differentiate between those who
entered support- vs. professional-level STEM jobs; and (d) gender differences in work
value profiles partially explained the differential representation of women across STEM
sub-disciplines and the overall underrepresentation of women in STEM fields.
Keywords: gender differences, work values, task values, STEM, career choice
Like the labor market and optional educational courses in general, women and men are
differentially represented across the various science, technology, engineering, and math (STEM)
fields (Valla and Ceci, 2014). For example, women are underrepresented in math-intensive
fields of STEM education, such as mathematics, physical science, engineering, and computer
science (hereafter math-intensive) but overrepresented in health, biological, and medical sciences
(hereafter life science, OECD, 2014, 2016). Further, on average across OECD countries, 15-year-
old girls are almost three times more likely as boys to aspire a career in a life science field, with
the reverse being true regarding gender differences in career aspirations in math-intensive fields
(OECD, 2016). The gender disparities are also apparent with tertiary degree enrolments, where
women accounted for 78% of total enrolments in life science courses, but only 30% of total
enrolments in science and engineering courses (OECD, 2014, also see Wang and Degol, 2017).
Eccles’ expectancy-value theory has been widely used to explain individual and gender differences
in educational and career choices (Eccles, 2009). During adolescence, one’s subjective task values
(i.e., enjoyment, importance, usefulness, and negative cost) placed on different school subjects
Guo et al. Work and Academic Values
are assumed to influence academic and career pathways more
so than one’s history of academic performance. Academic task
value in a domain has been found to be positively linked to
knowledge acquisition and aspirations in said domain, which
in turn prepares and constrains one’s pursuit toward certain
educational and occupational fields (Wang and Degol, 2013, 2017
for reviews).
Theories of career choice and development have also given
personal work values a crucial role in one’s educational and
career choices (Holland, 1997; Eccles, 2009). Work or career
values are the desired characteristics of one’s current or future
job and explain individual differences in vocational interests and
career choices (Super, 1962; Judge and Bretz, 1992; Berings et al.,
2004). Career choice is assumed to be made after various career
options and their associated characteristics (e.g., money, social
connect, family-balance) have been considered. These options are
evaluated and identified as whether or not they align with one’s
personal goals, values, and preferences (Eccles, 2009). Although a
large body of research using Eccles’s expectancy-value theory has
identified various personal work values or academic task values
that contribute to gender disparities within the STEM fields,
relatively few studies have examined the joint contributions of
both critical sets of values in explaining STEM career choices
(Wang et al., 2015; Eccles and Wang, 2016; see Wang and Degol,
2017, for a review). Furthermore, Eccles’ and other value theories
suggest that the relative importance of values matters most for
guiding career pathways because choices of college major and
career are made from a variety of options and their associated
characteristics (Eccles, 2009). However, we are not aware of any
study that has taken a person-oriented approach to examine the
different intraindividual pattern of personal work values and its
association with STEM participation.
To fill these gaps, this study examines how Finnish 11th-
graders’ personal work values and academic task values affect
their actual career choices (6 or 8 years postsecondary school).
We first examine the intraindividual patterns associated with
students’ ratings of the relative importance of work values across
five domains (society, family, monetary, career prospect, and
people-orientation) using a person-oriented approach. Second,
we investigate the incremental effects of gender differences
in work value profiles on the gender gap across STEM sub-
disciplines (non-STEM vs. math-intensive vs. life science fields)
above and beyond the established effects of academic task values.
Finally, because career choice processes may vary across required
educational levels, we test the generalizability of the predictive
patterns across two educational levels of STEM professions: the
professional and the support role levels STEM fields. This study
therefore provides a comprehensive test of the psychological
mechanisms proposed by Eccles’ expectancy-value theory that
underlie individual and gender differences in educational and
occupational choices.
Work Values, STEM Career Choice, and
Gender Differences
Work values have been at the center of several prominent theories
of vocational choice and development (e.g., Super, 1990; Holland,
1997; Eccles, 2009). Over the past several decades, an enormous
body of research has demonstrated that work values are one
of the most important influences leading people to different
occupations (Su and Rounds, 2015). However, work values have
been somewhat overlooked in the literature relating to STEM
occupational fields until recently (see Diekman et al., 2015,
2017 for reviews). To date, research has drawn on a variety of
instruments and classification of work values (see Johnson et al.,
2007 for a review). In this study, we reviewed recent studies on
work values and STEM career choices and identified several work
value types that are assumed to be related to gender differences in
preferences that may affect STEM career choices.
Social Values and Working With Others
Social values refer to valuing work that allows one to directly help
people and contribute to society, which is highly related to work
with people (i.e., a job that allows one to interact and help co-
workers and work in teams). These two work value components
have been elaborated in different theoretical frameworks, such
as communal goals (e.g., Diekman et al., 2011, 2015), social
interests [e.g., (Su et al., 2009; Su and Rounds, 2015); based
on Holland’s (1997) seminal work], and people-orientation (e.g.,
Woodcock et al., 2013). STEM fields are likely to deter people
who endorse these social work values because these fields are
often considered incompatible with goals of directly benefitting
others, collaboration, or altruism (Diekman et al., 2015, 2017).
Regardless of the theoretical framework used, research has shown
that women prefer jobs where they can help and work with other
people, whereas men prefer working with objects. Such gender
differences are associated with the gender disparities in STEM
fields (Su et al., 2009; Diekman et al., 2010, 2011; Woodcock
et al., 2013). More recently, the gender differences in preferences
of men and women to social work values are found to be useful
to explain gender imbalance within STEM fields (i.e., life science
vs. math-intensive; Su and Rounds, 2015; Wang et al., 2015;
Eccles and Wang, 2016). Indeed, math-intensive fields involve a
heavy thing-orientation component, whereas other STEM sub-
disciplines such as medicine, nutrition, biology, and psychology
science (life science) are more focused on working with and
helping people and other living beings (Su and Rounds, 2015).
Men and women who placed high value on having jobs associated
with people and altruistic concerns were more likely to choose
a life science rather than math-intensive career (e.g., Su and
Rounds, 2015; Eccles and Wang, 2016). Importantly, gender
differences in valuing working with people and altruism (favoring
women) significantly explained why women are over-represented
in STEM fields that are more people-oriented and less thing-
oriented (i.e., life sciences; e.g., Su and Rounds, 2015; Eccles and
Wang, 2016).
Material Value and Status
STEM fields are often considered more likely to provide
opportunities for agentic (rather than communal) goal fulfillment
(e.g., power, status, financial rewards; e.g., Brown and Diekman,
2010; Diekman et al., 2011). For example, even 6th graders (age
12) were found to be likely to associate science with power (Jones
et al., 2000). STEM fields, particularly math-intensive fields, also
Frontiers in Psychology | 2July 2018 | Volume 9 | Article 1111
Guo et al. Work and Academic Values
dominate the list of top-earning college majors (Valla and Ceci,
2014). Material value and status have been well-documented in
different theoretical conceptualizations of work values (e.g., Sagie
and Elizur, 1996; Ros et al., 1999; Lyons et al., 2010). Material
value is related to valuing work primarily for the salary or other
compensation, and status refers to valuing work for its prestige,
power, and authority (e.g., Ros et al., 1999). Research has shown
that men tend to place more value on jobs that yield high income,
power, and prestige compared to women (Eccles et al., 1999;
Abele and Spurk, 2011). For instance, even 6- to 11-year-old boys
showed greater interest than girls in professions recognized for
their lucrative remuneration (Hayes et al., 2018). Such gender
differences have been found to impede women’s STEM pursuits
(particularly in math-intensive fields, Eccles et al., 1999; Diekman
et al., 2010, 2015).
Work-Family Balance (Family Value)
Work-family balance is another deterrent to women in STEM
fields. Valuing work-family balance is directly related to gender
role identity, with a traditional feminine identity leading one to
place more emphasis on family and less on work and the reverse
for a traditional masculine identity (Eccles, 2009). Compared to
men, women are more willing to make occupational sacrifices
for the family and prefer work-centered lifestyle at lower rates
(Diekman et al., 2015; Wang et al., 2015; Wang and Degol,
2017). This gender difference emerges in late adolescence and
young adulthood since men and women begin to consider their
future more closely (Weisgram et al., 2010). Importantly, more
recent research has found that adolescents and young adults
perceive STEM careers afford family values less than other
values such as money, power, and altruism1; the perception
that science affords family values predicts interest in pursuing
science studies/careers (Diekman et al., 2015; Weisgram and
Diekman, 2017). Taken together, research has revealed that
endorsement of work-family balance directed women away
from masculine/STEM occupations (e.g., Frome et al., 2008;
Ferriman et al., 2009; Weisgram et al., 2010; Diekman et al.,
2015), particularly professional-level (e.g., scientist, Williams and
Ceci, 2012; Mason, 2014) and math-intensive occupations (e.g.,
Computer Science, Ceci and Williams, 2011; Beyer, 2014).
While the evidence reviewed above documented that each
single work value is associated with individual’s career aspirations
and choices, the relative hierarchical importance of these values
may play a more critical role in clarifying one’s perceptions,
interests, and career goals (Jin and Rounds, 2012). The relative
hierarchy of personal work values has been well elaborated
in Eccles’ expectancy-value theory (Eccles, 2009). Specifically,
behavioral choices are assumed to depend upon a series of
value-based calculations that weigh the relative (not absolute)
subjective value across the variety of perceived available
options associated with different occupational characteristics. For
example, if one places a higher value on working with people
1The reason why STEM careers are perceived as being family “un-friendly” may be
due to the overall stereotype of scientists, media depictions of scientists, and prior
experience of science activities in school (Weisgram and Diekman, 2017, also see
Diekman et al., 2015 for a review).
than working with objects, machines, and tools, one is likely to
prefer occupations that allow them to interact with people (e.g.,
life science and humanities). Thus, people’s relative work values
channel their educational and occupational decision-making
and attainment. The extant studies, however, mainly focus on
between-person differences in different work values, which limits
our understanding of how individuals weigh up pros and cons for
each option that leads to career choices.
Based on the literature reviewed above, in this study we
focused on five core work values (a) Social value, (b) Working
with people, (c) Material value (d) Status, and (e) Life-work
balance. The inclusion of the five work values will enable us to
examine the different intraindividual patterns across various core
personal work values within sample and then to assess how these
pattern groups contribute to gender differences in occupational
choices related to STEM fields.
It is important to note that previous research has shown
that work values stabilize by late adolescence, when students’
intentions to pursue (or not to pursue) STEM majors are
crystallizing (e.g., Jin and Rounds, 2012; Lechner et al., 2017). For
example, Jin and Rounds (2012) conducted a meta-analysis study
and showed that different work values (e.g., social values and
status) are relatively stable from colleague years (ages 18–21.9) to
young adulthood (ages 22–25.9). Similarly, Lechner et al. (2017)
found such high rank-order and mean-level stability of work
values of those aged between 20 and 25. These results are in line
with a dynamic system perspective on work values development,
which posits that individuals’ value structure tends to become
more stable and coherent with age (Vecchione et al., 2012).
Although work values with STEM career were assessed at the
same time point in this study, previous research (e.g., Bardi et al.,
2014; Diekman et al., 2017) has suggested that people are more
likely to choose their career transitions based on their values
(self-selection processes) rather being socialized into their self-
chosen careers (socialization). For example, Bardi et al. (2014)
showed that in the transition to vocational training (of new
police recruits) and to different university majors (psychology
vs. business students), there were no significant value changes
which would imply socialization effects. Taken together, the
presence of self-selection processes and work values’ high stability
during postsecondary school transition support our hypothesis
that work values guide individuals’ choices toward (or away from)
STEM careers from early life stages (see below).
Incremental Role of Work Values on Career
Choices in High School
High school is a critical stage of adolescence when career
aspirations began to crystallize on the basis of youth academic
and career/work values (Eccles, 2009; Su et al., 2009). Youth
are granted options to enroll in courses that are of interest,
usefulness, and importance to them starting in high school,
further creating a divide in STEM knowledge and learning
experience between those who value and enroll in more advanced
STEM courses, and those who de-value and opt out of challenging
STEM courses. From a developmental perspective, Eccles (2009)
hypothesized that individuals develop higher academic value for
Frontiers in Psychology | 3July 2018 | Volume 9 | Article 1111
Guo et al. Work and Academic Values
tasks and careers that they perceive as being closely aligned with
their work values, leading them to preferentially select courses
that are positively linked to their personal needs and identities.
