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

net/publication/393984357

Exploring the Links Between Gender-(A)Typical Career Aspirations and

Educational Choices: Heterogeneous Developmental Pathways

Article in Developmental Psychology · July 2025

DOI: 10.1037/dev0002015

CITATIONS READS

0 28

2 authors:

Jeffrey M. DeVries Fani Lauermann

University of California, Irvine University of Bonn
27 PUBLICATIONS 454 CITATIONS 56 PUBLICATIONS 2,398 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Jeffrey M. DeVries on 25 July 2025.

The user has requested enhancement of the downloaded file.

GENDERED CAREER ASPIRATIONS 1

This article was accepted for publication in Developmental Psychology on May 9, 2025.

©American Psychological Association, 2025. This paper is not the copy of record and may

not exactly replicate the authoritative document published in the APA journal. The final

article is available, upon publication, at: http://dx.doi.org/10.1037/dev0002015

Exploring the Links between Gender-(A)typical Career Aspirations and Educational

Choices: Heterogeneous Developmental Pathways

Jeffrey M. DeVries1 & Fani Lauermann2

1
University of California Irvine
2
University of Bonn
GENDERED CAREER ASPIRATIONS 2

Abstract

Developmental psychologists often focus on categories with significant gender

differences, such as science, technology, engineering, and math (STEM), to understand the

causes of persistent gender-typical educational and career choices. However, the proportion

of men and women can vary substantially within the same category (e.g., STEM).

Accordingly, this study examined heterogeneous developmental trajectories of adolescents’

career preferences using data from a representative sample of German academic-track 9th-to-

12th-grade students (N = 4,759, 56% female). We linked adolescents’ career preferences with

census data on the proportion of male/female job incumbents in each preferred occupation.

Over half of the participants (52%) aspired to a gender-segregated career by 12th grade.

Growth mixture analyses revealed five distinct developmental patterns: stable preferences for

male-dominated (21%, n = 920), female-dominated (22%, n = 983), or gender-neutral careers

(48%, n = 2,276), and shifts from male-to-female-dominated (6%, n = 280) or neutral-to-

male-dominated (6%, n = 300) career aspirations. These patterns predicted differences in

mean level and growth of subject-specific academic beliefs, the gender ratios of subsequent

choices of advanced math or language arts classes, and gender-typical university majors up to

four years after high school. Aspiring to more male-dominated careers was linked to positive

academic development in the math domain; aspiring to more female-dominated careers was

related to positive development in the verbal domain. Some developmental processes were

gender-specific, as boys aspiring to neutral or female-dominated careers experienced more

positive development in the verbal domain than girls. Highly-performing girls in math tended

to change to male-dominated careers later in school, but highly-performing boys in math

were on a stable-male trajectory throughout high school.

Keywords. gender; career aspirations; college major choices; self-concept of ability;

achievement
GENDERED CAREER ASPIRATIONS 3

Exploring the Links between Gender-(A)typical Career Aspirations and Educational Choices:

Heterogeneous Developmental Pathways

Gender disparities in many occupational fields remain stark and can lead to economic

and societal inequalities (Federal Statistics Office of Germany, 2018; Miyamoto & Wicht,

2020; U.S. Census Bureau, 2019). Women are often underrepresented in math-intensive

fields, such as the physical sciences, technology, engineering, and mathematics (STEM), and

overrepresented in other domains, including some areas within the life and social sciences

(Parker et al., 2020; Parker et al., 2018). To understand why, much developmental research

has examined gender differences in the development of relevant academic skills (e.g., in math

and verbal domains; Breda & Napp, 2019; S. Wan et al., 2023; Wang et al., 2013),

achievement-related motivational beliefs (e.g., domain-specific self-concepts of ability;

Chow et al., 2012; Z. H. Wan, 2021 ), and gender-typical career aspirations, for instance, in

math-intensive or human services-related fields (e.g. Eccles & Wigfield, 2020; Gottfredson &

Lapan, 1997; Lauermann et al., 2017). This evidence suggests that gender disparities in

adolescents’ achievement, academic beliefs, and career aspirations correspond to later

disparities in educational choices and career attainment (e.g., math-intensive vs. life sciences-

related college majors and occupations; Lauermann et al., 2017; Parker et al., 2012).

However, this research has not directly examined the gender typicality of career

aspirations, as indicated by the ratio of female/male job incumbents; instead, it focuses on

proximate variables such as math- or STEM-related careers. Focusing on such broad

occupational categories can overlook important heterogeneity in the proportion of men and

women within these categories (e.g., in different STEM fields; Parker et al., 2020). Moreover,

possible interindividual differences in the development of men’s and women’s career

aspirations are often overlooked. To provide a detailed account of the role gender plays in the

formation of career aspirations and attainment, we examine assumptions grounded in three

GENDERED CAREER ASPIRATIONS 4

interrelated developmental theories and assess the gender-typicality of career aspirations

directly via the proportion of men/women in a given occupation.

First, according to Gottfredson’s (1996) theory of circumscription and compromise,

gender constrains the number of career options perceived as acceptable by younger children

and adolescents. Children tend to eliminate career options based on whether these options are

perceived as atypical for their gender. Notably, this circumscription process unfolds before

children can fully explore their career interests and individual strengths. Second, situated

expectancy-value theory (Eccles, 2009; Eccles & Wigfield, 2020) conceptualizes career

aspirations as achievement-related choices predicted by individuals’ subjective expectancies

of success and their valuing of a given subject or career path relative to alternative options.

Gender differences in students’ expectancy of success and task values (including the

anticipated costs entailed in making a given gender-(a)typical choice) can thus result in

gender differences in career preferences and choices. Finally, the differential effects model

(Parker et al., 2012) combines these complementary predictions. Within this model, gender

may both directly and indirectly affect career aspirations. Gender may directly influence what

career options are considered, as is the case for gender-based circumscription in Gottredson’s

theory. Gender may also indirectly affect individuals’ career aspirations through

psychological mediators such as students’ success expectancies and valuing of academic

domains (including perceived costs), as proposed in Eccles and colleagues’ SEVT.

These three frameworks provide interrelated accounts of why individuals may aspire

to jobs that are typical or atypical of their gender. However, studies rarely look directly at the

“gender typicality” of adolescents’ career aspirations and career-related choices,

operationalized, for instance, based on objective indicators such as the proportion of women

in a career. Instead, most research has focused on aspirations within specific career categories

(e.g., STEM jobs; Sahin et al., 2018; Wang et al., 2013; Wille et al., 2020) and the
GENDERED CAREER ASPIRATIONS 5

developmental trajectory of childhood and adolescent career aspirations, university majors,

and career attainment within those categories (e.g., Lauermann et al., 2017; Parker et al.,

2014). As a result, we know more about gender differences in aspirations to specific career

categories and little about the nature and development of adolescents’ preferences for male-

versus female-dominated fields. Furthermore, there is a lack of agreement on suitable career

categories for studying gender differences, necessitating numerous categorization schemes,

such as differentiating between physical and biological sciences (e.g., Parker et al., 2014) and

focusing on subfields within STEM that either include or exclude the medical field and the

social sciences (e.g., see Parker et al., 2020). Evidence focusing on career categories may

overlook important heterogeneity, as men and women often make vastly different choices

within the same category (e.g., STEM; Parker et al., 2020). For instance, women earn nearly

half of all biomedical and environmental engineering degrees but barely an eighth of those in

mechanical or electrical engineering (Cheryan et al., 2017; Yoder, 2014).

Finally, the limited research that does look at gender typicality in career aspirations

tends to focus on mean-level differences. For instance, Basler et al. (2021) identify different

average trajectories of gender typicality across different academic tracks, and Lawson et al.

(2018) demonstrate that the trajectories of gender-typical career aspirations differ for boys

and girls based on categorizing the child’s personality features as either gender-typical or

gender-atypical. This evidence generally shows that gender does indeed tend to constrain

individuals’ career choices, as most boys and girls are deterred from pursuing gender-atypical

careers. Thus, we have some evidence concerning broad, group-based developmental patterns

of gender-based career aspirations. However, we know little about potential heterogeneity in

those patterns, possible implications for students’ academic development, and students who

differ significantly from the average. Such students may pursue jobs that are atypical of their
GENDERED CAREER ASPIRATIONS 6

gender, and understanding their motivations and development is crucial to understanding the

overall role of gender typicality in career aspirations.

The present study builds on prior work on career aspirations and career-related

choices by focusing on gender as an individual background characteristic and a feature of

occupational choices. We examine the effects of gender by using the gender typicality of a

career as an outcome variable and not a covariate. We use a nationally representative

longitudinal sample of German adolescents and link adolescents’ self-reported career

aspirations in grades 9 through 12 with concurrent census data regarding the proportion of

women in each occupation. These data allow us to examine the gender typicality of

adolescents’ career preferences in their cultural/national context, how these gendered

preferences may change during adolescence, and whether there are individual differences in

developmental trajectories. We also explore associations with theorized antecedents such as

domain-specific academic beliefs (math and verbal self-concepts of ability), standardized

achievement, and subsequent university study program selection (i.e., university major).

Thus, we strive to understand which academic and background characteristics (e.g.,

achievement and achievement-related beliefs) are predictive of gender-typical career

aspirations, whether gender typicality is a relatively stable or a variable factor of adolescents’

career preferences, and whether gender typicality has long-term predictive effects on career-

related choices such as university major selection.

Circumscription and Compromise Shaping Gendered Career Choices

The theory of circumscription and compromise (TCC; Gottfredson, 1981, 1996;

Gottfredson & Lapan, 1997) is a developmental model of career aspirations with four stages:

(1) orientation, (2) circumscription based on gender, (3) circumscription based on

socioeconomic status (SES), and (4) compromise. After an initial orientation stage (around

ages 3-5), when children form career aspirations based on their perceptions of size, power,
GENDERED CAREER ASPIRATIONS 7

and adult roles, children enter a circumscription stage during childhood (around ages 6-8). In

this stage, younger children exclude (circumscribe) careers based on perceptions of gender

typicality. For instance, many girls may believe that most scientists are men: when asked to

draw a “scientist,” young girls are likelier to draw a man than a woman, suggesting gender-

stereotypical beliefs about science (Jones & Hite, 2021). As a result, they might exclude

science-related careers from their aspirations before fully exploring their abilities, self-

concepts, and interests. In late childhood, children enter the third stage (around ages 9-13)

and integrate awareness of SES into their decision-making. This can lead them to constrain

their choices to options viewed as attainable and desirable for someone with their social

status. Lastly, in the fourth stage (adolescence), adolescents seek to reconcile their career

aspirations with the realities of their personal and social contexts. In this compromise stage,

adolescents are not just weighing career options based on gender and social status but also on

how these options fit with their developing self-concept. Adolescents are also influenced by

external factors such as family expectations, cultural norms, and economic conditions. They

might find themselves considering or even choosing careers that are atypical for their gender

as they negotiate between their interests, abilities, and the practical aspects of job availability

and educational attainment. Adolescents’ maturing cognitive abilities enable a more nuanced

understanding of the interplay between personal desires and external constraints.

The type and degree of compromise are critical in shaping adolescents’ career

aspirations (Gottfredson & Lapan, 1997). Choosing a career atypical for one’s gender, SES,

and relevant abilities requires a greater compromise, which may not always be achievable.

For instance, students are likelier to drop out of vocational training programs perceived as

more gender atypical, especially during their first year of training (Beckmann et al., 2023).

Moreover, compromises on each dimension can vary in magnitude. Gottfredson (1996)

argued that pursuing a gender-atypical career is less likely than following one that is atypical
GENDERED CAREER ASPIRATIONS 8

based on other aspects, such as SES. Several studies show that girls are more likely to

compromise than boys, thus pointing to notable between-group heterogeneity (Dinella et al.,

2014; Miyamoto & Wicht, 2020). Compromise rates also vary based on masculine or

feminine personality traits and gendered skills, such as students’ math and verbal skills

(Basler et al., 2021; Lawson et al., 2018). However, studies directly examining individual

differences in the formation of gender-(a)typical career aspirations and their associations with

the development of adolescents’ academic skills are still lacking.

The Role of Gender in Situated Expectancy-Value Theory

According to SEVT (Eccles & Wigfield, 2020), individuals’ achievement-related

choices, including their educational and career choices, are shaped by two proximal

predictors: their expectations of success and the subjective valuing of different options.

Expectancies refer to an individual’s self-concept of ability and confidence in succeeding at a

task, while task values reflect the task’s perceived importance, usefulness, enjoyment, and

anticipated costs 1. For instance, choosing to pursue a math-intensive (or reading-intensive)

major and occupation is positively predicted by students’ ability self-concept and expected

success, the perceived utility, and intrinsic interest in math-related (or reading-related) tasks,

and negatively by the anticipated costs (e.g., effort- and stress-related costs; e.g., Lauermann

et al., 2015; Lauermann et al., 2017; Wolter et al., 2019; Parker et al., 2020). Gender

differences in adolescents’ expectancies and values can thus lead to corresponding gender

differences in educational and occupational choices, such as taking more advanced math and

verbal courses (C. Brown & Putwain, 2021; Gaspard et al., 2020; Kirkham & Chapman,

2021; Mueller & Winsor, 2016) or enrolling in more math-intensive majors (Lauermann et

al., 2015; Lauermann et al., 2017). Substantial evidence has examined such differences,

1
Similar to some prior research (e.g., Eccles et al. , 1993; Gaspard et al., 2020; Lauermann et al., 2017) we do
not differentiate between self-concept of ability and expected future success in this study due to the typically
high correlation of these two constructs.
GENDERED CAREER ASPIRATIONS 9

showing that the effects of gender on future educational and career choices are often

mediated via students’ expectancies and the subjective valuing of academic fields such as

math, science, or language arts, whereas moderation analyses by gender typically produce

null results (e.g., Guo et al., 2015; Guo et al., 2017). Thus, similar psychological processes

seem to drive boys’ and girls’ educational and career choices, and gender-specific

achievement-related choices have been linked to mean-level differences in adolescents’

expectancies and values (Lauermann et al., 2017; Parker et al., 2014; Parker et al., 2020;

Parker et al., 2018).

However, other aspects of the theory, such as the possible sources of gender-atypical

educational and career choices, have received comparatively less attention (for important

exceptions focusing on parents, see Kang et al., 2019; Ramos Carranza & Simpkins, 2021;

Starr et al., 2022). In SEVT, gender stereotypes and role model effects can contribute to the

underrepresentation of women in math and science (Eccles, 1983, 2009) due to the

“importance of adult behavior as a standard or model for children’s behavior” (Eccles, 1983,

p. 116). Research has shown that the underrepresentation of female role models in the math

domain may be a critical reason for the underrepresentation of female students in math

courses, majors, and careers (e.g., Eccles, 1983; Morgenroth et al., 2015). Relevant research

has been conducted on both the success expectancy and value components of SEVT.

Attainable role models can boost expectations of success in career training and aspiration,

thus boosting motivation to continue in studies or career training (Gartzia et al., 2021). Role

models have also been shown to boost interest in STEM careers in multiple ways. They affect

interests and awareness of career options (Zhang et al., 2023), and same-gender role models

can boost achievement in male-dominated subjects like math (e.g., female role models who

achieved the top math test scores boosted later math scores of girls in their class, (Leroy et

al., 2022). Large-scale programs featuring female role models also boosted female students’
GENDERED CAREER ASPIRATIONS 10

aspirations for science-related careers (Breda et al., 2023). Thus, students’ career

aspirations—e.g., aspirations to careers with a high versus low proportion of (fe)male role

models—not only follow from but may also affect the development of relevant academic

skills and interests. For example, aspiring to a math- or science-related career affects

students’ academic beliefs, such as self-concept in math, which in turn affects career

aspirations and later career attainment in the math domain (Lauermann et al., 2017). Such

reciprocal developmental patterns are predicted in SEVT (Eccles, 2009) but are rarely

studied. Moreover, gender differences in subject-specific achievement and ability self-

concepts across math versus verbal domains can drive differences in career aspirations and

training. Students with higher math and moderate verbal ability are likelier to enroll in STEM

than students who do well in both math and verbal domains (e.g., Wang et al., 2013).

Students who excel in verbal domains but are less skilled in math are more likely to

specialize in the humanities (e.g., see Parker 2014). Even when there are no achievement

differences, gender differences in students’ math self-concepts have been documented; when

such differences exist, they generally favor boys (Lauermann et al., 2017; Parker et al., 2018).

International evidence suggests that girls typically have a relative advantage in verbal over

math domains, even in contexts in which girls outperform boys in math; this relative verbal-

over-math advantage partially explains gendered differences in math-related educational and

career outcomes (Breda & Napp, 2019; S. Wan et al., 2023).

The Role of Gender in the Differential Effects Model

Parker et al.’s (2012) differential effects model combines key aspects of the

circumscription and compromise and the SEVT frameworks. This model suggests that

students’ gender and SES have differential effects on future educational and career-related

choices: gender plays a critical role in shaping students’ areas of specialization (e.g., college

majors), whereas students’ SES is an important factor influencing their educational

GENDERED CAREER ASPIRATIONS 11

aspirations and attainment (e.g., university enrollment). Students’ expectancies and values

mediate the effects of gender and social status. Across two large samples collected from

Germany and the U.K., the authors showed that students with comparatively worse math and

better verbal self-concept of ability were less likely to pursue a college major in the physical

sciences, math, or engineering. Moreover, even after controlling for differences in SES, math

and verbal achievement, and verbal, math, and science ability self-concepts, women were still

less likely to pursue a college major in the physical sciences, math, or engineering.

Accordingly, the analyses corroborated both direct effects (women are less likely to pursue

male-dominated careers) and indirect effects (via students’ subject-specific self-concepts) of

gender on pursuing gender-(a)typical educational (and thus presumably also occupational)

goals.

Other studies have shown similar direct and indirect effects of gender. In Marsh et

al.’s (2019) longitudinal study, differences in math and reading self-concepts partly mediated

gender disparities in STEM enrollment at the high school level. However, a persistent, direct

effect of gender on women’s enrollment in STEM majors remained. Notably, the indirect

effect of gender through motivation variables such as self-concept varies within the sciences

and not just between STEM and non-STEM fields. In Parker et al.’s (2020) meta-analysis,

boys had higher self-concepts of ability in the physical sciences, and girls had higher self-

concepts of ability in the biological sciences. The same pattern was observed for other

intrinsic, attainment, and utility value measures. These lines of research support the key

assumptions of the differential effects model but might overlook significant heterogeneity in

the role of gender in shaping adolescents’ career aspirations and attainment that is not

captured by broad categories of fields, such as STEM versus non-STEM. Additionally, the

differential effects model is applied to a specific stage of career training: choosing a

university major upon entering college, which represents only a single decision point. As
GENDERED CAREER ASPIRATIONS 12

described by the theories it relies on (i.e., circumscription and compromise and SEVT), such

a decision follows a developmental trajectory, as students may vacillate between more or less

gender-(a)typical career options over time.

Heterogeneous Developmental Trajectories of Gender-(A)typical Career Aspirations

We are aware of only two studies that have treated gender typicality as an outcome

variable in longitudinal models of career aspirations (Basler et al., 2021; Lawson et al.,

2018). Both studies demonstrated a similar U-shaped pattern of gender typicality in career

aspirations, where typicality increases in early adolescence and drops off after a peak in mid-

adolescence. Both studies also showed that boys were more likely to aspire to gender-typical

careers than girls, and both also separated boys and girls by additional variables, such as

parental ratings of gender-typical skills and interests like dancing or sports. Lawson et al.

(2018) showed that boys aspired to more gender-typical careers during childhood and

adolescence but obtained slightly less gender-typical occupations later; however, the gender

typicality of men’s careers was still relatively high.

Meanwhile, on average, girls aspired to more gender-neutral careers, but as adult

women, they tended to obtain more gender-typical jobs (Lawson et al., 2018). Gender-typical

skills (e.g., dancing, sports, or math) mediated the associations between gender and the

gender typicality of later choices, as those who had more gender-typical skills showed greater

aspirations towards gender-typical careers. Basler et al. (2021) reported similar findings, but

individuals in higher academic tracks in secondary school were more likely to aspire to and

obtain gender-neutral careers. This evidence shows that gendered aspirations can change

significantly over time, and there are notable interindividual differences in the amount and

propensity of change based on math and verbal self-concepts and achievement. However, the

available evidence has not yet focused on potentially heterogeneous trajectories of gender-

typical career aspirations and the co-development of gendered academic skills and beliefs.
GENDERED CAREER ASPIRATIONS 13

The Present Study

To our knowledge, no study has directly examined the connections between

adolescents’ subject-specific achievement and self-perceptions (e.g., subject-specific self-

concepts) and potential changes in the gender typicality of career aspirations over time. In

addition, the few existing longitudinal studies on the gender typicality of career aspirations

have not examined potential heterogeneity in developmental trajectories (e.g., moving in and

out of gender-typical choices). The present study extends prior research in two key ways.

First, we match career aspirations directly to the proportion of women in the field. In this

way, we examine changes in the gender typicality of career aspirations directly rather than

inferring it based on a chosen field such as STEM. Second, we model career aspirations as a

dynamic developmental process and study interindividual differences in pursuing (a)typical

career goals. To do so, we use growth-mixture modeling of the gender typicality of preferred

careers reported by a nationally representative sample of German adolescents each year

between 9th and 12th grade. Third, we study the potential longer-term implications of different

developmental trajectories for students’ educational outcomes and areas of specialization by

linking this dynamic developmental process to the gender typicality of university majors of

our participants. Finally, we examine the co-development of academic skills and beliefs in

the math and verbal domains, often perceived as gender-typical in our study context.

Thus, the present study has three main research questions (RQ): First (RQ1), we

attempt to identify separate trajectories (latent classes) that show the yearly change in the

proportion of women in aspired careers from 9th to 12th grade (i.e., gender typicality).

Because prior research on changes in gender typicality of career aspirations is scarce, it is

difficult to predict how many distinct career choice trajectories will emerge. Nonetheless,

based on the theories and evidence reviewed in the previous sections, we expect most
GENDERED CAREER ASPIRATIONS 14

adolescents to exhibit a relatively stable gender-typical trajectory. However, some, especially

girls, may experience fluctuations in the gender typicality of their career aspirations.

Second (RQ2), we expect the trajectories will differ based on individual factors, such

as students’ gender, SES, math and verbal achievement, and self-concepts of math and verbal

ability. Based on TCC and the differential effects model, we expect a strong association

between gender and students’ aspirations toward gender-typical careers but make no

predictions for students’ SES, which is less relevant for areas of specialization. Consistent

with SEVT, and as discussed in the previous section, prior work has supported reciprocal

links between students’ domain-specific self-concepts of ability and their math-related career

aspirations, as students’ academic and career-related aspirations tend to influence each other

over time. 2 Following this research, we expect those whose aspirations change toward a more

male-dominated career will have a higher math ability self-concept, and those whose

aspirations change to a more female-dominated career will have a higher verbal ability self-

concept. The same pattern may occur for students’ math and verbal achievement, controlling

for differences in general cognitive ability.

Lastly (RQ3), we expect that gendered career choice trajectories will relate to the

gender typicality of future educational choices, such as enrollment in advanced math and

language courses in high school and the gender-typicality of the chosen university majors.

We base our prediction on prior work identifying high levels of gender segregation in

advanced high school math and language courses and specific majors (e.g., mechanical

engineering and biology). We expect that those who change their aspirations to a male-

dominated career and those who aspire to such a career throughout high school will be more

likely to pursue an advanced math course in secondary school and enroll in a male-dominated

2
While academic task values (e.g., science interest) are likewise predicted to be important, we are limited by
our dataset, which does not include measures of intrinsic, utility, or attainment value across multiple time points.
Thus, similar to Parker et al. (2012), our analyses focus only on the expectancy component of SEVT.
GENDERED CAREER ASPIRATIONS 15

major in university. Similarly, individuals who aspire towards a female-dominated career or

whose aspirations change to one will be more likely to take advanced language arts courses in

secondary school and opt for a more female-dominated major.

For all analyses involving covariates (RQ2-RQ3), we examined whether the identified

developmental patterns are gender-specific (i.e., whether gender moderates the associations

between career aspiration trajectory and relevant covariates or outcomes).

Methods

Participants and Data

This study used data from the 9th-grade cohort (Starting Cohort 4; SC4) of the

German National Education Panel Study (NEPS; Blossfeld et al., 2019). A correlation matrix

of all variables used is shown in Table 1. All academic track (i.e., Gymnasium) students who

responded to at least one measurement point regarding their ideal career (N = 4,759) were

included in the analyses. Due to our interest in decisions about university majors, data was

limited to academic track secondary-level students (i.e., the typical path to qualify for

university entrance). Measurements were taken yearly from 9th to 12th grade with a single

follow-up after four years after 12th grade. Participation varied by wave (n9th grade = 3,559,

ncollege = 2,692). The number of female participants varied slightly by measurement point (53-

58%).

Gender

Gender (0 = male, 1 = female) was coded based on parent responses at each

measurement point (NEPS variable “tx80501”). In the few (n = 14) cases in which the gender

code changed over time, only the most recent value was used. Such incidents reflected stable

changes (i.e., they did not change back).

SES
GENDERED CAREER ASPIRATIONS 16

The International Socioeconomic Index (ISEI; see Ganzeboom et al., 1992), scaled

based on education and income associated with having a specific occupation, is provided in

the NEPS data for parents’ occupations. ISEI values are on a 100-point scale with higher

values relating to a higher level of SES (e.g., a carpenter has an ISEI rating of 31, and an

architect has an ISEI rating of 77). If parents had different ISEI values, the higher value was

used in our analyses.

Nonverbal Reasoning

Nonverbal reasoning was evaluated as an indicator of general cognitive ability by the

sum score on the NEPS nonverbal reasoning scale (variable “dgg9_SC3b” in NEPS),

assessed in 9th grade. The test involved matching pictures and symbols in a matrix design and

was used as a control variable in our analyses (see Haberkorn & Pohl, 2013). Scores were

based on the total number of items solved, ranging form 0 to 12. Nonverbal reasoning is

included in the present study as a control variable.

Math and Verbal Test Performance

Math and reading achievement in NEPS was measured via comprehensive

standardized tests in 9th and 12th grade (Duchhardt & Gerdes, 2013; Haberkorn et al., 2012).

We used Warm’s weighted likelihood estimate (WLE) for both achievement scores. Values

were z-standardized based on 9th grade data (M = 0, SD = 1) so that higher values in 12th

grade reflect a growth in achievement relative to 9th grade.

Math and Verbal School Grades

Students self-reported their math and German grades in 9th and 12th grade. As is

typical for the German school system, grades ranged from 1 (very good) to 6 (insufficient) in

9th grade and from 15 (very good) to 0 (insufficient) in 12th grade. All grades were recoded so

that higher numbers consistently indicated better 9th- and 12th-grade performance.

Self-Concepts of Math and Verbal Ability

GENDERED CAREER ASPIRATIONS 17

Math and verbal self-concepts of ability were measured in 9th and 12th grade by three-

item scales (e.g., “I get good grades in [Math/German],” αMath = .89, αVerbal = .81), and

response options ranged from 1 (does not apply at all) to 4 (applies completely; NEPS

variables t66001a_g1 and t66000a_g1).

Gender Typicality of Career Aspirations

In 9th, 10th, 11th, and 12th grade, students reported their ideal careers when they

became adults. Responses were coded using the German classification of occupational titles

(KldB 2010) and matched to the concurrent proportion (0.00–1.00) of female employees in

German census reports (Federal Statistics Office of Germany, 2018). For example, a value of

0.15 means that 15% of German workers in that job are women.

Advanced Math and Verbal Courses

In 11th and 12th grade, students in academic-track schools can take advanced courses

in different subjects (i.e., “Leistungskurse”). These courses reflect a level of specialization

before applying to university programs. We record whether the students attended advanced

courses in math and German based on students’ self-reports (variables “te13010” and

“te13020” in NEPS).

Gender Typicality of University Majors

At the end of secondary education, the participants were asked what university majors

they would apply for. Four years later, they were asked for their current or most recent major.

Both were categorized by the ISCED-97 scale and matched to the current proportion (0.00–

1.00) of female students in each study field in German universities (Federal Ministry for

Education and Research, 2021). For example, a value of 0.15 means that 15% of the people

studying that major in German universities are women.

Analyses
GENDERED CAREER ASPIRATIONS 18

We discuss our analyses in three parts, corresponding to our main research questions.

All analyses were conducted with Mplus 8.6 (Muthén & Muthén, 1998-2017). The syntax for

each analysis is reported in the online supplemental materials.

Analysis 1: Identifying Career Aspiration Trajectories

A growth-mixture model (Ram & Grimm, 2009) with random intercept and slope

values examined the change in the gender typicality of career aspirations of students from 9th

to 12th grade with full information maximum likelihood estimation (FIML). A nested model

was calculated using clustering around the individual schools to account for hierarchical data.