For example, a girl whose interests in activities that allow her
to interact with and help others may choose to focus on classes
and activities that fulfill her personal goals through a preference
for those related to humanities. As such, she may come to place
more academic task values on humanities than on other subjects.
Both types of task values help her accumulate knowledge and
skills associated with humanities and prepares her for entry into
humanities-related majors or careers (Eccles, 2009; Lee et al.,
Gender differences in task values attached to various academic
subjects prevalent in STEM (e.g., math/science) and non-STEM
fields (e.g., humanities, arts) are related to gendered educational
and career aspirations and choices (Eccles, 2009; Wang and
Degol, 2013, 2017). More specifically, men are likely to perceive
math and physical science more important, useful, and enjoyable
than women, whereas women are likely to have higher task values
for language, social studies, and artistic subjects (e.g., arts, music;
e.g., Chow et al., 2012; Wang et al., 2015; Eccles and Wang, 2016).
These gendered differences in the academic task values partially
contribute to overall underrepresentation of women in STEM
fields (Chow et al., 2012; Guo et al., 2015) as well as the differential
representation of gender across math-intensive and life science
fields (Su and Rounds, 2015; Eccles and Wang, 2016).
While the relative hierarchy of work values and academic task
values comprise essential parts of individuals’ identity and can
direct both men and women to different educational and career
paths, their joint contributions to the prediction of career choices
across STEM sub-disciplines (math-intensive and life science)
have rarely been investigated. Given well-documented evidence
as to the effect of academic values (Wang and Degol, 2013, 2017
for reviews), of particular interest in this study is to explore the
incremental influence of work value profiles on career choices
over and above academic values. Furthermore, recent research
has stressed the need to distinguish between occupational choice
processes for two fundamentally different occupational levels
(professional- vs. support-level; Su and Rounds, 2015). For
example, gender differences in interests in support-level life
science careers (e.g., medical services) favoring women are larger
than those in profession-level careers (e.g., medical science; Su
and Rounds, 2015). In this study, to gain a better understanding
about work values and academic task values that contribute to the
differential participation of women across STEM sub-disciplines,
career choices were operationalized into three categories: math-
intensive, life science, and non-STEM occupations, and two
social status groups: professional- and support- level occupations
(see below for more details).
The present investigation aims to examine the incremental
contribution of the work value hierarchy in predicting
individual and gender differences in STEM choices after
accounting for academic task values. By taking into account
both academic task values and work values as well as different
domains and levels of occupations within STEM fields, this
study provides a greater understanding of the motivational
dynamics leading men and women to different STEM
career pathways during transition into early adulthood. To
achieve this aim, two overarching research questions are
examined. For each research question, specific hypotheses
(predictions) and empirical analysis questions are presented as
Overarching Research questions 1 (Q1): How does the
relative work value hierarchy influence individual differences
in STEM choices above academic task value?
Question 1a (Q1a). How many distinct hierarchical patterns
(profiles) of work values will be captured? It is difficult to predict
the exact number of groups that present the qualitatively and
quantitatively distinct patterns of work values given the limited
existing empirical research. As such, we leave it as an exploratory
research question to be explored.
Hypothesis 1 (H1a). We expect that work value profiles
will significantly discriminate between people entering non-
STEM, life science, and math-intensive fields above and beyond
academic task values (Eccles, 2009; Su et al., 2009). While we
are not able to propose clear a priori expectations as to the
intraindividual patterns of work profiles, we are tempted to
derive our hypotheses from previous findings related to each
work value. As such, we expect that the groups where family
value and social value/working with others are prioritized will
be more likely to enter non-STEM fields rather than life science
and particularly careers in math-intensive STEM fields (e.g.,
Ferriman et al., 2009; Diekman et al., 2011; Woodcock et al.,
2013; Wang et al., 2015); the groups where money and career
prospect (status) values are dominant will be more likely to
pursue life science and particularly math-intensive careers over
non-STEM careers, compared to other profile groups (e.g.,
Eccles et al., 1999; Abele and Spurk, 2011; Diekman et al., 2015).
Question 1b (Q1b). Given the limited literature in
comparisons between professional-level vs. support-level
occupations within STEM fields, we leave as an open research
question whether different predictive patterns merge for
individuals entering professional-level vs. support-level STEM
Overarching Research Questions 2 (Q2): How does the
relative work value hierarchy differ by gender and influence
gender imbalance in STEM choices above academic task
Hypothesis 2a (H2a): We anticipate that men will be over-
represented in the profiles where monetary and career prospect
are more highly endorsed than other values (i.e., social value,
working with people, and family value), with the reverse being
true for women (e.g., Diekman et al., 2015; Su and Rounds,
Hypothesis 2b (H2b): These gender differences will help
explain gender imbalances in STEM choice above and beyond
the gender differences in academic task values (Eccles, 2009;
Wang and Degol, 2013, 2017).
Frontiers in Psychology | 4July 2018 | Volume 9 | Article 1111
Guo et al. Work and Academic Values
The data set used in the present study is part of the larger
Finnish Educational (FinEdu) Transition Studies. FinEdu is a
multiple wave longitudinal follow-up study initiated in 2004
tracking two cohorts comprising 675 9th graders from nine
comprehensive schools (Cohort 1, mean age =16 years) and
584 11th graders from 13 upper secondary schools (Cohort
2, mean age =18 years) in a medium-sized city in Middle
Finland. Students were tracked every two years, through high
school, higher education and employment. For this study, we
utilized questionnaire data from both cohorts on academic
task values when students were in 11th grade (2006 data for
cohort 1 and 2004 data for cohort 2, total N=1,259) as
well as on their work values and STEM participation in the
2013 follow-up (cohort 1: 6 years after postsecondary school,
mean age =25 years; cohort 2: 8 years after postsecondary
school, mean age =27 years; total N=892, 71% response
rate, see Figure 1). Girls comprised 59.2% of the sample and
almost all participants (99%) reported Finnish as their mother
Finnish Context
According to the last Programme for International Student
Assessment (PISA) survey (OECD, 2016), Finnish 15-year-old
adolescents had relatively high science and math performance
(3rd and 8th across OECD countries, respectively). However,
only 17% of them expressed their career aspirations in STEM
fields (11% in life science fields), which was much lower than
that in the U.S. (38%) and average OECD countries (25%). While
Finland is a gender equality pioneer in terms of the low gender
gaps in education, health, and economic/political participation
and offers great gender equality in work/family policies, women
are still overall under-represented in STEM fields (Esping-
Andersen, 2002; Hausmann and Tyson, 2015). The Finnish
egalitarian context, allowing equal possibilities for men and
women to pursue STEM careers, provides a unique opportunity
to investigate the underlying motivational mechanism that
directs individual and gendered career development and
Academic Task Values
Academic task values in five school subject domains, including
Finnish, math/science, humanities, foreign language, and
practical subjects/arts, were measured by the task values scale
developed from expectancy-value theory at Grade 11 (Eccles,
2009). The scale comprised three items “How interesting
(important, useful) do you think each of the following subjects
is?” to assess the interest, importance, and usefulness of each
subject domain. All task values items were coded on a 7-point
scale (from “not at all” to 7 “very much”). The domain-specific
latent task values constructs demonstrated satisfactory reliability
across time (0.78–0.86, see Appendix A in Supplementary
Material for more details).
Work Values
We used a set of 16 items derived from the Meaning of Work
Study (MOW International Research Team, 1987) and Fit-
Choice scale (Watt and Richardson, 2007) to measure five aspects
of work values 6 or 8 years after postsecondary school (age 25
or 27 depending on cohorts). The items measured monetary
(e.g., “the job allows me to earn a good salary), career prospect
(e.g., “the job provides good opportunities for upgrading and
promotion”), society (e.g., “the job allows me an opportunity
to serve society”), family (e.g., “the job has hours that fit with
family responsibilities”), and people-oriented work values (e.g.,
“the job allows me to work together with others”). Respondents
rated the importance they attached to different job characteristics
on a 7-point scale (from strongly disagree to strongly agree). Scale
reliabilities for all work values were acceptable (0.81–0.90, see
Appendix A in Supplementary Material for factor structure of the
five work values).
STEM Participation
Participants’ STEM participation was measured 6–8 years after
high school transition (ages of 25–27). It should be noted that
a special feature of Finnish educational system is the high
graduation age for academic track students in university. On
average, the age of completion of university degree in Finland
is between 25 and 28 years (Sortheix et al., 2015). As such, we
assessed participants’ actual STEM participation (i.e., studying
or working in STEM fields) based on two questions: (1) “What
is your field of study at the moment?” and (2) “What is
your professional field at the moment?”. At that time point,
54% of participants had entered the workforce, for which we
used question 2 to measure their actual STEM participation,
otherwise, we used question 1. Supplemental multiple-group
analysis indicates that separating those who have actually entered
the workforce and those who are still in studying into two groups
results in similar findings in relation to the prediction of STEM
participation (see Appendix B in Supplementary Material). In
this study, therefore, we focus on the results based on combining
two groups to avoid complication.
We operationalized STEM occupations into two subsets:
math-intensive and life science (see Eccles and Wang, 2016).
Within STEM occupations, the categories below were further
classified into support-level and professional-level occupations
based on the skills and training required (Su and Rounds,
2015), which allows us to test the generalizability of the
predictive patterns across two levels of STEM professions.
Mathematics, Physical Sciences, Engineering, Computer Science,
Health Sciences, Biological Science, Medical Science, and math
and science teachers2were categorized as the profession-level
job, whereas Science Technicians, Engineering Technicians,
Mechanics and Electronics and Medical services were categorized
as the technical-level jobs (see Su and Rounds, 2015; OECD,
2016, p. 283 for the detailed classification, also see Appendix C in
2In this study STEM occupations were defined based on the skills and training
required. Given that in Finland math and science teachers require master’s degrees
specializing in math or a specific science domain, we categorized this occupation
as part of the STEM careers, which is what has been defined in the U.S. (e.g., Miller
et al., 2015; Wang et al., 2015).
Frontiers in Psychology | 5July 2018 | Volume 9 | Article 1111
Guo et al. Work and Academic Values
FIGURE 1 | Work values profiles. Figure presented here is based on factor scores that were standardized within each individual. Percentages represent the proportion
of population classified into the respective profiles.
Supplementary Material for a classified list of majors/professions
for the present study).
Demographic Factors and Matriculation Scores
Gender was coded as 0 (women) or 1 (man). Parent occupational
status was indicated by parents’ occupations reported at Grade
11 (age 18). Matriculation examination results in Finnish and
math were also included in this study. Given that students
participated in university entrance exam at different time
points, Matriculation examination scores, the only standardized
testing in Finland throughout the whole educational career,
were collected during the postsecondary school transition. Each
participant only had one matriculation score for each subject (see
Appendix D in Supplementary Material for more details).
Analytic Strategies
Missing Data Analysis
The missing completely at random (MCAR) test (Jamshidian
et al., 2014) revealed that data was not missing complete at
random, p<0.01. To determine whether the students who
participated in 11th grade differed from those who dropped out
between the ages of 18 and 27, a series of independent samples
contingency table analyses and t-tests were conducted with both
demographic variables and other variables used in the analyses.
We found men were significantly more likely to drop out of
the study than women during the post-high school transition
(t=5.32, p<0.00). Participants with lower GPA at Grade 11
(0.40 SD) were also significantly more likely to drop out of the
study during the transition (t=2.34, p<0.00). It should be
noted that missing data were not associated with work values
and academic task values. In all analyses, we operated under
the assumption that data were not MCAR but were missing
at random. Full Information Maximum Likelihood estimation
was used to cope with the missing data. Gender and GPA at
Grade 11 were included as auxiliary variables in the data analyses
[confirmatory factor analysis (CFA) and latent profile analysis
In the present study, analyses were conducted with Mplus
7.13 using the robust maximum likelihood estimator. First, we
conducted a CFA to examine the factor structures of work values
and academic task values. Subsequently, LPA, a person-oriented
modeling technique, was used to identify characteristically
distinct sub-populations of work values across the five domains.