Initially, growth mixture models with unequal variances for the intercept and slope

parameters were considered. However, these models produced unacceptable results with two

or more classes due to negative variances, multicollinearity, or both, as evidenced by a not

positive definite covariance matrix. In cases like this, Sijbrandij et al. (2020) recommend

selecting fixed effects models (e.g., a latent class growth analysis) or constraining residual

variances to be equal across classes. We rejected the fixed effects model because we expect

values to vary across and within classes based on prior latent growth models in similar data

(Basler et al., 2021; Lawson et al., 2018). We next considered a covariance pattern growth

mixture model (see McNeish et al., 2022), but this model did not converge, which is a

common problem with this type of analysis. As a result, we fixed the residual variances to be

equal for the intercept and slope between classes (i.e., a homoscedastic model). From one to

five class models were examined. Models with six or more classes were excluded because

invalid results were obtained, specifically invalid variable covariance matrices due to

multicollinearity. Table 2 describes the fit metrics of each class model. A 5-class model was

chosen because it produced the best-fitting values and yielded readily interpretable classes of

sufficient size.
GENDERED CAREER ASPIRATIONS 19

To assess possible bias due to a homoscedastic model, a posthoc analysis of the

selected model showed only a marginal difference in the observed intercept variance between

classes in our sample (Bartlett’s K2 (4) = 23.59, p <.001, ω = 0.081). This effect was much

smaller than those modeled in numerous simulation studies examining homoscedastic and

heteroscedastic growth mixture models (Enders & Tofighi, 2008; McNeish & Harring, 2017;

McNeish et al., 2022). Moreover, the most common difficulty reported in homoscedastic

models is the incorrect enumeration of classes, which is more common in low sample sizes

(e.g., <500; McNeish et al., 2022). Although estimation problems did arise in the

unconstrained model (collinearity issues), the preferred number of classes was the same as in

the constrained model, and parameter estimates were highly similar, except for C4 (neutral-

to-female), whose slope was somewhat lower (10.8 instead of 15.8). Finally, the greatest

parameter bias between homoscedastic and heteroscedastic models in simulation studies was

reported in intercept and quadratic components (see Enders & Tofighi, 2008). We have no

quadratic parameters, and the difference in variance between our intercept parameters was

negligible (see above). We prefer the homoscedastic model because the heteroscedastic

models produced invalid results, had very similar parameter estimates, and led to the same

class enumeration.

Analysis 2: Predictors of Career Aspiration Trajectories

We examined the second research question in three steps. First, we tested four

separate mixed ANOVAs to examine mean-level differences in students’ math and verbal

achievement, school grades, and self-concepts of ability, depending on students’ gender,

grade level, and gender-(a)typical career aspiration trajectory, with grade-level as a within-

subjects variable. These analyses allowed us to identify potential gender differences in the

levels of performance and self-concepts across career aspiration trajectories. Second, multiple

imputations were performed in Mplus 8.6 (Muthén & Muthén, 1998-2017) to examine the
GENDERED CAREER ASPIRATIONS 20

relations of trajectory membership with a broader range of predictor variables. Gender,

household income, parental ISCED, parental ISEI, nonverbal reasoning scores, 9th- and 12th-

grade achievement in math and reading, 9th- and 12th-grade math and reading self-concept,

9th- and 12th-grade math and German school grades, major applied for, and major studied

were multiply imputed with 100 simulations from a MCMC sequence (see Asparouhov &

Muthén, 2022 for details on multiple imputation and the integration of results using Mplus).

For a comparison of the imputed datasets with the original data, see Online Supplement S1.

Predictive effects were estimated using a three-step inclusion method (i.e., “R3STEP” in

Mplus). Variables included SES, 9th- and 12th-grade German and math grades, 9th- and 12th-

grade math and reading self-concept of ability, 9th- and 12th-grade German and math

achievement, nonverbal reasoning, and gender. Third, a path analysis examined the role of

trajectory membership on the growth of subject-specific achievement and self-concept

between 9th and 12th grade, with separate models calculated based on gender. For both math

and verbal domains, grade 12 achievement, self-concept, and grades were regressed on 9th-

grade scores, trajectory membership, and ISEI. Career trajectory membership was dummy-

coded across trajectories 1, 2, 3, and 4. The stable-male trajectory served as a reference

group.

Analysis 3: Links between Career Aspiration Trajectories and Educational Choices.

Students’ educational choices were incorporated as outcome variables (i.e., “BCH”

auxiliary variables) into the growth-mixture model in Mplus. These outcomes included

enrollment in advanced German and math high school courses, the gender typicality of the

university major they applied to, and the major they later studied.

Results

Analysis 1: Five Trajectories of the Gender-(A)Typicality of Aspired Careers

GENDERED CAREER ASPIRATIONS 21

The five-class solution had the best fit to the data (see Table 2) and revealed five

distinct career-choice trajectories (see Figure 1). The trajectories reflected stable-male (C1),

stable-female (C2), and stable-neutral (C3) career aspirations, as well as changed from male-

to-female (C4) and neutral-to-male (C5) career aspirations. As shown in Table 3, the most

common job aspirations in the stable-male trajectory (C1) included athletes, CEOs, and

pilots, while typical jobs in the stable-female (C2) trajectory included teachers, veterinarians,

and psychologists. Stable-neutral (C3) aspirations included architects, medical practitioners,

journalists, and psychologists. Typical aspirations in the male-to-female trajectory (C4)

shifted from CEOs, athletes, and pilots at the beginning of high school to medical workers,

psychologists, and teachers at the end of high school. Finally, typical aspirations in the

neutral-to-male (C5) trajectory shifted from medical practitioners, architects, and lawyers at

the beginning of high school to CEOs, athletes, and telecommunication engineers later on.

Thus, the most common career aspirations show clear differentiation by gender-typicality

across the five trajectories. A more detailed list of the most common career aspirations in

each trajectory and time point is provided in the online supplementary materials.

Slightly more than half (52%) of all participants followed a gender-differentiated

trajectory, showing a strong gender-specific preference in career aspirations by 12th grade

(C1, C2, C4, and C5, Figure 1). While most students had stable preferences for male-

dominated (C1), female-dominated (C2), or gender-balanced careers (C3), a significant

minority (13% of cases) experienced a substantial change in gender typicality of their aspired

career (C4 and C5, Figure 1). On average, the stable-male (C5) and stable-neutral (C3)

trajectories changed by only 2.5% and 1.4% each year, representing a very small shift to

careers with more women in them (ps < .001). A slightly larger change of -4.4% each year on

average emerged in the stable-female (C2) trajectory, showing a modest shift toward careers

with fewer women in them (p < .001). In the neutral-to-male (C5) trajectory, the trajectories
GENDERED CAREER ASPIRATIONS 22

changed by an average of 10.9% at each measurement point, indicating a stark shift towards

careers with fewer female workers (p < .001). Meanwhile, the largest shift emerged in the

male-to-female (C4) trajectory, with an average change of 15.8% per year, indicating a large

shift towards careers employing a larger proportion of women (p < .001).

Analysis 2: Differences in Students’ Gender-(A)Typical Career Aspiration Trajectories

by Gender, SES, Achievement Indicators, and Self-Concepts of Ability

Gender and SES

As shown in Table 4, after controlling for SES, nonverbal reasoning, achievement,

self-concept, and grades, disproportionately more girls were in the stable-female (C2), stable-

neutral (C3), male-to-female (C4), and neutral-to-male (C5) trajectories than the stable-male

(C1) trajectory (ps < 0.001). This pattern indicates that girls in secondary school were less

likely than boys to aspire to careers dominated by men. If they did so, they were likelier to

opt for a male-dominated career later in secondary school than their male peers. As expected,

no relations existed between trajectory membership and SES (ps > 0.10).

Achievement Indicators

Nonverbal reasoning. There were no relations between nonverbal reasoning and

trajectory membership (ps> 0.10).

Math and verbal school grades. Participants in the stable-female (C2) trajectory had

better German grades in 9th grade than those in the stable-male trajectory (C5), p = 0.044, and

no other significant mean-level differences emerged for school grades (see Table 4).

However, as shown in Table 5, controlling for differences in SES and 9th-grade achievement,

grades, and math/verbal self-concepts, male students had larger gains in German grades by

12th grade if they were on the stable-female (C2), stable-neutral (C3), and male-to-female

(C4) trajectories, compared to the stable-male trajectory (C1). At the same time, boys in the

male-to-female (C4) trajectory had relatively smaller gains in math grades than boys on the
GENDERED CAREER ASPIRATIONS 23

stable-male (C1) trajectory. In contrast, girls’ math and verbal grade changes did not depend

on trajectory membership.

Math and verbal test scores. As shown in Figure 2, math achievement increased

from 9th to 12th grade, F (1, 3248) = 277.6, p <.001, and boys scored better in math than girls,

F (1, 3248) =75.0, p <.001. Math achievement varied based on career aspiration trajectory, F

(4, 3248) = 14.0, p < .001. However, the differences between boys and girls shrank

significantly from 9th to 12th grade in the neutral-to-male trajectory, as revealed by a gender-

by-trajectory interaction, F (4, 3248) = 3.1, p = .015. Gains from 9th to 12th grade were largest

in the stable-male (C1) and neutral-to-male (C5) trajectories, as evidenced by a grade level-

by-trajectory interaction, F (4, 3248) = 2.6, p = .033. Meanwhile, reading achievement also

increased from 9th to 12th grade, F (1,3211) =16.2, p < .001, and boys had lower achievement

than girls, F (1, 3211) = 33.8, p < .001. Career aspiration trajectory and the co-development

of reading achievement were not significantly related, p > .10, and there were no interactions

by gender, all ps > .10.

ANOVA results are supported by the predictor analysis shown in Table 4, which

controls for other predictor variables (i.e., gender, SES, gender, grades, and subject-specific

self-concepts of ability). In 9th grade, those in the neutral-to-male (C5) trajectory showed

lower math achievement than those in the stable-male (C1) trajectory, p = 0.025. However,

by 12th grade, this difference vanished, p > 0.25. In 12th grade, those in the stable-female (C2)

and stable-neutral (C3) trajectories tended to have worse math achievement than those in the

stable-male (C1) trajectory, p < 0.057 and p = 0.016, respectively. No significant mean-level

differences between trajectories emerged for reading achievement (see Table 4).

Finally, as shown in Table 5, controlling for differences in SES, grades, and 9th-grade

self-concepts, boys on the stable-female (C2), stable-neutral (C3), and male-to-female (C4)

career-aspirations trajectory had smaller gains in math achievement than boys in the stable-
GENDERED CAREER ASPIRATIONS 24

male (C1) trajectory, and boys in the neutral-to-male (C5) trajectory had similar levels to

boys on the stable-male (C1) trajectory. Meanwhile, girls on the neutral-to-male (C5)

trajectory had larger gains in math achievement than girls on the stable-male (C1) trajectory.

Boys’ verbal achievement gains did not differ based on trajectory membership, but girls on

the male-to-female (C4) trajectory had smaller gains in verbal achievement than girls on the

stable-male (C1) trajectory. Thus, relative gains in math achievement depended heavily on

trajectory membership for boys, but for girls, gains (in math achievement) differed only for

those on the neutral-to-male (C5) trajectory.

Math and Verbal Self-Concepts of Ability

As shown in Figure 2, math self-concept decreased from 9th to 12th grade, F (1,3354)

= 52.9, p = .001, and boys had an overall higher math self-concept than girls, F (1,3354) =

21.1, p < .001. Notably, students’ math self-concept varied by trajectory, F (4, 3354) = 18.4,

p < .001, and gender, as evidenced by a significant trajectory-by-gender interaction. F (4,

3354) = 3.0, p = .018. For boys, it was highest in the stable-male (C1) trajectory, but for girls,

it was highest in the neutral-to-male (C5) trajectory, where their self-concept was at the same

level as boys’ (see Figure 2).

Verbal self-concept also decreased from 9th to 12th grade, F (1,3374) = 4.7, p = .029.

Boys had lower verbal self-concepts than girls, F (1, 3374) = 72.7, p <.001, and their verbal

self-concept delined significantly more than girls’ from 9th to 12th grade, as indicated by a

significant grade-by-gender interaction F (1,3374) = 16.9, p < .001. Students in the stable-

female (C2) and stable-neutral (C3) trajectory had higher verbal self-concepts, F (4, 3374) =

7.3, p < .001, but this pattern varied in a three-way interaction between gender, grade level

(9th vs. 12th) and trajectory, F(4, 3374) = 5.0, p < .001. By 12th grade, the difference in verbal

self-concept between boys and girls narrows in the stable-female (C2) trajectory but widens

in the remaining four trajectories.

GENDERED CAREER ASPIRATIONS 25

After controlling for gender, SES, and achievement differences, there were several

differences in students’ math and verbal self-concepts of ability, using the stable-male (C1)

trajectory as a reference group. In 9th grade, those in the stable-female (C2) and stable-

neutral (C3) trajectory showed lower levels of math self-concept than the reference group

(C1). By 12th grade, those in the male-to-female (C4), as well as the stable-female (C2), and

stable-neutral (C3) trajectory had significantly lower math self-concepts than the reference

group (C1), all ps < 0.05. In 9th and 12th grade, those in the stable-female (C2) trajectory

tended to show a higher German self-concept, p = 0.063 and 0.051, respectively. In 12th

grade, those in the stable-neutral (C3) trajectory had a higher German self-concept, p =

0.003, while those in the male-to-female (C4) trajectory tended towards a higher German

self-concept, p = 0.086.

Table 5 shows the gender- and trajectory-specific relative changes in self-concept

from 9th to 12th grade. After controlling for SES and 9th-grade test scores and grades, boys in

the stable-female (C2), stable-neutral (C3), and male-to-female (C4) trajectories had larger

decreases in math self-concept and larger gains in verbal self-concept than boys on the stable-

male (C1) trajectory. In contrast, changes in girls’ self-concepts did not depend on trajectory

membership and showed a slight decline across all trajectories.

Analysis 3: Educational Choices and Career Aspirations

Links between students’ career aspiration trajectories and their educational choices—

course selection and university major—were in the expected direction, as shown in Table 6.

Tendencies toward female-dominated careers corresponded to a higher likelihood of enrolling

in advanced high-school classes in language arts and a lower likelihood of enrolling in

advanced math classes. That is, students on the stable-female (C2) and stable-neutral (C3)

trajectories were more likely to enroll in advanced German courses and less likely to enroll in

advanced math courses than those on the stable-male (C1) trajectory. Those on the male-to-
GENDERED CAREER ASPIRATIONS 26

female trajectory were less likely to attend advanced German courses than those on the

stable-female (C2) trajectory and less likely to attend advanced math courses than those on

the stable-male (C1) trajectory. Those on the neutral-to-male (C5) trajectory were less likely

to attend advanced German courses than those on the stable-female (C2) trajectory and less

likely to attend advanced math courses than those on the stable-male (C1) trajectory, yet

more likely to attend advanced math courses than those on the stable-female (C2) trajectory.