It assumes that there is an underlying categorical latent variable
that characterizes an individual’s class or profile based on the
observed data (Muthén, 2001). LPA is a probabilistic model-
based method in which estimated posterior probabilities of class
membership are used to group individuals into latent classes. It
should be noted that factor scores of the work values saved from
preliminary measurement models were used in the LPA (Morin
and Marsh, 2015). Particularly, we standardized five factor scores
of work values within each individual before conducting LPA,
which allowed us to disentangle shape differences from level
effects (Morin and Marsh, 2015). Several indicators were used
to select the optimal number of profiles (groups): the Akaike
Information Criterion (AIC), the Consistent AIC (CAIC), the
Bayesian information criterion (BIC), and the sample-adjusted
BIC (SABIC). A lower value on these indicators suggests a better-
fitting model. These information criteria should be graphically
presented through “elbow plots” illustrating the gains associated
Frontiers in Psychology | 6July 2018 | Volume 9 | Article 1111
Guo et al. Work and Academic Values
with additional profiles (Morin and Marsh, 2015). In these plots,
the point after which the slope flattens indicates the optimal
number of profiles in the data. To further secure our decision in
selecting the best model, we used the adjusted likelihood ratio test
(LMR-LRT) and the bootstrap likelihood ratio test (BLRT) (Lo
et al., 2001). Nonsignificant LMR-LRT and BLRT tests indicate
that a model with k-1 profile model would provide a better fit
compared to a kprofile model. Finally, we also relied on the
Entropy Index that summarizes classification accuracy (Lubke
and Muthén, 2007). The entropy varies from 0 to 1, with higher
values indicating fewer classification errors. While there appears
to be no definitive criteria for determining optimal numbers
of latent classes when estimating LPA models, researchers have
recommended the use of multiple statistical indices, along
with conceptual considerations and interpretability of the latent
groups (Morin and Marsh, 2015).
Second, based on the LPA results, a series of hierarchical
regressions were conducted to explore how the work value
profile memberships and academic task values predict STEM
participation and gendered effect on STEM participation.
Mixture models in Mplus provide class membership probabilities
for each individual. Rather than using an “all-or-none” approach
of assigning class membership to participants based on the
highest probability for one of the profiles, we employed each
individual’s estimated probability of membership for each class
as sampling probabilities (i.e., CPROB1-CPROB4 in SAVEDATA
of Mplus output) to 25 created imputations of class membership
and combined them with the original sample (Sahdra et al., 2017).
The subsequent hierarchical regression analyses are based on
25 imputations in Mplus. All data analyses were run separately,
and the results were aggregated appropriately in order to obtain
unbiased estimates (Rubin, 1987). Thus, this approach allows
us to account for classification uncertainty in the latent class
membership3and test mediation effects for gender.
The third step of our analyses focused on the question
whether work value profiles and academic task values can
explain the expected gender gap in STEM aspirations (Q3). We
approached this question by testing how much the expected
gender differences in STEM participation would be reduced by
adding work value profiles and academic task values to the model
using hierarchical regression analyses. That is, we calculated
the magnitude of the relative indirect effects of gender (Huang
et al., 2004; also see Wang et al., 2015), which can be loosely
interpreted as the percentage reduction in the unstandardized
regression coefficients (b) of gender between Model 1 (bM1) and
Model 2 (bM2):
3Note that a three-step approach implemented in Mplus also allows us to control
for measurement error in profile membership (Asparouhov and Muthén, 2014).
However, only DCATEGORICAL command is available for categorical distal
outcomes (i.e., STEM participation) in the three-step approach (Muthén and
Muthén, 1998-2012, p. 557). But DCATEGORICAL cannot be combined with
other continuous covariates in LPA. In other words, it is unfeasible to examine
the effect of class membership on STEM participation while controlling for the
continuous covariates (e.g., demographic factors and academic task values) using
the three-step approach with DCATEGORICAL (even using the manual three-
step approach proposed in Asparouhov and Muthén, 2014). Thus, an alternative
approach accounting for classification uncertainty was used in this study.
1b=bM1 bM2
Preliminary CFA
Results showed that the CFA model, in which five work
values and five subject-specific values were included, fit the
data well [e.g., the comparative fit index (CFI) =0.941, the
Tucker–Lewis index (TLI) =0.928, the root mean square error
of approximation (RMSEA) =0.042, see Appendix A in
Supplementary Material for the factor structure).
General Descriptive Results for Gender
Two sets of mean testing of work and academic task values
within- and between-gender differences were examined (see
Appendix E in Supplementary Material). For within-gender
comparison, women placed relatively high values on the
Finnish subject and the lowest values on math and science;
the reverse was true for men. Men placed relatively higher
value on monetary rewards and career growth, whereas
women placed relatively higher value on work-life balance
than on other job characteristics. More women chose non-
STEM than STEM occupations, whereas men were slightly
overrepresented in STEM occupations. More women entered
life science than math-intensive fields; men did the reverse.
However, when only professional-level jobs were considered,
the difference in entering life science vs. math-intensive
fields became substantively smaller for women, but not for
For between-gender comparison, men had higher math and
science values than women, whereas women had higher values
than men in the other four academic domains. Men placed
more value on salary and career prospect than women; women
placed more value than men on family, working with people,
and altruism. Within STEM, men were over-represented in math-
intensive fields; women were over-represented, but to a lesser
extent, in life science fields. The pattern of results was similar
when only professional-level jobs were considered (see Appendix
E in Supplementary Material).
Classes Description
The values for AIC, BIC, CAIC, and SABIC for the one- to
seven-profile solutions continued to decrease with the addition
of profiles (see Table 1). LMR-LRT became non-significant after
four-profile solutions, whereas significant BLRT was showed
from two- to seven-profile solutions. In accordance with previous
recommendations, we also relied on elbow plots to help in the
selection of the final solution. The elbow plot showed a relatively
clear plateau at four profiles, after which improvement in fit
became marginal (see Appendix F in Supplementary Material).
Examination of the 4-profile solution shows them to be fully
proper and interpretable. Examination of adjacent 3- and 5-
profile solutions confirmed the added value of the 4-profile
solution compared to the 3-profile solutions, and the lack of
added value of the 5-profile solution which resulted in the
estimation of additional very small profiles (4.1%) that brought
Frontiers in Psychology | 7July 2018 | Volume 9 | Article 1111
Guo et al. Work and Academic Values
no new information to the model (i.e., had the same shape as
other profiles). Finally, entropy values suggested classification
qualities for the models with four-profile solution was reasonable
(0.858). More importantly, four distinct profiles provided
substantive interpretation, in which monetary, prospect, family,
and society values discriminate between profiles (e.g., Eccles,
2009; Diekman et al., 2015, 2017; Wang and Degol, 2017). Thus,
we retained it as our final solution.
It should be noted that the values (means of each domain
in each profile) presented in Figure 1 were based on factor
scores that were standardized within each individual across the
five work values. Thus, the histogram bars cannot be directly
compared across different work value profiles. Profile 1 was
characterized by relatively higher monetary importance, followed
by traditional career prospects. This group also attached relatively
less value on social contribution and working with people. We
labeled this profile Monetary-oriented. It described 23.4% of the
participants (see Table 2). In Profile 2 (26.1%), career prospect
was rated the most important and family needs were rated as
least important. Thus, we labeled this group Prospect-oriented. In
contrast, Profile 3 (28.1%) rated family values the most important
and career prospects the least important; we labeled this group
Family-oriented. Profile 4 (22.5%) rated society contribution and
working with people the most important and monetary rewards
the least important; we labeled it Society-oriented.
Prediction of Stem Participation
Prediction of Math-Intensive and Life Science
We conducted hierarchical logistic regression to determine
which variables discriminate between people who entered life
science, math-intensive, and non-STEM careers. In the first set of
regression models, we only included gender, parent occupational
status, and matriculation scores as the predictors (see Table 3).
Gender significantly predicted STEM occupation across three-
pair comparisons. Math matriculation scores had a small effect on
entry into math-intensive over life science and non-STEM fields.
Second, we added five subject-specific task values, which
significantly predicted individual differences in the choice of non-
STEM vs. life science and math-intensive fields (see Table 3).
Consistent with previous research, math/science values were
positively associated with entry into life science over non-STEM
fields, whereas humanities values were positively associated
with entry into non-STEM over life science fields. Similarly,
math/science values were positively associated with entry into
math-intensive over non-STEM fields. Academic values in
humanities and foreign language were positively associated
with entry into non-STEM over math-intensive fields. However,
individual differences in the academic values did not explain
differences in entry into life science vs. math-intensive fields.
Finally, we added the four work value profiles into
the hierarchical regression model predicting STEM choices,
controlling for all the variables previously included in the model
(see Table 4). We also report odds ratios, reflecting the change
in likelihood of entering life science vs. math-intensive fields
associated with people being in a target work value profile vs.
a comparison profile. For example, an OR (Odds Ratio) of
3 suggests that individuals in a target profile (compared to a
comparison profile) are three times more likely to enter life
science over math-intensive fields. As expected (H1a), individuals
from the Society-oriented group had the greatest likelihood of
choosing life science over math-intensive fields, followed by those
from the Family-oriented group. In line with this, individuals
in the Society-oriented group were more likely to enter non-
STEM over math-intensive fields, compared to those from other
groups. Individuals in the Prospect-oriented and Monetary-
oriented groups were more likely to enter math-intensive fields
over non-STEM fields, compared to individuals in other profiles.
Both Prospect- and Monetary-oriented groups had a small and
similar likelihood of pursing life science over math-intensive
fields, however, the work value profiles were not significant
predictors of entry into life science vs. non-STEM fields.
Results also showed that the effects of academic task values
remain significant after including work values profiles in the
model. A more detailed examination of the relations between
academic task values and work value profiles are presented in
Appendix G in Supplementary Material.
Profession-Level vs. Support-Level STEM Fields
We ran the regressions separately for individuals entering
professional- and support-level STEM fields to compare the
predictive pattern for entering professional-level STEM vs. non-
STEM fields, with that for entering support-level STEM vs. non-
STEM fields (see Table 5). Three major differences emerged.
Firstly, the Society-oriented group had a greater likelihood of
attaining support-level life science rather than support-level
math-intensive careers, followed by the Family-oriented group.
Second, the Society- and Family-oriented groups were more
likely to pursue support-level life science over non-STEM careers
compared those in the other profiles. Finally, compared to the
Family-oriented group, the Society-oriented group was more
likely to enter non-STEM vs. support-level math-intensive fields,
whereas such an effect was insignificant when looking at entering
non-STEM vs. profession-level math-intensive fields.
Gender Effects
As seen in Table 3 (also see Appendix E in Supplementary
Material), there are significant gender differences in work value
profiles and mean-level academic task values. Men were over-
represented in the Monetary-oriented profile, whereas women
were over-represented in the Family- and Society-oriented
profiles. Men placed relatively high values on math and science
and the lowest values on Humanities and languages. As predicted
(H2a), academic task values explained (2.03–1.75)/2.03 =14% of
the gender differences (i.e., the relative indirect effects of gender)
in entering non-STEM vs. math-intensive fields (comparing
STEP1 to STEP2 in Table 4). When work value profiles and
academic values were included, both sets of values explained
(2.03–1.21)/2.03 =40% (comparing STEP1 to STEP3, in Table 5)
of gender differences in entering non-STEM vs. math-intensive
fields. Taken together, results indicated that work value profiles
further explained 40–14% =26% of gender differences in
entry into non-STEM over math-intensive fields, controlling for
academic values. Similarly, results also indicated that further
Frontiers in Psychology | 8July 2018 | Volume 9 | Article 1111
Guo et al. Work and Academic Values
TABLE 1 | Fit indices from LPA models.