The neutral-to-male (C5) trajectory was not significantly different from the stable-

male (C1) trajectory in the gender ratio in majors they applied to or studied, p > 0.10, as both

groups applied to and studied male-dominated fields (see Table 6). In contrast, participants in

the stable-female (C2), stable-neutral (C3), and male-to-female (C4) trajectories applied to

and studied majors that had significantly more women than the stable-male trajectory (C1),

all ps < 0.001. Accordingly, career aspirations reflected actual major decisions. Specifically,

students on the stable-male (C1) and neutral-to-male (C5) trajectory chose majors dominated

by men, students on the stable-female (C2) and male-to-female (C4) trajectory chose majors

dominated by women, and students on the stable-neutral (C3) trajectory chose relatively

gender-balanced majors. Thus, although students on change trajectories differed in the

number of advanced math and language courses they took, they pursued similarly gender-

(a)typical majors.

Discussion

This study showed individual differences in the development of gender typicality of

career aspirations during secondary school, which were linked to students’ high-school

course enrollment and later college major selection. Although several studies have used

group-based latent growth models to look at changes in gender typicality of career aspirations

(e.g., Basler et al., 2021; Lawson et al., 2018), these studies looked for mean level changes

and did not explore potentially heterogeneous trajectories over the development of career
GENDERED CAREER ASPIRATIONS 27

aspirations. We identified five distinct developmental trajectories, including three stable

trajectories and two with sizeable changes in gender typicality, namely those who shift away

from male-dominated careers (male-to-female trajectory) and those who shift towards them

(neutral-to-male trajectory). Adolescents’ gender, self-concept of math ability, and math

achievement were important predictors for different trajectories, revealing co-developmental

processes between career aspirations and academic beliefs and skills. Additionally, boys

showed trajectory-specific changes in subject-specific (math/verbal) self-concept from 9th to

12th grade, while girls did not, pointing to gender-specific developmental processes. Finally,

career aspiration trajectories were related to subsequent educational choices and career

training activities (i.e., enrollment in advanced courses in German or math and gender-

(a)typical university majors).

RQ1: Gender-(A)typical Career Aspiration Trajectories

Gottfredson and Lapan (1997) predict that boys are more likely to pursue a gender-

typical career and less likely to make compromises based on gender typicality. Other research

has also shown that boys are more likely to pursue gender-typical careers (Basler et al., 2021;

Lawson et al., 2018). Lawson et al. argue that boys may be limiting their career options along

gender lines at an earlier stage of development (also see Gottfredson, 1996), resulting in more

gender-typical career aspirations during secondary school. Our data provide partial support

for this claim and contribute some novel observations. We found that most participants

showed only minor changes in the gender typicality of their ideal career from 9th to 12th

grade, although career aspirations do change to an extent even within stable gender-typicality

trajectories. For instance, over the course of the study, aspirations to become a pilot dropped

from 11% to 5% among stable-male trajectory members, and interest in becoming a

secondary education teacher fell from 12% to 6% among stable-female trajectory members.
GENDERED CAREER ASPIRATIONS 28

However, even when students changed their specific career aspirations, compromise across

gender lines remained rare.

The analyses further suggest that a sizeable minority of students—often overlooked in

mean-level analyses of career aspirations—follow gender-atypical trajectories. About 8% of

girls follow a gender-atypical trajectory (i.e., the stable-female or neutral-to-male

trajectories). Similarly, over an eighth (14%) of boys followed a gender atypical trajectory

(i.e., the stable-female or male-to-female trajectories). Thus, a greater proportion of boys

aspire to gender-atypical careers, which—to an extent—conflicts with TCC’s assertion that

boys are less likely to compromise their career aspirations by aspiring to a gender-atypical

career. Prior research looking at mean-level gender typicality of career aspirations has also

found that boys are less likely to aspire to gender-atypical careers (Basler et al., 2021;

Gottfredson & Lapan, 1997; Lawson et al., 2018). An explanation of this seeming

contradiction in results comes from the heterogeneous trajectories we identified. Most girls in

our study did not follow the stable-female but the gender-neutral trajectory. At the same

time, more boys were following a stable-male career aspiration trajectory than girls following

a stable-female trajectory. Thus, by identifying heterogenous trajectories, we can see that

boys were more likely than girls to aspire to either gender-typical or gender-atypical careers,

while girls were more likely to aspire to gender-neutral careers. When looking only at mean

levels, the large size of the gender-neutral trajectory hides the small but notable minority of

boys aspiring to female-dominated careers. Indeed, previous work has shown girls are more

likely to have high aptitudes in both math and verbal domains (Breda & Napp, 2019; S. Wan

et al., 2023; Wang et al., 2013), and our analyses suggest that they are more likely to aspire to

gender-neutral careers. At the same time, boys were more likely than girls to follow a change

trajectory, with approximately 18% of all boys on the male-to-female or neutral-to-male

trajectories, while only 7% of girls were on the same trajectories. These changes are similarly
GENDERED CAREER ASPIRATIONS 29

lost when mean-level analyses are used. Notably, as discussed subsequently, boys’ career

aspiration changes were related to their academic development, but girls’ were not.

RQ2: Factors Relating to Each Trajectory: The Co-Development of Gendered

Academic Skills and Beliefs

Students’ subject-specific self-concepts and achievement were predictive of gendered

career aspirations. All trajectories except the neutral-to-male (C5) were likely to have similar

levels of achievement to the stable-male (C1) trajectory in both math and reading, while

those in the neutral-to-male (C5) trajectory had a lower math achievement. At the same time,

those in the neutral-to-male (C5) trajectory had a similar level of math self-concept as those

in the stable-male (C1) trajectory. By 12th grade, those in the neutral-to-male (C5) trajectory

had a math self-concept and achievement comparable to those in the stable-male (C1)

trajectory. These results provide context for previous work relating self-concept of math

ability with aspirations toward male-dominated occupations. Such work focused on indirect

measures of gender-typicality, such as aspirations to math-intensive careers (Lauermann et

al., 2017) and university majors (Parker et al., 2012), and it did not examine heterogenous

trajectories of development. By focusing on heterogeneous trajectories, we were able to show

that different career aspiration trajectories correspond to different developmental patterns for

students’ academic abilities and beliefs in the math and verbal domains. For instance,

aspiring to a male-dominated career was associated with higher math self-concept and

achievement, while aspiring to a female-dominated career was associated with higher verbal

self-concept and achievement, and changes in the gender typicality of career aspirations

corresponded to changes in math and verbal self-concept and achievement.

As expected, we found no significant effects of SES relating to career aspiration

trajectory, self-concept of ability, achievement, or changes in achievement or self-concept of

ability. Our sample of students in the academic track in the German system may partially
GENDERED CAREER ASPIRATIONS 30

explain this result. These students are already similar in SES, are enrolled in academic track

schools, and are thus generally expected to attend higher education (Parker et al., 2012). In

the differential effects model (Parker et al., 2012), SES influences the likelihood of attending

a university, but not necessarily the area of specialization, such as gender-(a)typical college

major selection. Our results generally align with this conclusion. However, Basler et al.

(2021) identified a sizeable gender-by-school track interaction when looking at gendered

career aspirations, where adolescent boys and girls on a higher track aspired to less gendered

careers. They attributed this finding to the availability of higher-status jobs with a relatively

equal ratio of male and female workers. We could not replicate this finding in our data, and

analyses across a broader range of students and school types may be necessary to examine the

context-specific effects of SES on gendered career aspirations.

Our analyses also revealed gender differences in the mean level of students’ math and

verbal achievement and self-concepts of ability across different career aspiration trajectories

and significant gender-by-trajectory differences in self-concepts and in changes in self-

concept and achievement over time. Compared to boys on other trajectories, boys in the

neutral-to-male and stable-male trajectories had significantly higher 12th-grade math self-

concept and achievement and significantly worse German grades than predicted by their 9th-

grade self-concept, achievement, and grades. In contrast, girls on the neutral-to-male

trajectory showed a relatively higher gain in math achievement, while girls on other

trajectories showed no trajectory-specific gains from 9th to 12th grade.

This finding represents a somewhat surprising caveat to our prediction that math and

verbal achievement and self-concept of ability will correspond to aspiration towards a male-

or female-dominated career. The co-development of boys’ self-concept and achievement with

their career aspiration trajectories support predictions made by Eccles (2009) and Lauermann

et al. (2017), where career aspiration changes toward math-intensive fields relate to positive
GENDERED CAREER ASPIRATIONS 31

changes in math task values. However, the changes in girls’ self-concepts did not differ based

on the gender-(a)typicality of their career aspirations. One possibility is that girls tend to be

more broadly academically motivated than boys, whereas boys’ developing academic

interests tend to be more specialized (e.g., in math-related fields). For instance, Denissen et

al. (2007) showed that the positive relations between students’ subject-specific interests, self-

concepts, and performance increased over primary and secondary school, pointing to an

increasing specialization of motivation and performance across domains. However, the

degree of alignment in boys’ interests, self-concepts, and performance was significantly

higher than that for girls, as girls were more likely than boys to do well in school subjects for

which they reported comparatively lower academic motivation.

The change trajectories we identified also add nuance to previous results that focused

on relative math and verbal achievement and self-concept predicting STEM career aspirations

and entry. For instance, Wang et al. (2013) observed that many girls with high ability in both

language and math did not enter STEM careers, likely because they had more choices

available to them. In our study, girls on the stable-female trajectory did have higher verbal

achievement and self-concept and lower math achievement and self-concept than girls on

other trajectories. At the same time, a significant minority of girls (approximately 8%) had

high math and verbal achievement and aspired to male-dominated careers (i.e., girls on the

stable-male or neutral-to-male trajectories). These girls excelled in multiple domains and

aspired to more male-dominated careers. Wang et al. (2013) argued that such girls tend to be

pushed out of male-dominated career paths because they have more options available to

them, but this important minority in our study does the opposite. These heterogeneous

trajectories reveal individually different developmental patterns and need to be studied in

more detail to understand better why some learners aspire to gender-atypical careers.
GENDERED CAREER ASPIRATIONS 32

Another interesting pattern emerges when comparing the math and verbal

achievement and self-concept of ability of girls on the stable-male and neutral-to-male

trajectories. Girls on the neutral-to-male trajectory have consistently higher math self-

concepts and achievement levels than girls on any other trajectory, but boys on the stable-

male trajectory have consistently higher math self-concepts of ability and achievement levels

than any other boys. These results suggest that girls with higher ability levels in math opt for

a male-dominated career later in development than boys. The same tendency was evident for

boys on the male-to-female (C4) trajectory compared to boys on the stable-female (C2)

trajectory, but it did not reach significance. Specifically, the verbal ability of boys on the

male-to-female trajectory was slightly higher than that of boys on the stable-female trajectory

but not at the level of statistical significance. It is possible that a more comprehensive

assessment of verbal ability—beyond reading skills—is needed to contribute to a better

understanding of gender differences in this domain (see Lepper et al., 2022; McGeown et al.,

2016 on gender differences in motivation and skills in the verbal domain).

RQ3: What are the Longer-Term Educational Implications of Gender-(A)Typical

Career Aspiration Trajectories?

Career aspiration trajectories related to future educational and career training choices.

Students in the stable-female (C2) and stable-neutral (C3) trajectories were more likely to

take advanced German courses in secondary school and less likely to take advanced math

courses. Similarly, those in the stable-male (C1) trajectory were likelier to attend advanced

math courses. However, the educational pathways of those in the change trajectories are

different. Students following the male-to-female (C4) trajectory were less likely to take math

courses than the stable-male students and less likely to take advanced German courses than

the stable-female students. Meanwhile, those in the neutral-to-male (C5) trajectory were less

likely to take advanced German courses and more likely to take math courses than the stable-
GENDERED CAREER ASPIRATIONS 33

female students (C2). These results are likely because students in the neutral-to-male

trajectory have the highest math self-concept and opt for a male-dominated career later than

those on the stable-male trajectory.

Students’ gendered career aspiration trajectories also predict the university major

programs they apply to and study. Notably, the gender ratio of university majors was similar

for those in the neutral-to-male (C5) and stable-male (C1) trajectory, and there was a

significantly higher proportion of women in the majors of stable-female (C2), stable-neutral

(C3), and male-to-female (C4) trajectories. Career aspiration trajectories reflect both the

gender ratios in application and attendance of university programs. Moreover, students in all

profiles chose university majors aligned with their career aspirations in 12th grade, even if

that aspiration differed from 9th grade. Changes in career aspirations during secondary school

meaningfully reflected later career training activities.

Although other studies have not compared gender-(a)typical career aspirations with

math and verbal self-concept of ability, our results are consistent with findings that students

who maintained a higher level of math self-concept relative to verbal self-concept took more

math and advanced math courses in high school and worked in more STEM careers as adults

(Gaspard et al., 2020). Similarly, Wang et al. (2013) showed that students’ high math self-

concept of ability relative to their verbal self-concept was predictive of working in a STEM

field. In addition, the gender typicality of career aspirations is closely related to the types of

advanced courses students took—i.e., in the math or verbal domains—and gender-(a)typical

choices in university majors. Moreover, by focusing on the gender typicality of career

aspirations, we could directly examine specific developmental patterns (i.e., heterogenous

developmental trajectories on a continuous rather than binary variable) and how those

patterns led to university major choices consistent with the changed aspiration. Such changes
GENDERED CAREER ASPIRATIONS 34

represent individually different, meaningful decisions relating to math and verbal self-

concept of ability and achievement and future educational and career-training choices.

Limitations and Future Work

A particular strength of this study was our ability to examine heterogeneous

trajectories in individuals’ gender-(a)typical career preferences and academic development

using a large, nationally representative sample. However, there are a few notable limitations

that should be addressed in future work. First, the study did not look at the period before 9th

grade, when some developments in career aspirations are likely to occur. Second, we only

focused on academic track participants. In both cases, this allowed for a direct comparison of

career aspirations to major selection in university. However, as more data is collected in

NEPS, it may be possible to relate career aspirations in secondary school and, more broadly,

to other career training activities (e.g., vocational school programs or apprenticeships). Doing

so might also allow future researchers to find predicted effects based on differences in SES.

Third, this study examined the gender typicality of career aspirations. Other work can expand

on this by looking at other areas of career segregation and identity development, such as

ethnicity. Finally, this study focused on the self-concept of ability and achievement, but the

available data did not allow us to examine task values. As a result, we know relatively little

about how task values and other important motivational beliefs may differ across the

identified trajectories. Students on different trajectories may differ in beliefs about gender

and what sort of job qualities (e.g., competitive vs. cooperative) participants are looking for.