1-Class 10 8147.816 16315.631 16367.012 16335.247 16377.013 NA NA NA
2-Class 21 7396.78 14835.559 14943.459 14876.753 14964.460 <0.001 <0.001 0.894
3-Class 32 7061.118 14186.236 14350.655 14249.008 14382.654 0.004 <0.001 0.884
4-Class 43 6727.368 13540.736 13761.673 13625.085 13804.673 0.006 <0.001 0.858
5-Class 54 6655.016 13418.03 13749.49 13695.49 13523.96 0.076 <0.001 0.858
6-Class 65 6598.419 13326.84 13725.81 13660.81 13454.34 0.834 <0.001 0.869
7-Class 76 6548.516 13249.03 13715.53 13639.53 13398.11 0.832 <0.001 0.881
LL, Model log-likelihood; #fp, number of free parameters; AIC, Akaike’s Information Criterion; CAIC, consistent AIC; BIC, Bayesian Information Criterion; ABIC, sample-size adjusted
BIC; LMR, Lo-Mendell-Rubin likelihood ratio; BLRT, p-value for bootstrap likelihood ratio test.
TABLE 2 | Gender distribution across the four work value profiles.
N=295 (23.4%)
Prospect oriented
N=329 (26.1%)
Family-oriented N =355
Society-oriented N =280
Mean test between
Mean (SE) Mean (SE) Mean (SE) Mean (SE) F(3, 1259)
Gender distribution 55.36***
Women 123 (16.5%)a200 (26.8%)a224 (30.1%)a198 (26.6%)a
Men 172 (33.5%)b129 (25.1%)b131 (25.5%)b82 (16.0%)b
***p<0.001. aThe percentage of women; bThe percentage of men.
inclusion of work value profiles in the hierarchical regression
model (STEP3) explained a notable proportion of gender
differences in entering life science vs. math-intensive fields (24%),
however, work value profiles or academic and task values did not
close the gender gap in entering life science vs. non-STEM fields
(the gender effects got slightly larger between STEP1 and STEP3).
Table 6 summaries key findings of the present study and indicates
whether these findings supported our expectations. By employing
a person-oriented approach and incorporating two crucial sets of
predictors of STEM participation (work values and academic task
values), the findings provided evidence of how intraindividual
patterns of work values contribute to the gendered career
pathways above and beyond academic task values.
One of the unique contributions of the current study is
identifying four relative-priority profiles of core personal work
values and linking them to long-term STEM participation (6 or 8
years postsecondary school). As hypothesized, individuals in the
Family- and Society-groups moved toward life science and non-
STEM occupational pathways. Conversely, those in the Prospect-
and Monetary-oriented groups moved toward math-intensive
rather than other fields. These findings are inconsistent with
previous studies based on a variable-oriented approach showing
that work values placed on monetary rewards and career-
focus did not predict STEM participation when controlling
for other work values (Wang et al., 2015; Eccles and Wang,
2016). A potential explanation is that previous studies focused
on between-person differences in work values and found that
students tended to rate monetary/career-focused values similarly
(indicated by smaller standard deviations, see Wang et al., 2015
for more details) relative to other work values, thus leading to
insignificant effects on STEM aspirations. These between-person
differences in work values could mask individual decision-
making processes in career choices. Instead, using a person-
oriented approach allows us to assess the nuanced details about
how individuals prioritize different work values and weight
different options. Students who place importance on potential
earned income and career prospects than other work values
would move toward math-intensive fields that dominate the list
of top-earning college majors and yield more predictable career
advancement pathways (Valla and Ceci, 2014).
However, inconsistent with our expectations, work value
profiles did not differentiate between those who entered life
science vs. non-STEM fields. By further exploring the predictive
pattern for individuals entering professional- and support-level
STEM fields, results indicate that individuals in the family-
and society-oriented profiles are more likely to move toward
support-level life science vs. non-STEM fields (but these profiles
did not discriminate among those choosing profession-level life
science vs. non-STEM fields). Indeed, support-level life science
occupations may be perceived to more directly link to working
with people as well as to be more time flexibility and fewer
professional responsibilities, such as counselors and nurses,
compared to profession-level life science (e.g., bioengineers,
epidemiologist) and non-STEM fields (Kimmel et al., 2012).
These occupational characteristics make support-level life science
careers more attractive for individuals prioritizing social and
family values. Following the same motivational mechanism,
being in the Family-oriented or Society-oriented groups increased
the likelihood of choosing a support-level life science rather
Frontiers in Psychology | 9July 2018 | Volume 9 | Article 1111
Guo et al. Work and Academic Values
TABLE 3 | Hierarchical logistic regression predicting individuals’ participation in math-intensive and life science fields.
Predictors STEP 1 STEP 2
Non-STEM vs.
Life science vs.
Life science vs.
Non-STEM vs.
Life science vs.
Life science vs.
coef OR coef OR coef OR coef OR coef OR coef OR
Gender 2.03 (0.22)*** 0.13 2.50 (0.27)*** 0.08 0.47 (0.20)* 0.63 1.75 (0.23)*** 0.17 2.44 (0.27)*** 0.09 0.69 (0.21)** 0.50
0.24 (0.13) 1.27 0.30 (0.16) 1.48 0.15 (0.11) 1.16 0.21 (0.13) 1.23 0.30 (0.16) 1.42 0.14 (0.11) 1.15
0.01 (21) 1.01 0.11 (0.26) 1.12 0.11 (0.18) 1.12 0.01 (0.21) 1.01 0.12 (0.26) 1.13 0.11 (0.18) 1.12
Math matriculation 0.36 (0.11)** 0.70 0.44 (0.13)** 0.64 0.19 (0.20) 0.83 0.22 (0.20) 0.80 0.32 (0.27) 0.73 0.20 (0.21) 0.82
Finnish 0.08 (0.15) 0.73 0.13 (0.17) 1.19 0.21 (0.12) 1.35
Math and Science 0.75 (0.13)*** 0.47 0.23 (0.14) 0.79 0.52 (0.10)*** 1.68
Humanities 0.69 (0.15)*** 1.99 0.18 (0.15) 1.20 0.51 (0.12)*** 0.60
Foreign language 0.22 (0.13) 1.24 0.01 (0.13) 0.99 0.23 (0.11)* 0.79
Practical subjects
and arts
0.03 (0.12) 1.03 0.10 (0.13) 1.11 0.07 (0.10) 1.07
*p<0.05, **p<0.01, ***p<0.001.
than math-intensive careers (see Table 6). In contrast, work
value profiles did not predict differences at the professional-level
between life science vs. math-intensive choices, indicating that
young adults are likely to believe that professional-level STEM
careers have similar characteristics, such as entailing similar
level of family-work balance conflict (e.g., Diekman et al., 2015;
Eccles and Wang, 2016). These distinctions in the prediction of
STEM participation provide a greater understanding of how work
values motivate men and women to enter different STEM sub-
disciplines and underscore the importance of assessing pathways
leading to profession- and support-level STEM fields separately.
Consistent with previous studies (Wang and Degol, 2013,
2017), high task values in math and science coupled with low
values in humanities moved both genders toward STEM career
pathways, however, academic task values did not differentiate
between those people going into life science vs. math-intensive.
A possible explanation is that math and science values were
measured as a single motivational construct, while different
STEM occupations require different levels of math cognitive
ability and domain-specific science skills. More recent studies
have shown that high school students can distinguish academic
values in math and different science domains, although students
appear to have similar levels of achievement in those subjects
(e.g., Guo et al., 2017).
The pattern of results regarding the gender gaps in task
values and work values and their contribution to gendered career
pathways herein were largely aligned with our expectations.
Women’ participation in STEM was partially influenced by the
ways they prioritize different core personal work values and view
different academic subjects, according to their personal goals
and identities. Women viewed working with and helping people,
and committing to family responsibilities, as important personal
goals in the same way they placed high values on humanities in
high school. Both value beliefs increased the likelihood of women
moving away from math-intensive classes and activities in school.
The opposite is true for men. For both genders, these processes
constrain their options in educational and occupational pursuits
of STEM and non-STEM fields, however, gender difference
in both sets of value beliefs did not explain entry into life
science vs. non-STEM fields. This suggests that women generally
may believe that life science careers are congruent with their
personal goals, values, and preferences, as are the values of
non-STEM careers, which may help explain why women are
over-represented in life science fields. Thus, more research is
needed to include other social-psychological factors (e.g., gender-
related stereotypes and biases) and provide a comprehensive
picture of why such gender differences exist and how they are
Limitations and Further Research
Several limitations and caveats of this study must be noted. First,
we did not include academic ability self-concept (i.e., expectancy
of success), a variable that has been shown to significantly
predict STEM participation (Wang and Degol, 2013), even
though academic task values are better predictors of choice
behaviors (Wang and Degol, 2013). Additionally, as mentioned
earlier, math and science task values were operationalized as a
single construct, which substantially limited our ability to detect
motivational mechanism channeling people to different STEM
sub-disciplines. Thus, the further inclusion of multiple domain-
specific science expectancies and task values would provide a
more comprehensive understanding of the roles of motivational
beliefs in shaping career pathways.
Third, the current findings are correlational and no causal
inferences should be made. Particularly, the data related to core
personal work values and STEM participation were collected
Frontiers in Psychology | 10 July 2018 | Volume 9 | Article 1111
Guo et al. Work and Academic Values
TABLE 4 | Work value profiles and academic task values predicting individuals’ participation in math-intensive and life science fields (Con’t).
Predictors STEP 3
Non-STEM vs.
Life science vs. Math-intensive Life science vs. Non-STEM
coef OR coef OR coef OR
Gender 1.21 (0.34)*** 0.30 1.84 (0.28)*** 0.16 0.63 (0.21)** 0.53
Parent occupational status 0.21 (0.14) 1.23 0.29 (0.16) 1.45 0.16 (0.12) 1.17
Finnish matriculation 0.02 (0.20) 1.02 0.09 (0.24) 1.09 0.11 (0.19) 1.12
Math matriculation 0.10 (0.20) 0.90 0.30 (0.19) 0.69 0.26 (0.21) 0.77
Finnish 0.08 (0.16) 0.99 0.08 (0.17) 1.08 0.16 (0.12) 1.28
Math and Science 0.70 (0.14)*** 0.50 0.16 (0.15) 0.85 0.54 (0.13)*** 1.55
Humanities 0.73 (0.17)*** 2.08 0.24 (0.16) 1.27 0.49 (0.14)*** 0.68
Foreign language 0.26 (0.14) 1.27 0.02 (0.14) 1.02 0.24 (0.12)* 0.80
Practical subjects and arts 0.03 (0.13) 0.97 0.03 (0.14) 1.03 0.06 (0.11) 1.06
Vs. P1 (Monetary-oriented)
P2 (Prospect-oriented) 0.24 (0.28) 1.27 0.21 (0.33) 1.51 0.03 (0.27) 1.19
P3 (Family-oriented) 0.66 (0.28)* 1.93 1.09 (0.34)** 2.97 0.43 (0.26) 1.54
P4 (Society-oriented) 1.41 (0.42)** 4.10 1.91 (0.47)*** 6.75 0.50 (0.28) 1.65
Vs. P2 (Prospect-oriented)
P3 (Family-oriented) 0.42 (0.29) 1.52 0.68 (0.32)* 1.97 0.26 (0.23) 1.30
P4 (Society-oriented) 1.17 (0.41)** 3.22 1.50 (0.44)** 4.48 0.32 (0.24) 1.38
Vs. P4 (Society-oriented)
P3 (Family-oriented)0.75 (0.36)* 0.47 0.82 (0.37)* 0.45 0.07 (0.22) 0.93
*p<0.05, **p<0.01, ***p<0.001.
within a single wave. Although some of work value scales4
(i.e., monetary, career prospects) included in this study have
been shown to be highly stable during postsecondary school
transition (from age 20 to 25) based on the same sample
(Lechner et al., 2017), the current data still does not allow us
to examine the stability of the four work value profiles and
their long-term prediction of STEM participation across time.
Indeed, the transition from education to employment requires
individuals to invest in new social roles and to adapt their
behaviors and motivations to these new roles’ requirement. Even
though entering the workforce would potentially affect one’s
work values, recent studies have shown that family and work
transitions have very small and limited effects on work values,
especially when compared to stable background characteristics
such as gender and family socioeconomic status (Sortheix et al.,
2015; Lechner et al., 2017). Still, the robustness of our findings
could be strengthened by carefully constructed longitudinal panel
studies and experimental interventions to better understand the
causal mechanisms in the career decision-making process.
Fourth, given that different fields within non-STEM fields
involve different levels of earning prospects, social interaction,
and math-intensity (e.g., economic sciences vs. archeology),
4Another three work value scales (i.e., Family-work balance, social values, working
with others) included in this study was only assessed when participants are at age
25 (Cohort 1) or age 27 (Cohort 2), and thus we are not able to test the stability of
these three constructs over time.
it would be beneficial to further differentiate fields across
non-STEM sub-disciplines and explore how work value profiles
contribute to individual and gender differences in entry into
different career fields.
Fifth, the participants of this study were drawn from
central Finland, which did not allow to examine and compare
the roles of socio-cultural and national differences in family,
school, and work environment. For example, nations also
differ in the perception of the gendered stereotypes linked
to STEM and non-STEM occupations (Eccles and Wang,
2016). Thus, the cross-cultural variations in socialization and
gender-role processes that influence choices of occupational
pathways indicate that more comparative studies in more diverse
settings are needed to advance our understanding of career
Lastly, this study found that gender imbalance in the Society-
oriented profile (favoring women) significantly contribute to
gender differences in STEM representations. Relatedly, Baron-
Cohen (2003) offered a similar explanation from a biological
perspective, arguing that men and women have different brain
types and women are better at empathizing (vs. systematizing)
and more interested in careers involving social relations.
However, the extent to which such gender differences reflect
genetically based or hormonally based biological process,
or social cultural processes, or more likely the interaction
between these two broad types of developmental forces is
not the focus of our paper and cannot be determined with
Frontiers in Psychology | 11 July 2018 | Volume 9 | Article 1111
Guo et al. Work and Academic Values
TABLE 5 | Separate groups regression predicting STEM participation for individuals entering profession-level vs. support-level fields.
Predictors Profession-level Support-level
Non-STEM vs.
Life science vs.
Life science vs.
Non-STEM vs.
Life science vs.
Life science vs.
coef OR coef OR coef OR coef OR coef OR coef OR
Gender 1.11 (0.30)** 0.33 1.95 (0.48)** 0.14 0.61 (0.23)* 0.54 1.53 (0.42)*** 0.22 1.65 (0.45)*** 0.19 0.51 (0.24)* 0.60
Parent occupational status 0.12 (0.19) 1.13 0.24 (0.18) 1.27 0.42 (0.24) 1.52 0.23 (0.18) 1.26 0.32 (0.24) 1.38 0.14 (0.14) 1.15
Finnish matriculation 0.12 (0.23) 0.89 0.35 (0.36) 1.42 0.47 (0.31) 1.60 0.12 (0.31) 1.13 0.15 (0.35) 1.16 0.03 (0.22) 1.03
Math matriculation 0.23 (0.24) 0.79 0.12 (0.39) 0.89 0.11 (0.37) 1.12 0.14 (0.33) 1.15 0.35 (0.39) 0.70 0.49 (0.26) 0.61
Finnish 0.53 (0.18)** 0.59 0.21 (0.27) 0.81 0.31 (0.13)* 1.36 0.07 (0.21) 1.07 0.22 (0.23) 1.38 0.15 (0.11) 1.28
Math and Science 1.00 (0.17)*** 0.37 0.09 (0.26) 1.09 1.09 (0.22)*** 2.97 0.45 (0.17)** 0.64 0.03 (0.19) 0.97 0.42 (0.11)*** 1.52
Humanities 0.71 (0.17)*** 2.03 0.14 (0.26) 1.15 0.56 (0.24)* 0.57 0.78 (0.20)*** 2.18 0.34 (0.20) 1.40 0.44 (0.12)*** 0.64
Foreign language 0.14 (0.17) 1.15 0.27 (0.19) 0.76 0.41 (0.14)** 0.66 0.35 (0.14)* 1.42 0.12 (0.16) 1.13 0.23 (0.11)* 0.79
Practical subjects and arts 0.08 (0.14) 1.08 0.01 (0.19) 0.99 0.10 (0.16) 0.90 0.14 (0.17) 0.87 0.06 (0.19) 0.94 0.08 (0.10) 1.08
Vs. P1 (Monetary-oriented)
P2 (Prospect-oriented) 0.45 (0.34) 1.57 0.22 (0.49) 1.25 0.23 (0.42) 1.26 0.25 (0.39) 1.28 0.61 (0.47) 1.84 0.36 (0.33) 1.43
P3 (Family-oriented) 0.94 (0.36)** 2.56 0.22 (0.55) 1.25 0.73 (0.49) 2.08 0.50 (0.37) 1.65 1.31 (0.45)** 3.71 0.81 (0.32)* 2.25
P4 (Society-oriented) 1.27 (0.51)* 3.56 0.92 (0.69) 2.51 0.35 (0.51) 1.42 1.58 (0.62)* 4.85 2.51 (0.66)** 8.30 0.93 (0.34)* 2.53
Vs. P2 (Prospect-oriented)
P3 (Family-oriented) 0.50 (0.38) 1.65 0.00 (0.53) 1.00 0.50 (0.44) 1.65 0.25 (0.40) 1.28 0.70 (0.33)* 2.01 0.45 (0.26) 1.57
P4 (Society-oriented) 0.82 (0.40)* 2.27 0.70 (0.66) 2.01 0.12 (0.48) 1.13 1.32 (0.52)* 3.42 1.90 (0.54)** 6.68 0.57 (0.30) 1.60
Vs. P4 (Society-oriented)
P3 (Family-oriented) 0.33 (0.51) 1.39 0.70 (0.70) 0.50 0.38 (0.52) 1.46 1.07 (0.52)* 2.91 1.20 (0.53)* 0.30 0.12 (0.24) 1.03
*p<0.05, **p<0.01, ***p<0.001.
the kind of data we have. Given the very classic nature
of this debate, we encourage researchers to pursue this
Implications and Practice
Despite these limitations, the current study has implications
for intervention and practice. Firstly, the findings suggest that
individuals in the Monetary-oriented groups (those who placed
the least importance on social value) have the greatest likelihood
of choosing math-intensive careers. This, however, does not
mean that interventions aim at fostering math/science-course
participation should focus on promoting student’s monetary
values, which are more highly valued by men. Rather, to reduce
the gender gap in math/science-course taking, STEM educators
could place greater emphasis on demonstrating the societal
relevance of math-intensive skills and careers as well as how
math-intensive fields can be collaborative and beneficial to
society and have the ability to improve people’s lives (Diekman
et al., 2015; Wang and Degol, 2017). By doing this, we may be able
to make STEM careers, particularly for math-intensive careers,
more relatable and accessible to women in everyday life and
thus attract more women to participate in these types of STEM
activities (Valla and Ceci, 2014; Su and Rounds, 2015). At the
same time, special efforts should also be made to ensure that
students are well-informed of the whole variety of occupational
options in STEM fields and what characteristics are attached to
those occupations. This information enables women and men to
better relate their personal goals and identities to different STEM
Second, relatively high family-value coupled with relatively
low career prospect work value is another key factor directing
people away from math-intensive careers. Given the STEM
labor market shortages, interventions promoting women’s career
ambition and helping them view potential bright future in
STEM careers might be useful in recruiting science-talented and
capable women to embark on STEM career paths, particularly
in math-intensive fields. Additionally, integrating more family-
friendly workplace policies in math-intensive fields might make
these professions more enticing to people wanting better work-
life balance. Such policy moves could counter the stereotype
that math-intensive careers are more time demanding and
higher work commitment than non-STEM careers (Diekman
et al., 2015). This conception comes into direct conflict with
women’s work-life balance. Interventions designed to eliminate
this conception might be an effective way to increase women’s
participation in math-intensive fields. Interestingly, such a
perceived stereotype does not exist within life science and non-
STEM fields.
Third, high math and science task values help to propel
students toward STEM pathways. Interventions designed to
increase students’ perceptions of the relevance of math and
science to their lives through teachers and parents have been
Frontiers in Psychology | 12 July 2018 | Volume 9 | Article 1111
Guo et al. Work and Academic Values
TABLE 6 | Summary for the key findings.
Hypothesis Support for predictions Inconsistent with predictions Leave as research questions
Q1 How does the relative
work value hierarchy
influence individual
differences in STEM
choices above
academic task value?
H1a: work value profiles significantly discriminate
between people entering STEM fields.
Likelihood of choosing Non-STEM vs.
math-intensive fields:
society >family >prospect =monetary;
Likelihood of choosing life science vs.
math-intensive fields:
society >family >prospect =monetary.
The work value profiles did not
predict entry into life science vs.
non-STEM fields.
Q1a: Number of work value profiles.
Four profiles (monetary-, prospect-,
family-, and society-oriented);
Q1b: Predictions for comparison
between professional-level vs.
support-level STEM fields.
Likelihood of choosing support-level
life science vs. math-intensive fields:
society >family >prospect =monetary;
Likelihood of choosing support-level
life science vs. non-STEM fields:
society =family >monetary;
Likelihood of choosing non-STEM
vs. support-level math-intensive
fields: society >family
Q2 How does the relative
work value hierarchy
differ by gender and
influence gender
imbalance in STEM
choices above
academic task value?
H2a: Gender distribution in profiles:
Women are over-represented in family- and
society-oriented groups; Men were
over-represented in the Monetary-oriented profile
H2b: Mediation effects
Work value profiles were partially mediated
gender differences (favoring men) in entering
math-intensive vs. life science and non-STEM
Work value profiles or academic
and task values did not close the
gender gap in entering life
science vs. non-STEM fields.
found to be effective in triggering students’ interest and academic
performance in STEM topics (Lazowski and Hulleman, 2016).
Other interventions focusing on positive school experience
in relation to science, such as providing increased exposure
to women scientist role models and challenging stereotypes
of science masculinity, have been also proved to be useful
in promoting women’s motivation and engagement in STEM
activities (Wang and Degol, 2017).
Finally, given that both work values and academic task values
significantly contribute to gender differences in STEM fields, we
call for more interventions that target both types of value beliefs
and seek to enhance students’ perception of value beliefs attached
to activities and careers based on a long-term longitudinal design.
Importantly, these interventions should be implemented at early
ages since academic and work values are closely associated
with ability development, academic engagement, and STEM
educational and occupational preparedness (Eccles, 2009). On
the other hand, we think it is important to think carefully about
using our findings to socially engineer the next generations’
career choices. Ideally, individuals should be helped to make the
best career choices they can for themselves, influenced as little
as possible by stereotypes, gendered patterns of socialization,
and government policy. To say we would like women to be
just as likely as men to consider becoming an engineer or a
nurse or a medical doctor is one thing; to say we want to
persuade a particular women or man to become a computer
scientist rather than a journalist or professional musician is quite
The study was consistent with the ethical principles of human
subjects. First, we told the detailed content of the study to the
participants. Second, participants signed the informed consent
on a voluntary basis.
JG conceived of the study, performed the statistical analysis,
and wrote up the paper; JE conceived of the study, provided
suggestions for the statistical analysis, and commented the
manuscript; FS participated in the design of the study and
commented the manuscript; KS-A conceived of the study,
participated in the coordination of the study, and commented
the manuscript. All authors read and approved the final
This study was supported by the Academy of Finland grant
#273872, #308351, and #307598 to KS-A.
The Supplementary Material for this article can be found
online at:
Frontiers in Psychology | 13 July 2018 | Volume 9 | Article 1111
Guo et al. Work and Academic Values
Abele, A. E., and Spurk, D. (2011). The dual impact of gender and the influence of
timing of parenthood on men’s and women’s career development: longitudinal
findings. Int. J. Behav. Dev. 35, 225–232. doi: 10.1177/0165025411398181
Asparouhov, T., and Muthén, B. (2014). Auxiliary Variables in Mixture Modeling:
Using the BCH Method in Mplus to Estimate a Distal Outcome Model and an
Arbitrary Secondary Model. Mplus Web Notes.
Bardi, A., Buchanan, K. E., Goodwin, R., Slabu, L., and Robinson, M. (2014). Value
stability and change during self-chosen life transitions: self-selection versus
socialization effects. J. Pers. Soc. Psychol. 106, 131–147. doi: 10.1037/a0034818
Baron-Cohen, S. (2003). The Essential Difference: The Truth about the Male and
Female Brain. New York, NY: Basic Books.
Berings, D., De Fruyt, F., and Bouwen, R. (2004). Work values and personality
traits as predictors of enterprising and social vocational interests. Pers. Indiv.