Conclusion

Gendered career aspirations reflect a dynamic developmental process with individual

differences. Over half of the adolescents in our representative sample aspired to a highly

gendered career trajectory by the end of high school. Although for the majority of

adolescents, the gender typicality of their career aspirations did not change over secondary
GENDERED CAREER ASPIRATIONS 35

school, it did for a significant minority of students, and this change was linked to their

academic development. While most students aspired to gender-typical careers, some aspired

to a career dominated by the opposite gender. By examining this significant minority, we

identified distinct motivational processes, including content-specific changes in math and

verbal achievement and self-concept, as well as corresponding differences in high-school

course enrollment and preferences for gender-(a)typical college major selection. Higher math

self-concept was related to maintaining a constant preference for or changing to a male-

dominated career. In contrast, actual math achievement, as measured by achievement tests

and grades, was less predictive of types of or changes in career aspirations.

Boys’ career aspiration trajectory predicted changes from 9th to 12th grade in math

self-concept, math achievement, reading achievement, and language self-concept. Girls

showed no consistent changes in domain-specific self-concept or ability across career

aspiration trajectories, with one exception. Changes towards a male-dominated career

corresponded with increased math achievement, likely as girls shifted their study choices

correspondingly. Gender remains a strong predictor of gendered career aspirations, but

individual differences and the relations to self-concept show diverse developmental

trajectories for adolescents’ gendered career aspirations.

Acknowledgment

This paper uses data from the National Educational Panel Study (NEPS; see Blossfeld

& Roßbach, 2019). The NEPS is carried out by the Leibniz Institute for Educational

Trajectories (LIfBi, Germany) in cooperation with a nationwide network. The second author

gratefully acknowledges funding by the Deutsche Forschungsgemeinschaft (German

Research Foundation) under Germany’s Excellence Strategy–EXC2126/1-390838866.

GENDERED CAREER ASPIRATIONS 36

References

Asparouhov, T., & Muthén, B. O. (2022). Multiple Imputation with Mplus.

https://www.statmodel.com/download/Imputations7.pdf

Basler, A., Kriesi, I., & Imdorf, C. (2021). The development of gendered occupational

aspirations across adolescence: examining the role of different types of upper-

secondary education. Longitudinal and Life Course Studies, 12(2), 173–199.

https://doi.org/10.1332/175795920X16015782777176

Beckmann, J., Wicht, A., & Siembab, M. (2023). Career Compromises and Dropout from

Vocational Education and Training in Germany. Social Forces, 102(2), 658–680.

https://doi.org/10.1093/sf/soad063

Blossfeld, H.‑P., Roßbach, H.‑G., & von Maurice, J. (Eds.). (2019). Education as a Lifelong

Process – The German National Educational Panel Study (NEPS) (2nd). Springer,

VS.

Breda, T., Grenet, J., Monnet, M., & Van Effenterre, C. (2023). How Effective are Female

Role Models in Steering Girls Towards STEM? Evidence from French High Schools.

The Economic Journal, 133(653), 1773–1809. https://doi.org/10.1093/ej/uead019

Breda, T., & Napp, C. (2019). Girls' comparative advantage in reading can largely explain the

gender gap in math-related fields. Proceedings of the National Academy of Sciences

of the United States of America, 116(31), 15435–15440.

https://doi.org/10.1073/pnas.1905779116

Brown, C., & Putwain, D. W. (2021). Socio-economic status, gender and achievement: the

mediating role of expectancy and subjective task value. Educational Psychology, 1–

19. https://doi.org/10.1080/01443410.2021.1985083
GENDERED CAREER ASPIRATIONS 37

Cheryan, S., Ziegler, S. A., Montoya, A. K., & Jiang, L. (2017). Why are some STEM fields

more gender balanced than others? Psychological Bulletin, 143(1), 1–35.

https://doi.org/10.1037/bul0000052

Chow, A., Eccles, J. S., & Salmela-Aro, K. (2012). Task value profiles across subjects and

aspirations to physical and IT-related sciences in the United States and Finland.

Developmental Psychology, 48(6), 1612–1628. https://doi.org/10.1037/a0030194

Denissen, J. J. A., Zarrett, N. R., & Eccles, J. S. (2007). I like to do it, I'm able, and I know I

am: Longitudinal couplings between domain-specific achievement, self-concept, and

interest. Child Development, 78(2), 430–447. https://doi.org/10.1111/j.1467-

8624.2007.01007.x

Dinella, L. M., Fulcher, M., & Weisgram, E. S. (2014). Sex-typed personality traits and

gender identity as predictors of young adults' career interests. Archives of Sexual

Behavior, 43(3), 493–504. https://doi.org/10.1007/s10508-013-0234-6

Duchhardt, C., & Gerdes, A. (2013). NEPS Technical Report for Mathematics: Scaling

Results of Starting Cohort 4 in Ninth Grade (NEPS Working Papers No. 22). National

Education Panel Study (NEPS). https://www.neps-

data.de/Portals/0/Working%20Papers/WP_XXII.pdf

Eccles, J. S. (1983). Expectancies, values and academic behaviors. In J. T. Spence (Ed.), A

series of books in psychology. Achievement and achievement motives: Psychological

and sociological approaches. W.H. Freeman.

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. Educational Psychologist, 44(2), 78–89.

https://doi.org/10.1080/00461520902832368

Eccles, J. S., & Wigfield, A. (2020). From expectancy-value theory to situated expectancy-

value theory: A developmental, social cognitive, and sociocultural perspective on

GENDERED CAREER ASPIRATIONS 38

motivation. Contemporary Educational Psychology, 61, 101859.

https://doi.org/10.1016/j.cedpsych.2020.101859

Eccles, J. S., Wigfield, A., Harold, R. D., & Blumenfeld, P. (1993). Age and Gender

Differences in Children's Self- and Task Perceptions during Elementary School. Child

Development, 64(3), 830–847. https://doi.org/10.1111/j.1467-8624.1993.tb02946.x

Enders, C. K., & Tofighi, D. (2008). The Impact of Misspecifying Class-Specific Residual

Variances in Growth Mixture Models. Structural Equation Modeling: A

Multidisciplinary Journal, 15(1), 75–95. https://doi.org/10.1080/10705510701758281

Federal Ministry for Education and Research. (2021). Students in Major programs, ISCED-

education reports and gender. https://www.datenportal.bmbf.de/portal/en/Table-

2.5.21.html

Federal Statistics Office of Germany. (2018). Employees subject to social insurance

contributions at their place of work: Länder, reference date, sex. https://www-

genesis.destatis.de/genesis/online?operation=previous&levelindex=3&step=2&titel=T

able+structure&levelid=1716498048092&levelid=1716497983757#abreadcrumb

Gartzia, L., Morgenroth, T., Ryan, M. K., & Peters, K. (2021). Testing the motivational

effects of attainable role models: Field and experimental evidence. Journal of

Theoretical Social Psychology, 5(4), 591–602. https://doi.org/10.1002/jts5.121

Gaspard, H., Lauermann, F., Rose, N., Wigfield, A., & Eccles, J. S. (2020). Cross-Domain

Trajectories of Students' Ability Self-Concepts and Intrinsic Values in Math and

Language Arts. Child Development, 1800–1818. https://doi.org/10.1111/cdev.13343

Gottfredson, L. S. (1981). Circumscription and compromise: A developmental theory of

occupational aspirations. Journal of Counseling Psychology, 28(6), 545–579.

https://doi.org/10.1037/0022-0167.28.6.545
GENDERED CAREER ASPIRATIONS 39

Gottfredson, L. S. (1996). Gottfredson's Theory of Circumscription and Compromise. In D.

Brown & L. Brooks (Eds.), Career Choice and Development (3rd, pp. 179–232).

Jossey-Bass.

Gottfredson, L. S., & Lapan, R. T. (1997). Assessing Gender-Based Circumscription of

Occupational Aspirations. Journal of Career Assessment, 5(4), 419–441.

https://doi.org/10.1177/106907279700500404

Guo, J., Marsh, H. W., Morin, A. J. S., Parker, P. D., & Kaur, G. (2015). Directionality of the

Associations of High School Expectancy-Value, Aspirations, and Attainment.

American Educational Research Journal, 52(2), 371–402.

https://doi.org/10.3102/0002831214565786

Guo, J., Marsh, H. W., Parker, P. D., Morin, A. J., & Dicke, T. (2017). Extending

expectancy-value theory predictions of achievement and aspirations in science:

Dimensional comparison processes and expectancy-by-value interactions. Learning

and Instruction, 49, 81–91. https://doi.org/10.1016/j.learninstruc.2016.12.007

Haberkorn, K., & Pohl, S. (2013). Cognitive Basic Skills (Non-Verbal) Data in the Scientific

Use File. https://www.neps-

data.de/Portals/0/NEPS/Datenzentrum/Forschungsdaten/SC3/com_cogbasic2013_en.p

df

Haberkorn, K., Pohl, S., Hardt, K., & Wiegand, E. (2012). NEPS Technical Report for

Reading: Scaling Results of Starting Cohort 4 in Ninth Grade (NEPS Working Papers

No. 16). National Education Panel Study (NEPS). https://www.neps-

data.de/Portals/0/Working%20Papers/WP_XVI.pdf

Jones, L. K., & Hite, R. L. (2021). A global comparison of the circumscription and

compromise theory of career development in science career aspirations. School

Science and Mathematics, 121(7), 381–394. https://doi.org/10.1111/ssm.12492

GENDERED CAREER ASPIRATIONS 40

Kang, H., Calabrese Barton, A., Tan, E., D. Simpkins, S., Rhee, H., & Turner, C. (2019).

How do middle school girls of color develop STEM identities? Middle school girls’

participation in science activities and identification with STEM careers. Science

Education, 103(2), 418–439. https://doi.org/10.1002/sce.21492

Kirkham, J., & Chapman, E. (2021). Gender, achievement level and sociocultural factors in

the mathematics course choices of Year 10 students in Western Australia. The

Australian Educational Researcher. Advance online publication.

https://doi.org/10.1007/s13384-020-00425-5

Lauermann, F., Chow, A., & Eccles, J. S. (2015). Differential Effects of Adolescents’

Expectancy and Value Beliefs about Math and English on Math/Science-Related and

Human-Services-Related Career Plans. International Journal of Gender, Science and

Technology, 7(2), 205–228.

http://genderandset.open.ac.uk/index.php/genderandset/article/view/393/674

Lauermann, F., Tsai, Y.‑M., & Eccles, J. S. (2017). Math-related career aspirations and

choices within Eccles et al.'s expectancy-value theory of achievement-related

behaviors. Developmental Psychology, 53(8), 1540–1559.

https://doi.org/10.1037/dev0000367

Lawson, K. M., Lee, B., Crouter, A. C., & McHale, S. M. (2018). Correlates of gendered

vocational development from middle childhood to young adulthood. Journal of

Vocational Behavior, 107, 209–221. https://doi.org/10.1016/j.jvb.2018.05.002

Lepper, C., Stang, J., & McElvany, N. (2022). Gender Differences in Text‐Based Interest:

Text Characteristics as Underlying Variables. Reading Research Quarterly, 57(2),

537–554. https://doi.org/10.1002/rrq.420
GENDERED CAREER ASPIRATIONS 41

Leroy, N., Max, S., & Pansu, P. (2022). Is Emma or Liam the Top Scorer in Math? The

Effects of a Counter-Stereotypical Role Model on Math Achievement. Sex Roles,

86(9-10), 587–603. https://doi.org/10.1007/s11199-022-01289-y

Marsh, H. W., van Zanden, B., Parker, P. D., Guo, J., Conigrave, J., & Seaton, M. (2019).

Young Women Face Disadvantage to Enrollment in University STEM Coursework

Regardless of Prior Achievement and Attitudes. American Educational Research

Journal, 56(5), 1629–1680. https://doi.org/10.3102/0002831218824111

McGeown, S. P., Osborne, C., Warhurst, A., Norgate, R., & Duncan, L. G. (2016).

Understanding children's reading activities: Reading motivation, skill and child

characteristics as predictors. Journal of Research in Reading, 39(1), 109–125.

https://doi.org/10.1111/1467-9817.12060

McNeish, D., & Harring, J. R. (2017). The Effect of Model Misspecification on Growth

Mixture Model Class Enumeration. Journal of Classification, 34(2), 223–248.

https://doi.org/10.1007/s00357-017-9233-y

McNeish, D., Harring, J. R., & Bauer, D. J. (2022). Nonconvergence, covariance constraints,

and class enumeration in growth mixture models. Psychological Methods. Advance

online publication. https://doi.org/10.1037/met0000456

Miyamoto, A., & Wicht, A. (2020). Developmental trajectories of the socioeconomic status

of occupational aspirations during adolescence. Journal of Adolescence, 26–35.

https://doi.org/10.1016/j.adolescence.2020.08.001

Morgenroth, T., Ryan, M. K., & Peters, K. (2015). The Motivational Theory of Role

Modeling: How Role Models Influence Role Aspirants’ Goals. Review of General

Psychology, 19(4), 465–483. https://doi.org/10.1037/gpr0000059

GENDERED CAREER ASPIRATIONS 42

Mueller, C. E., & Winsor, D. L. (2016). Math and verbal academic self-concept: Subject-

specificity across four distinctive groups of high ability adolescents. Learning and

Individual Differences, 50, 240–245. https://doi.org/10.1016/j.lindif.2016.08.002

Muthén, L. K., & Muthén, B. O. (1998-2017). Mplus User’s Guide: Eighth Edition. Muthén

& Muthén.

Parker, P. D., Nagy, G., Trautwein, U., & Lüdtke, O. (2014). Predicting career aspirations

and university majors from academic ability and self-concept. In I. Schoon, J. S.

Eccles, I. Schoon, & J. S. Eccles (Eds.), Gender Differences in Aspirations and

Attainment (pp. 224–246). Cambridge University Press.

https://doi.org/10.1017/CBO9781139128933.015

Parker, P. D., Schoon, I., Tsai, Y.‑M., Nagy, G., Trautwein, U., & Eccles, J. S. (2012).

Achievement, agency, gender, and socioeconomic background as predictors of

postschool choices: A multicontext study. Developmental Psychology, 1629–1642.

https://doi.org/10.1037/a0029167

Parker, P. D., van Zanden, B., Marsh, H. W., Owen, K., Duineveld, J. J., & Noetel, M.