Diff. 36, 349–364. doi: 10.1016/S0191-8869(03)00101-6
Beyer, S. (2014). Why are women underrepresented in computer science? Gender
differences in stereotypes, self-efficacy, values, and interests and predictors
of future CS course-taking and grades. Comp. Sci. Educ. 24, 153–192.
doi: 10.1080/08993408.2014.963363
Brown, E. R., and Diekman, A. B. (2010). What will I be? Exploring gender
differences in near and distant possible selves. Sex Roles 63, 568–579.
doi: 10.1007/s11199-010-9827-x
Ceci, S. J., and Williams, W. M. (2011). Understanding current causes of women’s
underrepresentation in science. Proc. Natl. Acad. Sci. U.S.A. 108, 3157–3162.
doi: 10.1073/pnas.1014871108
Chow, A., Eccles, J. S., and Salmela-Aro, K. (2012). Task value profiles across
subjects and aspirations to physical and IT-related sciences in the United States
and Finland. Dev. Psychol. 48, 1612–1628. doi: 10.1037/a0030194
Diekman, A. B., Brown, E. R., Johnston, A. M., and Clark, E. K. (2010). Seeking
congruity between goals and roles: a new look at why women opt out of science,
technology, engineering, and mathematics careers. Psychol. Sci. 21, 1051–1057.
doi: 10.1177/0956797610377342
Diekman, A. B., Clark, E. K., Johnston, A. M., Brown, E. R., and Steinberg, M.
(2011). Malleability in communal goals and beliefs influences attraction to stem
careers: evidence for a goal congruity perspective. J. Pers. Soc. Psychol. 101,
902–918. doi: 10.1037/a0025199
Diekman, A. B., Steinberg, M., Brown, E. R., Belanger, A. L., and Clark, E. K. (2017).
A goal congruity model of role entry, engagement, and exit: understanding
communal goal processes in stem gender gaps. Pers. Soc. Psychol. Rev. 21,
142–175. doi: 10.1177/1088868316642141
Diekman, A. B., Weisgram, E. S., and Belanger, A. L. (2015). New routes
to recruiting and retaining women in STEM: policy implications of a
communal goal congruity perspective. Soc. Issues Policy Rev. 9, 52–88.
doi: 10.1111/sipr.12010
Eccles, J. S. (2009). Who am I and what am I going to do with my life? Personal
and collective identities as motivators of action. Educ. Psychol. 44, 78–89.
doi: 10.1080/00461520902832368
Eccles, J. S., and Wang, M.-T. (2016). What motivates females and males to
pursue careers in mathematics and science? Int. J. Behav. Dev. 40, 100–106.
doi: 10.1177/0165025415616201
Eccles, J. S., Barber, B., and Jozefowicz, D. (1999). “Linking gender to education,
occupation, and recreational choices: applying the Eccles et al. model of
achievement-related choices,” in Sexism and Stereotypes in Modern Society: The
Gender Science of Janet Taylor Spence, eds W. B. Swann, J. H. Langlois, and L.
A. Gilbert (Washington, DC: American Psychological Association), 153–192.
Esping-Andersen, G. (2002). Why We Need a New Welfare State. Oxford: Oxford
University Press.
Ferriman, K., Lubinski, D., Benbow, C. P., Geary, D. C., Gottfredson, L. S., Gough,
H. G., et al. (2009). Work preferences, life values, and personal views of
top math/science graduate students and the profoundly gifted: developmental
changes and gender differences during emerging adulthood and parenthood. J.
Pers. Soc. Psychol. 97, 517–532. doi: 10.1037/a0016030
Frome, P. M., Alfeld, C. J., Eccles, J. S., and Barber, B. L. (2008). “Is the
desire for a family-flexible job keeping young women out of male-dominated
occupations?” in Explaining Gendered Occupational Outcomes: Examining
Individual and Social Explanations Through School and Beyond, eds H. M. G.
Watt and J. S. Eccles (Washington, DC: APA), 195–214.
Guo, J., Marsh, H. W., Parker, P. D., Morin, A. J. S., and Dicke, T. (2017). Extending
expectancy-value theory predictions of achievement and aspirations in science:
dimensional comparison processes and expectancy-by-value interactions.
Learn. Instruct. 49, 81–91. doi: 10.1016/j.learninstruc.2016.12.007
Guo, J., Parker, P. D., Marsh, H. W., and Morin, A. J. S. (2015). Achievement,
motivation, and educational choices: a longitudinal study of expectancy
and value using a multiplicative perspective. Dev. Psychol. 51, 1163–1176.
doi: 10.1037/a0039440
Hausmann, R., and Tyson, L. D. (2015). The Global Gender Gap Report 2015.
Geneva: World Economic Forum.
Hayes, A. R., Bigler, R. S., and Weisgram, E. S. (2018). Of men and money:
characteristics of occupations that affect the gender differentiation of children’s
occupational interests. Sex Roles 78, 775–788. doi: 10.1007/s11199-017-0846-8
Holland, J. L. (1997). Making Vocational Choices: A Theory of Vocational
Personalities and Work Environments, 3rd Edn. Odessa, FL: Psychological
Assessment Resources.
Huang, B., Sivaganesan, S., Succop, P., and Goodman, E. (2004). Statistical
assessment of mediational effects for logistic mediational models. Stat. Med.
23, 2713–2728. doi: 10.1002/sim.1847
Jamshidian, M., Jalal, S., and Jansen, C. (2014). MissMech: an R package for testing
homoscedasticity, multivariate normality, and missing completely at random
(MCAR). J. Stat. Softw. 56, 1–31. doi: 10.18637/jss.v056.i06
Jin, J., and Rounds, J. (2012). Stability and change in work values: a
meta-analysis of longitudinal studies. J. Vocat. Behav. 80, 326–339.
doi: 10.1016/j.jvb.2011.10.007
Johnson, M. K., Mortimer, J. T., Lee, J. C., and Stern, M. J. (2007).
Judgments about work dimensionality revisited. Work Occup. 34, 290–317.
doi: 10.1177/0730888407303182
Jones, G. M., Brader-Araje, L., Carboni, L. W., Carter, G., Rua, M. J.,
Banilower, E., et al. (2000). Tool time: Gender and students’ use
of tools, control, and authority. J. Res. Sci. Teach. 37, 760–783.
doi: 10.1002/1098-2736(200010)37:8<760::AID-TEA2>3.0.CO;2-V
Judge, T. A., and Bretz, R. D. (1992). Effects of work values on job choice decisions.
J. Appl. Psychol. 77:261.
Kimmel, L. G., Miller, J. D., and Eccles, J. S. (2012). Do the paths to
STEMM professions differ by gender? Peabody J. Educ. 87, 92–113.
doi: 10.1080/0161956X.2012.642276
Lazowski, R. A., and Hulleman, C. S. (2016). Motivation interventions
in education: a meta-analytic review. Rev. Educ. Res. 86, 602–640.
doi: 10.3102/0034654315617832
Lechner, C. M., Sortheix, F. M., Göllner, R., and Salmela-Aro, K. (2017). The
development of work values during the transition to adulthood: a two-country
study. J. Vocat. Behav. 99, 52–65. doi: 10.1016/j.jvb.2016.12.004
Lee, B., Lawson, K. M., and McHale, S. M. (2015). Longitudinal associations
between gender-typed skills and interests and their links to occupational
outcomes. J. Vocat. Behav. 88, 121–130. doi: 10.1016/j.jvb.2015.02.011
Lo, Y., Mendell, N., and Rubin, D. B. (2001). Testing the number of components
in a normal mixture. Biometrika 88, 767–778. doi: 10.1093/biomet/88.3.767
Lubke, G., and Muthén, B. O. (2007). Performance of factor mixture models as a
function of model size, covariate effects, and class-specific parameters. Struct.
Equ. Model. 14, 26–47. doi: 10.1080/10705510709336735
Lyons, S. T., Higgins, C. A., and Duxbury, L. (2010). Work values:
Development of a new three-dimensional structure based on confirmatory
smallest space analysis. J. Organ. Behav. 31, 969–1002. doi: 10.1002/
Mason, M. A. (2014). How to Level the Playing Field for Women in Science.The
Chronicle of Higher Education. Available online from:
article/How-to- Level-the-Playing- Field/145037/?cid=at&utm_source=at&
utm_medium=en, Accessed November 30, 2017.
Miller, D. I., Eagly, A. H., and Linn, M. C. (2015). Women’s representation in
science predicts national gender-science stereotypes: evidence from 66 nations.
J. Educ. Psychol. 107, 631–644. doi: 10.1037/edu0000005
Morin, A. J. S., and Marsh, H. W. (2015). Disentangling shape from level effects
in person-centered analyses: An illustration based on university teachers’
multidimensional profiles of effectiveness. Struct. Equ. Model. Multidiscipl. J.
22, 39–59. doi: 10.1080/10705511.2014.919825
MOW International Research Team (1987). The Meaning of Working. New York,
NY: Academic Press.
Frontiers in Psychology | 14 July 2018 | Volume 9 | Article 1111
Guo et al. Work and Academic Values
Muthén, B. O. (2001). “Latent variable mixture modeling,” in New Developments
and Techniques in Structural Equation Modeling, eds G. Marcoulides and R.
Schumacker (Mahwah, NJ: Erlbaum), 1–33.
Muthén, L. K., and Muthén, B. O. (1998-2012). Mplus User’s Guide, 7th Edn.
Los Angeles, CA: Authors.
OECD (2014). Education at a Glance 2014: OECD Indicators. Paris: OECD
OECD (2016). PISA 2015 Results (Volume I): Excellence and Equity in Education.
Paris: OECD Publishing.
Ros, M., Schwartz, S. H., and Surkiss, S. (1999). Basic individual values,
work values, and the meaning of work. Appl. Psychol. Int. Rev. 48, 49–71.
doi: 10.1111/j.1464-0597.1999.tb00048.x
Rubin, D. B. (1987). Multiple Imputation for Nonresponse in Surveys. New York,
NY: Wiley.
Sagie, A., and Elizur, D. (1996). The structure of personality values: a conical
representation of multiple life areas. J. Organ. Behav. 17, 573–586.
Sahdra, B. K., Ciarrochi, J., Parker, P. D., Basarkod, G., Bradshaw, E. L., and Baer,
R. (2017). Are people mindful in different ways? Disentangling the quantity
and quality of mindfulness in latent profiles and exploring their links to mental
health and life effectiveness. Eur. J. Pers. 31, 347–365. doi: 10.1002/per.2108
Sortheix, F. M., Chow, A., and Salmela-Aro, K. (2015). Work values and the
transition to work life: a longitudinal study. J. Vocat. Behav. 89, 162–171.
doi: 10.1016/j.jvb.2015.06.001
Su, R., and Rounds, J. (2015). All STEM fields are not created equal: people and
things interests explain gender disparities across STEM fields. Front. Psychol.
6:189. doi: 10.3389/fpsyg.2015.00189
Su, R., Rounds, J., and Armstrong, P. I. (2009). Men and things, women and people:
a meta-analysis of sex differences in interests. Psychol. Bull. 135, 859–884.
doi: 10.1037/a0017364
Super, D. E. (1962). The structure of work values in relation to status, achievement,
interests, and adjustment. J. Appl. Psychol. 46:231.
Super, D. E. (1990). “A life-span, life-space approach to career development,” in
Career Choice and Development: Applying Contemporary Theories to Practice,
eds D. Brown and L. Brooks & Associates (John Wiley & Sons).
Valla, J. M., and Ceci, S. J. (2014). Breadth-based models of women’s
underrepresentation in STEM fields: an integrative commentary on
Schmidt (2011) and Nye et al. (2012). Perspect. Psychol. Sci. 9, 219–224.
doi: 10.1177/1745691614522067
Vecchione, M., Caprara, G., Schoen, H., Castro, J. L. G., and Schwartz, S. H.