(2020). The Intersection of Gender, Social Class, and Cultural Context: a Meta-

Analysis. Educational Psychology Review, 32(1), 197–228.

https://doi.org/10.1007/s10648-019-09493-1

Parker, P. D., van Zanden, B., & Parker, R. B. (2018). Girls get smart, boys get smug:

Historical changes in gender differences in math, literacy, and academic social

comparison and achievement. Learning and Instruction, 54, 125–137.

https://doi.org/10.1016/j.learninstruc.2017.09.002

Ram, N., & Grimm, K. J. (2009). Growth Mixture Modeling: A Method for Identifying

Differences in Longitudinal Change Among Unobserved Groups. International

GENDERED CAREER ASPIRATIONS 43

Journal of Behavioral Development, 33(6), 565–576.

https://doi.org/10.1177/0165025409343765

Ramos Carranza, P., & Simpkins, S. D. (2021). Parent and Sibling Science Support for

Latinx Adolescents. Social Psychology of Education, 24(2), 511–535.

https://doi.org/10.1007/s11218-021-09620-3

Sahin, A., Ekmekci, A., & Waxman, H. C. (2018). Collective Effects of Individual,

Behavioral, and Contextual Factors on High School Students’ Future STEM Career

Plans. International Journal of Science and Mathematics Education, 16(S1), 69–89.

https://doi.org/10.1007/s10763-017-9847-x

Sijbrandij, J. J., Hoekstra, T., Almansa, J., Peeters, M., Bültmann, U., & Reijneveld, S. A.

(2020). Variance constraints strongly influenced model performance in growth

mixture modeling: A simulation and empirical study. BMC Medical Research

Methodology, 20(1), 276. https://doi.org/10.1186/s12874-020-01154-0

Starr, C. R., Gao, Y., Lee, G., Safavian, N., Rubach, C., Dicke, A.‑L., Eccles, J. S., &

Simpkins, S. D. (2022). Parents’ Math Gender Stereotypes and Their Correlates: An

Examination of the Similarities and Differences Over the Past 25 Years. Sex Roles,

87(11-12), 603–619. https://doi.org/10.1007/s11199-022-01337-7

U.S. Census Bureau. (2019). STEM and STEM-Related Occupations by Sex and Median

Earnings: ACS 2019. https://www.census.gov/data/tables/time-series/demo/income-

poverty/stem-occ-sex-med-earnings.html

Wan, S., Lauermann, F., Bailey, D. H., & Eccles, J. S. (2023). Girls' comparative advantage

in language arts explains little of the gender gap in math-related fields: A replication

and extension. Proceedings of the National Academy of Sciences of the United States

of America, 120(40), e2305629120. https://doi.org/10.1073/pnas.2305629120

GENDERED CAREER ASPIRATIONS 44

Wan, Z. H. (2021). Exploring the Effects of Intrinsic Motive, Utilitarian Motive, and Self-

Efficacy on Students’ Science Learning in the Classroom Using the Expectancy-

Value Theory. Research in Science Education, 51(3), 647–659.

https://doi.org/10.1007/s11165-018-9811-y

Wang, M.‑T., Eccles, J. S., & Kenny, S. (2013). Not lack of ability but more choice:

Individual and gender differences in choice of careers in science, technology,

engineering, and mathematics. Psychological Science, 24(5), 770–775.

https://doi.org/10.1177/0956797612458937

Wille, E., Stoll, G., Gfrörer, T., Cambria, J., Nagengast, B., & Trautwein, U. (2020). It Takes

Two: Expectancy-Value Constructs and Vocational Interests Jointly Predict STEM

Major Choices. Contemporary Educational Psychology, 61, 101858.

https://doi.org/10.1016/j.cedpsych.2020.101858

Wolter, I., Ehrtmann, L., Seidel, T., & Drechsel, B. (2019). Social or Economic Goals? The

Professional Goal Orientation of Students Enrolled in STEM and Non-STEM Majors

in University. Frontiers in Psychology, 10, 2065.

https://doi.org/10.3389/fpsyg.2019.02065

Yoder, B. L. (2014). Engineering by the Numbers. 2014 ASEE Profiles of Engineering and

Engineering Technology Colleges. https://monolith.asee.org/papers-and-

publications/publications/college-profiles/14EngineeringbytheNumbersPart1.pdf

Zhang, H., Couch, S., Estabrooks, L., Perry, A., & Kalainoff, M. (2023). Role models’

influence on student interest in and awareness of career opportunities in life sciences.

International Journal of Science Education, Part B, 13(4), 381–399.

https://doi.org/10.1080/21548455.2023.2180333
GENDERED CAREER ASPIRATIONS 45

Table 1

Correlation Matrix of All Variables

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22
1. Gender-Female -
2. SES -.01* -
3. Nonverbal Reasoning -.03*** .19*** -
4. 9th Grade Math Achievement -.17*** .29*** .48*** -
5. 12th Grade Math Achievement -.27*** .22*** .38*** .69*** -
6. 9th Grade Verbal Achievement .12*** .26*** .44*** .52*** .36*** -
7. 12th Grade Verbal Achievement .10*** .23*** .35*** .49*** .43*** .59*** -
8. 9th Grade Math Self-Concept -.25*** .08*** .23*** .37*** .42*** .05*** .09*** -
9. 12th Grade Math Self-Concept -.19*** .05*** .21*** .39*** .46*** .05*** .12*** .67*** -
10. 9th Grade Verbal Self-Concept .18*** .09*** .02*** .06*** .01 .20*** .20*** -.07*** -.07*** -
11. 12th Grade German Self-Concept .24*** .07*** -.04*** -.05*** -.10*** .19*** .18*** -.12*** -.13*** .48*** -
12. 9th Grade Math Grade -.06*** .14*** .25*** .39*** .39*** .20*** .22*** .66*** .52*** .06*** .04*** -
13. 12th Grade Math Grade -.05*** .10*** .18*** .34*** .39*** .15*** .23*** .46*** .72*** .04*** .04*** .48*** -
14. 9th Grade German Grade .21*** .17*** .13*** .23*** .16*** .32*** .32*** .06*** .06*** .57*** .38*** .37*** .21*** -
15. 12th Grade German Grade .19*** .12*** .07*** .13*** .11*** .26*** .27*** .05*** .07*** .33*** .60*** .22*** .42*** .43*** -
16. Math as an Advanced Course -.14*** .00 .12*** .27*** .32*** .05*** .08*** .31*** .32*** -.09*** -.16*** .22*** .20*** .01 .00 -
17. German as an Advanced Course .17*** -.01 -.04*** -.09*** -.12*** .03* .04* -.18*** -.20*** .11*** .18*** -.10*** -.11*** .13*** .09*** .15*** -
18. Gender Ratio in Major Studied .38*** .03** -.03** -.16*** -.26*** .11*** .06*** -.28*** -.32*** .21*** .31*** -.14*** -.17*** .16*** .20*** -.23*** .15*** -
19. Gender Ratio in Major Applied .43*** .05*** -.07*** -.20*** -.25*** .10*** .08*** -.23*** -.27*** .21*** .31*** -.08*** -.12*** .18*** .24*** -.25*** .13*** .73*** -
20. 9th Grade Career Aspiration Ratio .53*** -.04*** -.05*** -.13*** -.20*** .05*** .05*** -.19*** -.16*** .12*** .17*** -.09*** -.07*** .10*** .10*** -.14*** .10*** .35*** .32*** -
21. 10th Grade Career Aspiration Ratio .51*** -.03*** -.06*** -.15*** -.19*** .05*** .06*** -.19*** -.16*** .12*** .15*** -.09*** -.06*** .10*** .10*** -.14*** .09*** .37*** .39*** .64*** -
22. 11th Grade Career Aspiration Ratio .45*** -.04*** -.07*** -.18*** -.20*** .03*** .04*** -.17*** -.18*** .09*** .16*** -.08*** -.07*** .07*** .11*** -.13*** .08*** .42*** .40*** .52*** .63*** -
23. 12th Grade Career Aspiration Ratio .44*** -.04*** -.08*** -.21*** -.25*** .00 .01 -.20*** -.20*** .11*** .17*** -.10*** -.10*** .08*** .08*** -.16*** .10*** .42*** .45*** .46*** .54*** .62***
Note. All ratios are in terms of the proportion of women in their respective majors or aspired careers.
* p < .05
** p < .01
*** p < .001
GENDERED CAREER ASPIRATIONS 46

Table 2

Model Selection
# of Size of Smallest
AIC BIC Adjusted BIC Entropy
Classes Class
1 123,691 123,763 123,728 n.a. n.a.
2 115,915 115,992 115,954 0.647 1207
3 115,407 115,504 115,457 0.735 964
4 115,283 115,400 115,342 0.740 231
5 115,186 115,322 115,255 0.697 280
Note. AIC refers to adjusted information criteria. BIC refers to Bayesian information criteria.
Models above six classes produced invalid results and were excluded from consideration.
GENDERED CAREER ASPIRATIONS 47

Table 3

Most Common Career Aspirations by Trajectory

Grade 9 Grade 12
Percent Percent
Females Females
Aspiration (% percent of Trajectory) in field Aspiration (% percent of Trajectory) in field

Stable-Male (C1)
Aircraft Pilots and Related Associate 7% Managing Directors and Chief Executives
Professionals (11%) (7%) 22%
Athletes and Sports Players (10%) 9% Athletes and Sports Players (6%) 9%
Managing Directors and Chief Executives 22% Aircraft Pilots and Related Associate
(10%) Professionals (5%) 7%
Systems Analysts (5%) 16% Telecommunications Engineers (4%) 12%
Mechanical Engineers (3%) 14% Information and Communications
Technology User Support Technicians
(3%) 16%

Stable-Female (C2)
Secondary Education Teachers (12%) 69% Psychologists (6%) 80%
Psychologists (11%) 80% Secondary Education Teachers (6%) 69%
Veterinarians (6%) 75% Generalist Medical Practitioners (4%) 56%
Building Architects (5%) 72% Primary School Teachers (4%) 83%
Product and Garment Designers (5%) 85% Social Work and Counselling
Professionals (2%) 75%

Stable-Neutral (C3)
Generalist Medical Practitioners (10%) 56% Generalist Medical Practitioners (7%) 56%
Building Architects (6%) 50% Psychologists (3%) 80%
Journalists (5%) 52% Journalists (2%) 52%
Actors (5%) 45% Building Architects (2%) 50%
Lawyers (4%) 47% Police Officers (2%) 50%

Male-to-Female (C4)
Managing Directors and Chief Executives 22% Generalist Medical Practitioners (7%)
(22%) 56%
Athletes and Sports Players (13%) 9% Psychologists (5%) 80%
Aircraft Pilots and Related Associate 7% Secondary Education Teachers (3%)
Professionals (10%) 69%
Specialist Medical Practitioners (7%) 32% Specialist Medical Practitioners (3%) 58%
Chemists (4%) 29% Product and Garment Designers (2%) 85%

Neutral-to-Male (C5)
Generalist Medical Practitioners (9%) 56% Managing Directors and Chief Executives
(20%) 22%
Building Architects (9%) 50% Athletes and Sports Players (8%) 9%
Lawyers (5%) 47% Telecommunications Engineers (6%) 12%
Secondary Education Teachers (5%) 69% Aircraft Pilots and Related Associate
Professionals (6%) 7%
Actors (5%) 45% Mechanical Engineers (4%) 14%
Note. Career aspiration names come from the English-language description of the ISCO 08 code for the
profession. Percent females in field is the percentage of women in that field based on German census data used
in this study. Online Supplement S3 provides the ten most common aspirations across all measurement points
(Grade 9 to Grade 12).
GENDERED CAREER ASPIRATIONS 48

Table 4

Odds Ratios for Predictor Variables with Stable-Male Trajectory as Reference

Trajectory
C2: Stable-Female C3: Stable- C4: Male-to- C5: Neutral-to-
Neutral Female Male

M (SE) M (SE) M (SE) M (SE)

Gender (Female) 81.62 (24.62)*** 19.47 (5.14)*** 8.80 (2.87)*** 2.68 (1.01)**
SES 1.00 (0.01) 1.00 (0.01) 1.01 (0.01) 1.00 (0.01)
Nonverbal Reasoning 0.99 (0.04) 1.06 (0.04) 1.10 (0.08) 1.07 (0.06)
Math Grades
9th Grade 1.00 (0.13) 0.97 (0.10) 1.07 (0.20) 1.09 (0.15)
12th Grade 1.03 (0.04) 1.03 (0.04) 1.00 (0.06) 0.94 (0.05)
German Grades
9th Grade 1.33 (0.11)* 1.22 (0.10) 1.32 (0.15) 1.00 (0.19)
12th Grade 1.00 (0.05) 1.03 (0.04) 1.00 (0.06) 1.02 (0.06)
Math Test Scores
9th Grade 0.99 (0.09) 1.03 (0.08) 0.97 (0.12) 0.77 (0.09)*
12th Grade 0.80 (0.10)x 0.78 (0.08)* 0.88 (0.12) 0.99 (0.13)
Verbal Test Scores
9th Grade 1.00 (0.10) 1.02 (0.09) 1.01 (0.13) 1.00 (0.12)
12th Grade 1.20 (0.14) 1.15 (0.12) 1.16 (0.20) 1.12 (0.17)
Math Self-Concept
9th Grade 0.74 (0.10)* 0.79 (0.09)* 0.91 (0.19) 0.95 (0.17)
12th Grade 0.71 (0.11)* 0.71 (0.10)* 0.60 (0.13)* 1.30 (0.25)
Verbal Self-Concept
9th Grade 1.35 (0.22)x 1.23 (0.18) 1.00 (0.25) 0.98 (0.19)
12th Grade 1.40 (0.24)x 1.55 (0.23)** 1.52 (0.37) 1.13 (0.24)
Note.
x
p < 0.10
* p < 0.05
** p < 0.01
*** p < 0.001
GENDERED CAREER ASPIRATIONS 49

Table 5

Selected Standardized Group Path Analysis Loadings (C1: Stable-Male as Reference)