(2012). The role of personal values and basic traits in perceptions of the
consequences of immigration: a three-nation study. Brit. J. Psychol. 103,
359–377. doi: 10.1111/j.2044-8295.2011.02079.x
Wang, M., Degol, J., and Ye, F. (2015). Math achievement is important, but
task values are critical, too: examining the intellectual and motivational
factors leading to gender disparities in STEM careers. Front. Psychol. 6:36.
doi: 10.3389/fpsyg.2015.00036
Wang, M.-T., and Degol, J. (2013). Motivational pathways to STEM career choices:
using expectancy–value perspective to understand individual and gender
differences in STEM fields. Dev. Rev. 33, 304–340. doi: 10.1016/j.dr.2013.08.001
Wang, M.-T., and Degol, J. L. (2017). Gender gap in science, technology,
engineering, and mathematics (STEM): current knowledge, implications for
practice, policy, and future directions. Educ. Psychol. Rev. 29, 119–140.
doi: 10.1007/s10648-015-9355-x
Watt, H. M. G., and Richardson, P. W. (2007). Motivational factors influencing
teaching as a career choice: development and validation of the FIT-choice scale.
J. Exp. Educ. 75, 167–202. doi: 10.3200/JEXE.75.3.167-202
Weisgram, E. S., and Diekman, A. B. (2017). Making STEM “family friendly”:
the impact of perceiving science careers as family-compatible. Soc. Sci. 6:61.
doi: 10.3390/socsci6020061
Weisgram, E. S., Bigler, R. S., and Liben, L. S. (2010). Gender, values, and
occupational interests among children, adolescents, and adults. Child Dev. 81,
778–796. doi: 10.1111/j.1467-8624.2010.01433.x
Williams, W., and Ceci, S. (2012). When scientists choose motherhood. Am. Sci.
100:138. doi: 10.1511/2012.95.138
Woodcock, A., Graziano, W. G., Branch, S. E., Habashi, M. M., Ngambeki,
I., and Evangelou, D. (2013). Person and thing orientations: psychological
correlates and predictive utility. Soc. Psychol. Pers. Sci. 4, 116–123.
doi: 10.1177/1948550612444320
Conflict of Interest Statement: The authors declare that the research was
conducted in the absence of any commercial or financial relationships that could
be construed as a potential conflict of interest.
Copyright © 2018 Guo, Eccles, Sortheix and Salmela-Aro. This is an open-access
article distributed under the terms of the Creative Commons Attribution License (CC
BY). The use, distribution or reproduction in other forums is permitted, provided
the original author(s) and the copyright owner(s) are credited and that the original
publication in this journal is cited, in accordance with accepted academic practice.
No use, distribution or reproduction is permitted which does not comply with these
Frontiers in Psychology | 15 July 2018 | Volume 9 | Article 1111
... Five profiles were obtained, with four profiles distinguishable quantitatively (i.e., profiles with low to high levels for both work values) and one profile characterized by contrasting levels of work values (i.e., high intrinsic-low extrinsic). The second study (Guo et al., 2018) identified work values profiles in a sample of Finnish teenagers using a five-factor model of work values that included two social work values (i.e., working with others and contribution to society), two extrinsic work values (i.e., income and family), and one status work values (i.e., career advancement). This study identified four work values profiles characterized by: (1) strong endorsement of extrinsic-income and weak endorsement of social work values; (2) strong endorsement of status work value and low endorsement of extrinsic-family work value; (3) moderate level of extrinsic-family work values and low level of status, and (4) high levels of social work values and low level of extrinsic-income. ...
... In contrast, when individuals strongly endorse instrumentally-oriented work values (extrinsic, status), they report their needs for autonomy, competence, and relatedness as being more frustrated and less satisfied at work. Second, previous studies (Guo et al., 2018;Koh, 2016) showed that it is possible to identify unique profiles of work values to examine how subgroups of individuals display unique patterns of values combination (Meyer et al., 2013). Third, work values profiles have yet to be identified using the FFM-WV. ...
... The first objective was to identify work values profiles using latent profile analysis, based on the FFM-WV. We expected a small number of work values profiles (i.e., four or five), based on previous studies that found four or five profile solutions to be optimal even though they relied on different conceptualizations of work values (Guo et al., 2018;Koh, 2016). However, given that this study is only the third to apply a person-centered approach to work values and the first to do so with a fourfactor model, this hypothesis remain preliminary and not specifically tested. ...
Full-text available
The association between work values and key motivational variables has been repeatedly supported in previous studies. However, little attention has been devoted to understanding intraindividual patterns of work values and how combinations of work values relate to other motivational variables. This study aimed to identify profiles of work values based on a four-factor model (i.e., intrinsic, extrinsic, social, and status). It also investigated how profile membership relates to basic psychological need satisfaction and frustration at work using a self-determination perspective. A sample of French Canadian adults (N = 476) participated in this study by filling out an online questionnaire. Latent profile analyses revealed five distinct work values profiles. Results showed that participants in more positive profiles (i.e., high level of intrinsic, social, and status work values) generally reported higher level of need satisfaction and lower level of need frustration at work than participants belonging to more negative profiles (i.e., low level of intrinsic, social, and status work values). These results support the importance of considering work values in organizational and career development interventions, and to do so using a person-centered approach, to better understand need satisfaction and frustration at work.
... The identification of factors influencing educational choices that in turn lead to gender disparity in STEM disciplines are recently studied. Several studies provide different possible explanations, such as relative strengths -girls being better in other disciplines compared to boys (OECD, 2019); lower confidence or poorer perception of abilities in some STEM subjects of girls (OECD, 2019); role-model function of female STEM teachers (Bottia et al., 2015); influence of society and family through gender stereotypes (Makarova et al., 2019) or gender role beliefs (Dicke et al., 2019); expectations to what type of jobs different subjects can contribute, where girls have the tendency to favour people or society-oriented work values, which are not perceived to be coupled with STEM discipline (Guo et al., 2018); or monetary expectation in later stage of career, which on contrary are often perceived to be coupled with STEM subjects and seem to be more important for boys (Rapport andThibout, 2018, Guo et al, 2018). ...
... Concrete examples include missing or misleading knowledge about space careers and the prerequisites for choosing such a career path. Another mentioned misconception is that is not clear to everybody that STEM studies can play an important role in the contribution to solve societal problems, which is related to the findings that working for others is one factor for girls choosing a specific educational path (Guo et al., 2018). In line with Fitzsimmons (2018) quoting of the need to "urge industries traditionally dominated by one gender to send diverse role models to schools to talk about careers" 6 , we argue that increased publicity about career opportunities when choosing STEM education related to the EO*GI sector in society and education can motivate girls and young women to join. ...
Full-text available
Gender inequality is omnipresent in our society and in the field of education and training, the gender gap is especially evident in STEM (Science, Technology, Engineering and Mathematics) disciplines. While different studies have been conducted about potential reasons explaining this gap, little is known about gender inequality and underlying factors in the Earth Observation and Geoinformatics (EO*GI) domain. To close some parts of this knowledge gap, the initiative Women in Copernicus was established with the overall goal to make women working in the EO*GI field and especially in the Copernicus ecosystem more visible. This paper analyses the results of a survey of 462 women identifying reasons for not choosing STEM education and the barriers related to educational choices in their career path. The main obstacles that hinder choosing a STEM education for these women are stereotypes in society, missing female role models but also culture, television and society message transmitted by the media. The lack of self-confidence is an essential factor in this choice and is also experienced as a barrier during individual career paths. This analysis provides insights valuable for political decisions making targeting at a gender-balanced work environment and emphasizes the importance of attracting more girls and young women towards a STEM education and supporting them during their career to reach skills and occupational equality and strengthen the economic development of the EO*GI sector.
... At the same time, there is an erratic belief that men are attracted to more technical and rational professions of building and producing things. However, this justification is reductionist and binary (Diekman et al., 2010;Guo et al., 2018;Sikora & Pokropek, 2011;Su & Rounds, 2015). It seems that guilds are to be divided into two simplistic categories without considering the presence and importance of the environment. ...
Full-text available
Differences in the representation of diversity in higher education, emphasising the gender gap in some areas, are issues addressed from different research domains. Socially, gender roles have been constructed and are also related to professions. In this context, the Social Cognitive Career Theory explores the possible causes of segregation. This segregation is evident in Europe and Spain, as indicated by the European Institute for Gender Equality. This paper describes the design and validation process of an instrument to find out what opinions university students have about higher education studies in science, technology, engineering and mathematics (STEM), according to gender. After drafting the questionnaire, it was piloted in a non-experimental quantitative design in Spain. Subsequently, a validity and reliability study was applied to validate the items and construct their dimensionality. The process was implemented using Reliability Analysis and Exploratory Factor Analysis. Also, the dimensionality consists of five scales: Gender Ideology, Perception and Self-perception, Expectations about Science, Attitudes and Interests. Based on the results, it is concluded that the opinion about STEM studies is conditioned by personal elements, such as motivations, educational background and family and social influences, such as people who judged their decision, were their references or studied STEM programs. Finally, it is essential to pay socio-educational attention to the modulating components of decisions about which higher education studies to pursue. Awareness of the factors involved in the decision helps the educational community to establish mechanisms to prevent horizontal gender segregation. The instrument designed, validated and presented in this study provides a glimpse of possible causes for the gender gap in STEM higher education.
... Earlier studies suggest that men place more value on jobs that yield high incomes, power, and prestige compared to women (Abele & Spurk, 2011;Guo et al., 2018), which seems to impact negatively on women's decisions to enter STEM fields, especially in math-intensive fields (Diekman et al., 2010(Diekman et al., , 2015. The usual reasoning behind the gender gap in income motivation is that men are socialized into assuming the role of family breadwinner, meaning it is more important for them to earn high incomes. ...
Full-text available
Gender segregation in fields of study represents an important explanation for gender inequalities in the labor market, such as the gender wage gap. Research shows that horizontal gender segregation in higher education persists for a variety of reasons, including women’s greater communal goals and men’s greater motivation to earn high incomes. Yet with the male breadwinner model in decline, a key question is whether women’s motivation to earn high incomes might contribute to increasing women’s participation in female-atypical fields of study. Using data from the German Student Survey over a period of 30 years, our findings show that the proportion of women enrolled in female-atypical fields of study increased from 1984 to 2015. Moreover, women’s motivation to earn high incomes mediates the effect of time on enrollment in female-atypical fields of study. Their motivation to earn high incomes might thus be a factor contributing to the disruption of gender segregation in fields of study over time. Furthermore, contrary to expectations, the motivation to earn high incomes as a driving force for women to opt for gender-atypical fields of study is not stratified by social background.
... Males (both boys and adult men) were found to be more interested than females in jobs offering high monetary rewards (Hayes et al., 2018;Konrad et al., 2000;Weisgram et al., 2010); men also were more interested in leadership roles, which are traditionally deemed congruent with men's social roles (Neff et al., 2007) and incongruent with women's roles (Eagly & Karau, 2002). Interestingly, studies found that material goal endorsement was positively associated with career interests in mathematics-intensive fields (Eccles, 1999;Guo et al., 2018), which may partly explain the underrepresentation of women in math-intensive fields (Diekman et al., 2015). ...
Full-text available
other members of our team for their dedicated work. We also thank all the participating English professors and their students for making this work possible. Abstract Two national datasets of first-year college students, collected a decade apart, asking the same questions about career interests and life goal endorsement, allowed us to investigate the extent to which the life goals and career interests had converged among young men and women. We compared the gender differences in four types of goal endorsement (communal, material, intellectual, and free-time goals) by career interest groups (science, engineering, medicine, health, and other professions) between the two cohorts (2007 vs. 2017). Conversely, we compared the gender differences in career interests by goal endorsement between the two cohorts. Our specific focus was on science, technology, engineering, and mathematics (STEM) career interests. We found that significant differences have stubbornly persisted between male and female students preparing for STEM careers, particularly in the area of communal goals, whereas gender differences in communal, material, and intellectual goals have narrowed or disappeared for those interested in many non-STEM careers.
... We thus suggest that a 'configural', person-centered analysis can complement and further inform what is already known about the links between aspirations and well-beingrelated outcomes. Such strategies are gaining popularity in empirical psychology and have been successfully employed in person-centered analyses of several constructs, such as, achievement goal orientation (Pastor, Barron, Miller, & Davis, 2007;Tuominen-Soini, Salmela-Aro, & Niemivirta, 2008), work values (Guo, Eccles, Sortheix, & Salmela-Aro, 2018), mindfulness (Bravo, Boothe, & Pearson, 2016;Pearson, Lawless, Brown, & Bravo, 2015;Sahdra et al., 2017), self-concept (Marsh, Lüdtke, Trautwein, & Morin, 2009), and personality traits (Merz & Roesch, 2011). ...