Regression Path Male Female
Participants Participants
β (SE) β (SE)
12th-Grade Math Achievement on
C2: Stable-Female -0.06* (0.02) -0.02 (0.05)
C3: Stable-Neutral -0.06* (0.03) -0.03 (0.05)
C4: Male-to-Female -0.05* (0.03) -0.00 (0.03)
C5: Neutral-to-Male -0.03 (0.03) 0.08** (0.03)
12th-Grade Reading Achievement on
C2: Stable-Female -0.00 (0.03) -0.06 (0.06)
C3: Stable-Neutral 0.05 (0.03) -0.08 (0.06)
C4: Male-to-Female 0.01 (0.03) -0.07* (0.04)
C5: Neutral-to-Male 0.04 (0.03) 0.02 (0.03)
12th-Grade Math Self-Concept on
C2: Stable-Female -0.06* (0.03) -0.02 (0.04)
C3: Stable-Neutral -0.07** (0.03) -0.00 (0.04)
C4: Male-to-Female -0.07** (0.02) -0.04 (0.03)
C5: Neutral-to-Male 0.02 (0.03) 0.01 (0.02)
12th-Grade German Self-Concept on
C2: Stable-Female 0.13*** (0.03) 0.02 (0.06)
C3: Stable-Neutral 0.13*** (0.03) 0.05 (0.06)
C4: Male-to-Female 0.10*** (0.03) 0.04 (0.03)
C5: Neutral-to-Male -0.00 (0.03) -0.00 (0.04)
12th-Grade Math Grade on
C2: Stable-Female -0.03 (0.02) -0.01 (0.05)
C3: Stable-Neutral -0.07* (0.03) 0.05 (0.05)
C4: Male-to-Female -0.01 (0.03) -0.02 (0.03)
C5: Neutral-to-Male 0.01 (0.03) 0.03 (0.03)
12th-Grade German grade on
C2: Stable-Female 0.07* (0.03) 0.03 (0.06)
C3: Stable-Neutral 0.07* (0.03) 0.07 (0.06)
C4: Male-to-Female 0.05* (0.03) 0.02 (0.03)
C5: Neutral-to-Male 0.03 (0.03) -0.01 (0.03)
Note. Male and female participants were modeled as a 2-group SEM path analysis. For full
model results, please see the online supplementary materials.
* p < 0.05
** p < 0.01
*** p < 0.001
GENDERED CAREER ASPIRATIONS 50

Table 6

Education Choices by Career Trajectory

Proportion of Trajectory Attending Advanced Courses in Proportion of Women in

German Math Applied Major Enrolled Major
Trajectory M (SE) ΔC1 ΔC2 M (SE) ΔC1 ΔC2 M (SE) ΔC1 ΔC2 M (SE) ΔC1 ΔC2
C1: Stable-Male .45 (.05) .71 (.03) .33 (.02) .33 (.01)
C2: Stable-Female .58 (.03) .13** .38 (.03) -.33*** .60 (.01) .27*** .60 (.01) .27***
C3: Stable-Neutral .60 (.02) .15*** .02 .44 (.03) -.27*** .06 .56 (.01) .23*** -.04** .55 (.01) .22*** -.05***
C4: Male-to-Female .49 (.04) .04 -.09** .42 (.04) -.29*** .04 .55 (.02) .22*** -.05* .55 (.03) .22*** -.05*
C5: Neutral-to-Male .42 (.05) -.03 -.16** .58 (.04) -.13** .20** .33 (.03) .00 -.27** .37 (.03) .04 -.23***
Note. ΔC1 is the relative difference between trajectories C2 through C5 versus C1 (the stable-male trajectory). ΔC2 is the relative difference
between trajectories C3 through C5 versus C2 (the stable-female trajectory).
* p < 0.05
** p < 0.01
*** p < 0.001
 GENDERED CAREER ASPIRATIONS 51

Figures

Figure 1
Gender Typicality of Career Aspirations: Five Different Trajectories in Growth Mixture
Model Results

Note. This figure shows the mean trajectory of each of the five identified classes in our
growth-mixture analyses (N = 4,759). The average probability for each student to follow their
assigned trajectory (e.g., C1), relative to other trajectories (e.g., C2 – C5) was high for all five
trajectories (C1 = .79; C2 = .86; C3 = .80; C4 = 80, C5 = .76).
 GENDERED CAREER ASPIRATIONS 52

Figure 2
Math and Verbal Achievement and Self-Concept by Grade, Gender, and Trajectory

Note. C1 refers to the Stable-Male trajectory¸C2 refers to the Stable-Female trajectory, C3

refers to the Stable-Neutral trajectory, C4 refers to the Male-to-Female trajectory, and C5
refers to the Neutral-to-Male trajectory. The error bars show standard errors.
 GENDERED CAREER ASPIRATIONS 53

Supplementary Materials: Exploring the Links between Gender-(A)typical Career

Aspirations and Educational Choices: Heterogeneous Developmental Pathways

Supplement S1. Sample Summary: Observed and Imputed Data.

Supplement S2. Mplus Syntaxes
Supplement S3. Most Common Career Aspirations at All Measurement Points
Supplement S4. Full Results from Path Analysis
 GENDERED CAREER ASPIRATIONS 54

Supplement S1. Sample Summary: Observed and Imputed Data

Observed Cases Mean/Percent Mean

(Observed) (Imputed)
Proportion of female participants 4,759 55.5% n.a.

ISEI (SES) 3,127 58.49 57.84

Nonverbal Reasoning 4,483 9.87 9.86
Yearly grades (9th grade)
Math 4,513 2.64 2.82
Verbal 4,517 2.81 2.66
Final Grades (12th grade)
Math 3,460 9.60 8.95
Verbal 3,461 9.16 9.44
Test Scores (9th grade)
Math 4,609 0.93 0.92
Verbal 4,478 0.80 0.79
Test Scores(12th grade)
Math 3,350 1.40 1.34
Verbal 3,367 0.97 0.92
Self-concept (9th grade)
Math 4,506 2.56 2.55
Verbal 4,523 3.02 3.01
Self-concept (12th grade)
Math 3,537 2.50 2.45
Verbal 3,542 2.98 2.95
Note. ISEI refers to the International Socioeconomic Index. We use only the parent with the
higher value. Gender was not imputed because there were no missing data for gender. Math
and German grades in 9th grade range from 1 to 6, with 1 being the best. Math and German
grades in 12th grade range from 1 to 15, with 15 being the best.
 GENDERED CAREER ASPIRATIONS 55

Supplement S2. Mplus Syntaxes

Analysis 1: Growth Mixture Model

This is an example of the syntax for the growth mixture model with five classes.

TITLE: Analysis 1: Growth Mixture Model

DATA: FILE = "Analysis1.dat";
VARIABLE:
NAMES = IDt IJGen9 RJGen9 IJGen10 RJGen10 IJGen11 RJGen11
IJGen12 RJGen12 IJGen13 RJGen13 IJGen15 RJGen15 IJGen14
RJGen14 IDi Gender Gym Sch3 Inc LogInc ISCED CASMIN ISEI
Reas9 M9WLE M9SE M12WLE M12SE R9WLE R9SE R12WLE R12SE
GSC9 MSC9 ASC9 GSC12 MSC12 ASC12 GG9 MG9 GG12 MG12;
usevariables are IJGen9 IJGen10 IJGen11 IJGen12 Female;
missing are .;
useobs are Gym==1;
cluster is IDi;
IDvariable is IDt;
Classes are c(5); !set the number of classes
auxiliary are ISEI(R3step)
M9WLE(R3step) R9WLE(R3step) M12WLE(R3step) R12WLE(R3step) MSC9(R3step)
GSC9(R3step) GSC12(R3step) MSC12(R3step) GG9(R3step) MG9(R3step)
GG12(R3step) MG12(R3step) Reas9(R3step) Female(R3step);

DEFINE:
IF gender==2 THEN Female = 1;
IF gender==1 THEN Female = 0;

MODEL:
%OVERALL%
iI sI | IJgen9@0 IJgen10@1 IJgen11@2 IJgen12@3;
[iI sI];

OUTPUT: stdyx;

SAVEDATA:
file is graphdata verify.txt;
results are results;

PLOT:
type=Plot3;
Series = IJgen9(9) IJgen10(10) IJgen11(11) IJgen12(12);

ANALYSIS:
TYPE is complex mixture;
starts are 250 100;
 GENDERED CAREER ASPIRATIONS 56

Analysis 2: Imputation and Auxliary Predictors

Below are the Mplus syntaxes needed for imputation and the analysis based on imputed
values.
TITLE: Mint Study and Degree on Class, PEd SC, Gender - Imputation run
DATA: FILE = "Analysis3.dat";
VARIABLE:
NAMES = IDt IJGen9 RealGen_2 IJGen10 RealGen_3 IJGen11 RealGen_5
IJGen12 RealGen_7 IJGen13 RealGen_8 IJGen14 RealGen_10 IJGen15
RealGen_9 IDi Gender Gym Sch3 Inc LogInc ISCED CASMIN ISEI Reas9 M9WLE
M9SE
M12WLE M12SE R9WLE R9SE R12WLE R12SE GSC9 MSC9 ASC9 GSC12 MSC12
ASC12
GG9 MG9 GG12 MG12
GenMajor ApplyGen;
usevariables are
Gender LogInc ISCED ISEI
Reas9 M9WLE M12WLE M12SE R9WLE R12WLE GSC9 MSC9
GSC12 MSC12 GG9 MG9 GG12 MG12 GenMajor ApplyGen;
MISSING=.;
cluster is IDi;
IDvariable is IDt;
Auxiliary are IJGen9 IJGen10 IJGen11
IJGen12 Gym;

DEFINE:
Data Imputation:
impute = LogInc ISEI ISCED
Reas9
M9WLE R9WLE
M12WLE R12WLE
M9WLE R9WLE
GSC9 MSC9 GSC12 MSC12
GG9 MG9 GG12 MG12
GenMajor ApplyGen
gender(c);

NDATASETS = 100;
SAVE = TESTimp*.dat;
THIN = 100;

ANALYSIS:
TYPE is twolevel; biter=(1000);
processors are 8;
 GENDERED CAREER ASPIRATIONS 57

TITLE: Use imputed predictors on GMM

DATA: FILE = "TESTimplist.dat"; type is imputation;
VARIABLE:
NAMES =
GENDER LOGINC ISCED ISEI REAS9 M9WLE M12WLE
M12SE R9WLE R12WLE GSC9 MSC9 GSC12 MSC12 GG9
MG9 GG12 MG12 GENMAJOR APPLYGEN IJGEN9 IJGEN10
IJGEN11 IJGEN12 GYM IDT IDI;
usevariables are
IJGEN9 IJGEN10 IJGEN11 IJGEN12 Female;
missing are *;
useobs are Gym==1;
cluster is IDi;
IDvariable is IDt;
Classes are c(5);
auxiliary are ISEI(R3step)
M9WLE(R3step) R9WLE(R3step)
M12WLE(R3step) R12WLE(R3step)
MSC9(R3step) GSC9(R3step)
GSC12(R3step) MSC12(R3step)
GG9(R3step) MG9(R3step)
GG12(R3step) MG12(R3step)
Reas9(R3step)
Female(R3step);

DEFINE:
IF gender==2 THEN Female = 1;
IF gender==1 THEN Female = 0;

MODEL:
%OVERALL%
iI sI | IJgen9@0 IJgen10@1 IJgen11@2 IJgen12@3;
[iI sI];

OUTPUT:
stdyx;

PLOT:
type=Plot3;
Series = IJgen9(9) IJgen10(10) IJgen11(11) IJgen12(12);

ANALYSIS:
TYPE is complex mixture;
starts are 250 100;
processors are 8;

SAVEDATA:
file is imputation summary.txt;
 GENDERED CAREER ASPIRATIONS 58

Analysis 2: Change Analysis

This is the syntax used to conduct the path analysis of changes in self-concept and
achievement.
TITLE: Classes Predicting Grades/Achievement/self-concept
DATA: FILE = "analysis2.dat";
VARIABLE:
NAMES are
IDt Reas9 M9WLE M9SE M12WLE M12SE R9WLE R9SE R12WLE R12SE GSC9
MSC9 ASC9 GSC12 MSC12 ASC12 IdealGen9 IdealGen10
IdealGen11 IdealGen12 LogInc ISEI female M12WLE_y Mdiff PEd Rdiff MSCdiff
GSCdiff ASCdiff GG9 MG9 GG12 MG12
I S CI CS Prob1 Prob2 Prob3 Prob4 Prob5 C IDi;
usevariables are
MSC9 MSC12 GSC9 GSC12 M9WLE M12WLE R9WLE R12WLE ISEI
GG9 MG9 GG12 MG12
C1 C2 C3 C4;
missing are .;
cluster is IDi;
IDvariable is IDt;
grouping is FEMALE(0=M,1=F);

DEFINE:
IF C==1 THEN C1=1 ELSE C1=0;
IF C==2 THEN C2=1 ELSE C2=0;
IF C==3 THEN C3=1 ELSE C3=0;
IF C==4 THEN C4=1 ELSE C4=0;
!IF C==5 THEN C5=1 ELSE C5=0;

MODEL:
!C5 as reference
R12WLE M12WLE MSC12 GSC12 GG12 MG12 on
R9WLE M9WLE MSC9 GSC9 GG9 MG9
ISEI
C1 C2 C3 C4;

OUTPUT: stdyx;

ANALYSIS:
TYPE is complex;
processors are 8;
 GENDERED CAREER ASPIRATIONS 59

Analysis 3: Education choices

This is the syntax used to calculate the relationship between mixture class membership and

the gender ratio of each student’s university major.

TITLE: Mint Study and Degree on Class, PEd SC, Gender

DATA: FILE = "analysis3.dat";
VARIABLE:
NAMES =IDt IJGen9 RJGen9 IJGen10 RJGen10 IJGen11 RJGen11
IJGen12 RJGen12 IJGen13 RJGen13 IJGen15 RJGen15 IJGen14
RJGen14 IDi Gender Gym Sch3 Inc LogInc ISCED CASMIN ISEI
Reas9 M9WLE M9SE M12WLE M12SE R9WLE R9SE R12WLE R12SE
GSC9 MSC9 ASC9
GSC12 MSC12 ASC12
GG9 MG9 GG12 MG12
GenMajor ApplyGen
AdvGer(BCH) AdvMath(BCH);
usevariables are IJGen9 IJGen10 IJGen11 IJGen12;

MISSING=.;
IDvariable is IDt;
cluster is IDi;
subpopulation are Gym==1;
Classes are c(5);
auxiliary are
GenMajor(BCH) ApplyGen(BCH)
AdvGer(BCH) AdvMath(BCH);

DEFINE:
IF gender==2 THEN Female = 1;
IF gender==1 THEN Female = 0;
IF ISCED>6 THEN PEd = 1;
IF ISCED<7 THEN PEd = 0;

MODEL:
%OVERALL%
iI sI | IJGen9@0 IJGen10@1 IJGen11@2 IJGen12@3;
[iI sI];

ANALYSIS:
TYPE is complex mixture;
starts are 250 100;
processors are 8;

PLOT:
type=Plot3;
Series = IJgen9(9) IJgen10(10) IJgen11(11) IJgen12(12);

SAVEDATA:
file is graphdata.txt;
 GENDERED CAREER ASPIRATIONS 60

Supplement S3. Most Common Career Aspirations by Trajectory at All Measurement Points

Grade 9 Grade 10 Grade 11 Grade 12

Percent Percent Percent Percent
Aspiration (Percent of Females in Aspiration (Percent of Females in Aspiration (Percent of Females in Aspiration (Percent of Females in
Trajectory) Field Trajectory) Field Trajectory) Field Trajectory) Field