We conducted a person-centered analysis of the Aspiration Index to identify subgroups that differ in the levels of their specific (wealth, fame and image, personal growth, relationships, community giving, and health) and global intrinsic and extrinsic aspirations. In a Hungarian (N = 3,370; 77% female; age: M = 23.57), an Australian (N = 1,632; 51% female; age: M = 16.6), and an American sample (N = 6,063; 82.2% female; age: M = 21.86), we conducted separate bifactor exploratory structural equation models to disentangle the level of higher-order intrinsic and extrinsic aspirations from the shape of specific aspirations by using the resultant factor scores as indicators in latent profile analyses. The analyses yielded 3 replicable latent profiles: Disengaged from relationships and health (Profile 1); Aspiring for interpersonal relationships more than community relationships (Profile 2); and Aspiring for community relationships more than interpersonal relationships (Profile 3), with Profile 3 reliably experiencing the highest well-being. To demonstrate the incremental value of our approach to more traditional variable-centered methods, we used profile membership to predict well-being while controlling for the aspirations that comprise the profiles. Even in these conservative tests, profile membership explained additional variance in well-being. However, the real-life significance of the size of the incremental value appeared quite small. These studies make a unique contribution to the literature by identifying replicable latent profiles of aspiring, membership to which uniquely predicted well-being, over and above the constituent variables. (PsycInfo Database Record (c) 2021 APA, all rights reserved).
... These theoretical distinctions are reflected in the measures of students' motivational beliefs, which either focus on more stable, domainspecific motivation (traits) (e.g. Chow & Salmela-Aro, 2011;Gaspard, Wille, Wormington, & Hulleman, 2019;Guo, Eccles, Sortheix, & Salmela-Aro, 2018;Lazarides et al., 2020Lazarides et al., , 2016Nurmi & Aunola, 2005;Viljaranta, Aunola, & Hirvonen, 2016) or on situation-specific motivation that varies from moment to moment (states) (e.g. Dietrich, Moeller, Guo, Viljaranta, & Kracke, 2019;Gray, 2014;Tsai, Kunter, Lüdtke, Trautwein, & Ryan, 2008). ...
The present study employed a person-oriented approach to (1) identify elementary school students’ self-concept and intrinsic value profiles across the subjects Finnish language, mathematics and science and to examine (2) the stability and change of these motivational profiles from 2nd to 3rd grade, (3) gender differences in profile membership as well as (4) the relation to students’ STEM (Science, Technology, Engineering and Mathematics) occupational aspirations. Based on data from 383 Finnish students (56.7% girls) three profiles were identified: High motivation across all three subjects, low motivation across all subjects, and a math-motivated profile with low motivation in the other two subjects. Latent transition analyses revealed moderate stability, particularly in the high motivation profile. Girls were less likely to be and to remain in the math-motivated profile, but they were more likely than boys to remain in the high motivation profile. The math-motivated profile transition pattern was associated with students’ STEM occupational aspirations. Fulltext:,8E164KY0
Purpose Work values are a representation of people’s priorities as they reflect what is pertinent for them and what they want to accomplish. In light of this, the purpose of this study is to understand the priorities given to work values (extrinsic and intrinsic) by employees and also to explore whether these work values vary with the levels of work engagement and job burnout. Design/methodology/approach The study was based on the survey responses of 386 officers working in Indian manufacturing organisations engaged in different areas. Findings The findings reveal that security officers give much priority to extrinsic work values than intrinsic work values (IWVs). Moreover, IWVs vary with different levels of work engagement along with job burnout. The security officers belonging to the engaged group differ significantly with those belonging to the job burnout group in terms of IWVs. Moreover, work values also have a negative correlation with job burnout and a positive correlation with work engagement. Originality/value This study explores the variation in work values of security officers working in Indian manufacturing organisations with changes in levels of job burnout and work engagement, which is a novel contribution in the field. The findings also advocate that it is crucial for human resource managers, supervisors and key people in organisations to find out employees showing early signs of job burnout (exhaustion or disengagement) or early stages of strain and frustration as the priorities of work values of the employees are affected by these parameters. Such identified employees should be provided with required managerial support and necessary work resources immediately.
Purpose The influence of gender on high-tech entrepreneurship is of growing interest worldwide, as scholars argue that women face gendered barriers specific to this field. Although some gender-focussed research exists on the interplay of context and entrepreneurial learning, these issues have yet to be intensively studied, and the research aims to address this gap. Design/methodology/approach The research draws upon empirical evidence from the entrepreneurial learning of nine women opportunity entrepreneurs in the high-technology sector in Norway. It employs a qualitative phenomenological approach, with retrospective and in-depth interviews to capture and analyse the entrepreneurs' lived experiences and learning histories. Findings The entrepreneurs in this study highlight gendered learning experiences, leading them to make conscious and strategic decisions of both alignment and resistance to negotiate their enterprise in a highly masculine sector. Their prior learning histories of not belonging seem to underpin their preparedness for entrepreneurship in the sector. Counter to prevailing theorizing, not belonging is an enabling condition, allowing women entrepreneurs to subvert and challenge a highly masculinized context. This condition empowers them to mobilize their “otherness” to create change within their own ventures and make the rules on their own terms. Originality/value This interdisciplinary research deepens the understanding of the interplay between gender, entrepreneurial learning and context through the concept of belonging and extends theorization of the gendered dynamics in entrepreneurial learning histories. The paper proposes a framework of gendered entrepreneurial learning in a masculinized industry context, which highlights important implications for future gender and entrepreneurial learning research.
For some professionally, vocationally, or technically oriented careers, curricula delivered in higher education establishments may focus on teaching material related to a single discipline. By contrast, multidisciplinary, interdisciplinary, and transdisciplinary teaching (MITT) results in improved affective and cognitive learning and critical thinking, offering learners/students the opportunity to obtain a broad general knowledge base. Chemistry is a discipline that sits at the interface of science, technology, engineering, mathematics, and medicine (STEMM) subjects (and those aligned with or informed by STEMM subjects). This article discusses the significant potential of inclusion of chemistry in MITT activities in higher education and the real-world importance in personal, organizational, national, and global contexts. It outlines the development and implementation challenges attributed to legacy higher education infrastructures (that call for creative visionary leadership with strong and supportive management and administrative functions), and curriculum design that ensures inclusivity and collaboration and is pitched and balanced appropriately. It concludes with future possibilities, notably highlighting that chemistry, as a discipline, underpins industries that have multibillion dollar turnovers and employ millions of people across the world.
Full-text available
Occupational interests become gender differentiated during childhood and remain so among adults. Two characteristics of occupations may contribute to this differentiation: the gender of individuals who typically perform the occupation (workers’ gender) and the particular goals that the occupation allows one to fulfill, such as the opportunity to help others or acquire power (value affordances). Two studies tested hypotheses about whether U.S. 6- to 11-year-olds show gender differences in their interest in novel jobs that were depicted as (a) being performed by men versus women and (b) affording money, power, family, or helping values. In Study 1, 98 children rank-ordered their preferences for experimentally-manipulated novel jobs, and they answered questions about their occupational values and the value affordances of jobs in which men and women typically work. In Study 2, a second sample of 65 children was used to test the replicability of findings from Study 1. As hypothesized, children were more interested in jobs depicted with same- than other-gender workers in both studies. Boys showed greater interest than did girls in novel jobs depicted as affording money in Study 1, but not Study 2. Explicit knowledge that men and women typically work in jobs that afford differing values increased with participants’ age.
Full-text available
Two studies extended the communal goal congruity perspective to examine perceived incongruity between science careers and family caregiving goals. Study 1 examined beliefs about science careers among young adolescents, older adolescents, and young adults. Science careers were perceived as unlikely to afford family goals, and this belief emerged more strongly with age cohort. Study 1 also documented that the perception that science affords family goals predicts interest in pursuing science. Study 2 then employed an experimental methodology to investigate the impact of framing a science career as integrated with family life or not. For family-oriented women, the family-friendly framing of science produced greater personal favorability toward pursuing a science career. In addition, perceived fulfilment of the scientist described predicted personal favorability toward a science career path. We discuss the implications of these findings for research and for policy.
Full-text available
We sought to disambiguate the quantitative and qualitative components of mindfulness profiles, examine whether including 'nonattachment' as a subcomponent of mindfulness alters the profiles, and evaluate the extent to which the person-centred approach to understanding mindfulness adds predictive power beyond a more parsimonious variable-centred approach. Using data from a nationally representative sample of Americans (N = 7884; 52% female; Age: M = 47.9, SD = 16), we utilized bifactor exploratory structural equation modelling and latent profile analysis to separate the level and shape of previously identified profiles of mindfulness (Pearson, Lawless, Brown, & Bravo, ). Consistent with past research, we identified a judgmentally observing profile and a non-judgmentally aware group, but inconsistent with past research, we did not find profiles that showed high or low levels on all specific aspects of mindfulness. Adding nonattachment did not alter the shape of the profiles. Profile membership was meaningfully related to demographic variables. In models testing the distinctive predictive utility of the profiles, the judgmentally observing profile, compared to the other profiles, showed the highest levels of mental ill-health, but also the highest levels of life satisfaction and effectiveness. We discuss the implications of our study for clinical interventions and understanding the varieties of mindfulness.
Full-text available
This study addresses the development of work values—the desired characteristics of one’s current or future job—during young adulthood. Using two panel studies from Germany (N = 2,506) and Finland (N = 1,326), we investigated (a) mean-level and rank-order change and stability in work values across three biennial waves (age 20/21 to age 25/26); and (b) the influence of stable background characteristics as well as of major transitions in family and work roles on inter-individual differences and intra-individual changes in work values. Latent measurement models with three work value dimensions showed good fit in both countries: extrinsic (importance of job security and material rewards), intrinsic (importance of having an interesting, varied, and valuable job), and autonomy (working independently; making one’s own decisions). Analyses revealed high mean-level stabilities and moderate to high rank-order consistencies in work values across four years. Intrinsic work values emerged as the most highly endorsed value, and extrinsic work values as the most stable value, in both countries. Individual differences in work values emerged along the lines of sociodemographic background characteristics—especially gender, and to a lesser extent, parental socio-economic status (SES) and school track— whereas work and family transitions played only a limited role in explaining changes in work values across time. We discuss these results against the backdrop of previous research conducted mainly in North America and note some implications for work value research and for career counseling.
Full-text available
Although the gender gap in math course-taking and performance has narrowed in recent decades, females continue to be underrepresented in math-intensive fields of Science, Technology, Engineering, and Mathematics (STEM). Career pathways encompass the ability to pursue a career as well as the motivation to employ that ability. Individual differences in cognitive capacity and motivation are also influenced by broader sociocultural factors. After reviewing research from the fields of psychology, sociology, economics, and education over the past 30 years, we summarize six explanations for US women’s underrepresentation in math-intensive STEM fields: (a) cognitive ability, (b) relative cognitive strengths, (c) occupational interests or preferences, (d) lifestyle values or work-family balance preferences, (e) field-specific ability beliefs, and (f) gender-related stereotypes and biases. We then describe the potential biological and sociocultural explanations for observed gender differences on cognitive and motivational factors and demonstrate the developmental period(s) during which each factor becomes most relevant. We then propose evidence-based recommendations for policy and practice to improve STEM diversity and recommendations for future research directions.
The goal congruity perspective provides a theoretical framework to understand how motivational processes influence and are influenced by social roles. In particular, we invoke this framework to understand communal goal processes as proximal motivators of decisions to engage in science, technology, engineering, and mathematics (STEM). STEM fields are not perceived as affording communal opportunities to work with or help others, and understanding these perceived goal affordances can inform knowledge about differences between (a) STEM and other career pathways and (b) women's and men's choices. We review the patterning of gender disparities in STEM that leads to a focus on communal goal congruity (Part I), provide evidence for the foundational logic of the perspective (Part II), and explore the implications for research and policy (Part III). Understanding and transmitting the opportunities for communal goal pursuit within STEM can reap widespread benefits for broadening and deepening participation.