Stable-Male (C1)
Aircraft Pilots and Related 7% Aircraft Pilots and Related 7% Aircraft Pilots and Related 7% Managing Directors and
Associate Professionals (11%) Associate Professionals (9%) Associate Professionals (8%) Chief Executives (7%) 22%
Athletes and Sports Players 9% Athletes and Sports Players 9% Managing Directors and 22% Athletes and Sports Players
(10%) (8%) Chief Executives (7%) (6%) 9%
Managing Directors and Chief 22% Managing Directors and 22% Athletes and Sports Players 9% Aircraft Pilots and Related
Executives (10%) Chief Executives (7%) (7%) Associate Professionals
(5%) 7%
Systems Analysts (5%) 16% Mechanical Engineers (4%) 14% Mechanical Engineers (4%) 14% Telecommunications
Engineers (4%) 12%
Mechanical Engineers (3%) 14% Systems Analysts (3%) 16% Systems Analysts (3%) 16% Information and
Communications
Technology User Support
Technicians (3%) 16%
Chemists (2%) 29% Chemists (2%) 29% Industrial and Production 16% Mechanical Engineers (2%)
Engineers (2%) 14%
Applications Programmers 14% Physicists and Astronomers 17% Physicists and Astronomers 17% Chemists (2%)
(2%) (1%) (2%) 29%
Physicists and Astronomers 17% Applications Programmers 14% Chemists (2%) 29% Industrial and Production
(1%) (1%) Engineers (2%) 16%
Physicists and Astronomers 15% Industrial and Production 16% Specialist Medical 32% Air Traffic Controllers
(1%) Engineers (1%) Practitioners (1%) (1%) 29%
Specialist Medical 32% Air Traffic Controllers (1%) 29% Mechanical Engineers (1%) 8% Physicists and Astronomers
Practitioners (1%) (1%) 17%

Stable-Female (C2)
Secondary Education 69% Psychologists (10%) 80% Psychologists (11%) 80% Psychologists (6%)
Teachers (12%) 80%
Psychologists (11%) 80% Secondary Education 69% Secondary Education 69% Secondary Education
Teachers (9%) Teachers (8%) Teachers (6%) 69%
Veterinarians (6%) 75% Primary School Teachers 83% Generalist Medical 56% Generalist Medical
(3%) Practitioners (4%) Practitioners (4%) 56%
Building Architects (5%) 72% Social Work and Counselling 75% Primary School Teachers 83% Primary School Teachers
Professionals (3%) (4%) (4%) 83%
Product and Garment 85% Veterinarians (3%) 75% Social Work and Counselling 75% Social Work and
Designers (5%) Professionals (3%) Counselling Professionals 75%
 GENDERED CAREER ASPIRATIONS 61

(2%)
Primary School Teachers 83% Generalist Medical 56% Veterinarians (2%) 75% Veterinarians (2%)
(4%) Practitioners (2%) 75%
Specialist Medical 78% Specialist Medical 78% Specialist Medical 78% Journalists (2%)
Practitioners (4%) Practitioners (2%) Practitioners (2%) 52%
Social Work and Counselling 75% Product and Garment 85% Social Work Associate 90% Specialist Medical
Professionals (3%) Designers (2%) Professionals (2%) Practitioners (1%) 78%
Social Work Associate 90% Social Work Associate 90% Lawyers (1%) 47% Social Work Associate
Professionals (2%) Professionals (2%) Professionals (1%) 90%
Translators, Interpreters and 68% Journalists (2%) 52% Building Architects (1%) 72% Building Architects (1%)
Other Linguists (2%) 50%

Stable-Neutral (C3)
Generalist Medical 56% Generalist Medical 56% Generalist Medical 56% Generalist Medical
Practitioners (10%) Practitioners (8%) Practitioners (8%) Practitioners (7%) 56%
Building Architects (6%) 50% Building Architects (4%) 50% Building Architects (4%) 50% Psychologists (3%) 80%
Journalists (5%) 52% Journalists (4%) 52% Journalists (3%) 52% Journalists (2%) 52%
Actors (5%) 45% Police Officers (3%) 50% Psychologists (3%) 80% Building Architects (2%) 50%
Lawyers (4%) 47% Actors (3%) 45% Police Officers (3%) 50% Police Officers (2%) 50%
Police Officers (3%) 50% Lawyers (3%) 47% Lawyers (2%) 47% Actors (2%) 45%
Product and Garment 36% Managing Directors and 22% Secondary Education 69% Lawyers (2%)
Designers (2%) Chief Executives (2%) Teachers (2%) 47%
Musicians, Singers and 39% Musicians, Singers and 39% Actors (2%) 45% Secondary Education
Composers (2%) Composers (2%) Teachers (2%) 69%
Photographers (2%) 63% Psychologists (2%) 80% Musicians, Singers and 39% Conference and Event
Composers (2%) Planners (2%) 58%
Specialist Medical 32% Product and Garment 36% Conference and Event 58% Authors and Related
Practitioners (2%) Designers (1%) Planners (1%) Writers (1%) 46%

Male-to-Female (C4)
Managing Directors and Chief 22% Managing Directors and 22% Generalist Medical 56% Generalist Medical
Executives (22%) Chief Executives (9%) Practitioners (6%) Practitioners (7%) 56%
Athletes and Sports Players 9% Specialist Medical 32% Psychologists (5%) 80% Psychologists (5%)
(13%) Practitioners (5%) 80%
Aircraft Pilots and Related 7% Athletes and Sports Players 9% Lawyers (3%) 47% Secondary Education
Associate Professionals (10%) (4%) Teachers (3%) 69%
Specialist Medical 32% Aircraft Pilots and Related 7% Building Architects (3%) 50% Specialist Medical
Practitioners (7%) Associate Professionals (3%) Practitioners (3%) 58%
Chemists (4%) 29% Generalist Medical 56% Specialist Medical 32% Product and Garment
Practitioners (3%) Practitioners (3%) Designers (2%) 85%
Mechanical Engineers (3%) 14% Building Architects (3%) 50% Managing Directors and 22% Conference and Event
Chief Executives (2%) Planners (2%) 58%
Systems Analysts (3%) 16% Sports, Recreation and 44% Secondary Education 69% Specialist Medical 91%
 GENDERED CAREER ASPIRATIONS 62

Cultural Centre Managers Teachers (2%) Practitioners (2%)

(2%)
Physicists and Astronomers 15% Lawyers (2%) 47% Sports, Recreation and 44% Building Architects (1%)
(2%) Cultural Centre Managers
(2%) 50%
Applications Programmers 14% Specialist Medical 58% Specialist Medical 58% Police Officers (1%)
(2%) Practitioners (2%) Practitioners (2%) 50%
Air Traffic Controllers (2%) 29% Psychologists (2%) 80% Journalists (2%) 52% Hotel Managers (1%) 65%

Neutral-to-Male (C5)
Generalist Medical 56% Managing Directors and 22% Managing Directors and 22% Managing Directors and
Practitioners (9%) Chief Executives (8%) Chief Executives (13%) Chief Executives (20%) 22%
Building Architects (9%) 50% Generalist Medical 56% Aircraft Pilots and Related 7% Athletes and Sports Players
Practitioners (6%) Associate Professionals (6%) (8%) 9%
Lawyers (5%) 47% Aircraft Pilots and Related 7% Athletes and Sports Players 9% Telecommunications
Associate Professionals (4%) (6%) Engineers (6%) 12%
Secondary Education 69% Athletes and Sports Players 9% Mechanical Engineers (3%) 14% Aircraft Pilots and Related
Teachers (5%) (4%) Associate Professionals
(6%) 7%
Actors (5%) 45% Building Architects (3%) 50% Systems Analysts (2%) 16% Mechanical Engineers (4%) 14%
Bank Tellers and Related 60% Lawyers (3%) 47% Generalist Medical 56% Industrial and Production
Clerks (3%) Practitioners (2%) Engineers (3%) 16%
Journalists (3%) 52% Mechanical Engineers (2%) 14% Physicists and Astronomers 17% Physicists and Astronomers
(2%) (2%) 17%
Police Officers (2%) 50% Actors (2%) 45% Legislators (2%) 18% Legislators (2%) 18%
University and Higher 44% Legislators (2%) 18% Journalists (2%) 52% Air Traffic Controllers
Education Teachers (2%) (2%) 29%
Building Architects (2%) 72% Journalists (2%) 52% Graphic and Multimedia 50% Mechanical Engineers (1%)
Designers (1%) 5%
 GENDERED CAREER ASPIRATIONS 63

Supplement S4. Full Results from Path Analysis

Male Female Participants

Participants
β (SE) β (SE)
Intercepts
12th-Grade Math Achievement 1.10** (0.33) 0.61 (0.33)
12th-Grade Reading Achievement 0.50 (0.37) 1.67*** (0.34)
12th-Grade Math Self-Concept 2.04*** (0.28) 1.54*** (0.25)
12th-Grade German Self-Concept 2.70*** (0.39) 3.29*** (0.41)
12th-Grade Math Grade 3.38*** (0.32) 3.51*** (0.29)
12th-Grade German Grade 3.86*** (0.40) 4.58*** (0.36)
12th-grade Math Achievement on
9th-Grade Math Achievement 0.45*** (0.03) 0.51*** (0.03)
9th-Grade Reading Achievement 0.10** (0.03) 0.14*** (0.03)
9th-Grade Math Self-Concept 0.13** (0.04) 0.13*** (0.03)
9th-Grade German Self-Concept -0.02 (0.03) -0.04 (0.03)
9th-Grade Math Grade 0.06 (0.04) 0.06 (0.03)
9th-Grade German Grade 0.03 (0.04) 0.01 (0.03)
SES 0.03 (0.03) 0.03 (0.02)
12th-grade Reading Achievement on
9th-Grade Math Achievement 0.25*** (0.03) 0.28*** (0.03)
9th-Grade Reading Achievement 0.35*** (0.04) 0.35*** (0.03)
9th-Grade Math Self-Concept -0.06 (0.04) -0.18*** (0.04)
9th-Grade German Self-Concept 0.05* (0.04) 0.06 (0.04)
9th-Grade Math Grade 0.02 (0.05) 0.14** (0.04)
9th-Grade German Grade 0.08* (0.04) 0.06 (0.04)
SES 0.04 (0.03) 0.01 (0.02)
12th-grade Math Self-Concept on
9th-Grade Math Achievement 0.19*** (0.03) 0.14*** (0.03)
9th-Grade Reading Achievement -0.04 (0.02) 0.02 (0.02)
9th-Grade Math Self-Concept 0.45*** (0.04) 0.52*** (0.03)
9th-Grade German Self-Concept -0.04 (0.03) -0.04 (0.03)
9th-Grade Math Grade 0.16*** (0.03) 0.14*** (0.03)
9th-Grade German Grade 0.03 (0.03) 0.03 (0.03)
SES -0.00 (0.02) 0.01 (0.02)
12th-grade German Self-Concept on
9th-Grade Math Achievement -0.09** (0.03) -0.05 (0.03)
9th-Grade Reading Achievement 0.11** (0.03) 0.12*** (0.03)
9th-Grade Math Self-Concept -0.08 (0.04) -0.07* (0.03)
9th-Grade German Self-Concept 0.36*** (0.04) 0.36*** (0.04)
9th-Grade Math Grade 0.12* (0.05) 0.03 (0.04)
9th-Grade German Grade 0.13** (0.04) 0.16*** (0.04)
SES 0.06 (0.03) 0.04 (0.03)
12th-grade Math Grade on
9th-Grade Math Achievement 0.19*** (0.03) 0.10*** (0.03)
9th-Grade Reading Achievement 0.04 (0.03) 0.06* (0.03)
9th-Grade Math Self-Concept 0.24*** (0.04) 0.20*** (0.04)
9th-Grade German Self-Concept -0.07* (0.03) -0.05*** (0.03)
9th-Grade Math Grade 0.24*** (0.04) 0.31*** (0.04)
9th-Grade German Grade 0.17*** (0.03) 0.14*** (0.03)
SES 0.04 (0.03) 0.03 (0.02)
12th-grade German grade on
9th-Grade Math Achievement 0.02 (0.03) 0.05 (0.03)
9th-Grade Reading Achievement 0.13*** (0.04) 0.12*** (0.03)
9th-Grade Math Self-Concept -0.04 (0.04) -0.05 (0.04)
 GENDERED CAREER ASPIRATIONS 64

9th-Grade German Self-Concept 0.17*** (0.04) 0.13*** (0.03)

9th-Grade Math Grade 0.20*** (0.05) 0.14** (0.05)
9th-Grade German Grade 0.25*** (0.04) 0.31*** (0.04)
SES 0.11*** (0.03) 0.06* (0.03)
12th-Grade Math Grade with
12th-Grade Math Self-Concept 0.68*** (0.02) 0.65*** (0.02)
12th-Grade German Self-Concept 0.02 (0.06) 0.09** (0.03)
12th-Grade Math Achievement 0.21*** (0.03) 0.15*** (0.03)
12th-Grade Reading Achievement 0.13*** (0.04) 0.13*** (0.03)
12th-Grade German Grade 0.29*** (0.03) 0.34*** (0.04)
12th-Grade German grade with
12th-Grade Math Self-Concept 0.01 (0.03) 0.04 (0.03)
12th-Grade German Self-Concept 0.60*** (0.02) 0.57*** (0.03)
12th-Grade Math Achievement -0.01 (0.03) 0.07* (0.03)
12th-Grade Reading Achievement 0.09** (0.03) 0.11*** (0.03)
12th-Grade Reading Achievement with
12th-Grade Math Self-Concept 0.03 (0.04) 0.09** (0.03)
12th-Grade German Self-Concept 0.06 (0.03) 0.04 (0.03)
12th-Grade Math Achievement 0.23* (0.03) 0.14*** (0.03)
12th-Grade Math Achievement with
12th-Grade Math Self-Concept 0.21*** (0.03) 0.16*** (0.03)
12th-Grade German Self-Concept -0.09** (0.03) -0.04 (0.03)
12th-Grade German Self-Concept with
12th-Grade Math Self-Concept -0.11*** (0.03) -0.034 (0.03)
Variance Explained (R2)
12th-Grade Math Achievement 0.39*** (0.02) 0.49*** (0.03)
12th-Grade Reading Achievement 0.29*** (0.03) 0.34*** (0.02)
12th-Grade Math Self Concept 0.50*** (0.02) 0.52*** (0.02)
12th-Grade German Self Concept 0.30*** (0.03) 0.28*** (0.03)
12th-Grade Math Grade 0.41*** (0.03) 0.39*** (0.03)
12th-Grade German Grade 0.30*** (0.02) 0.30*** (0.03)
Note. See the description of grade and SES variables in the methods section. Trajectory
membership regressions are presented in Table 4.
* p < 0.05
** p < 0.01
*** p < 0.001

View publication stats