Journal for STEM Education Research

https://doi.org/10.1007/s41979-023-00099-5

RESEARCH BRIEF

Predicting STEM Major Choice: a Machine Learning

Classification and Regression Tree Approach

Chi‑Ning Chang1 · Shuqiong Lin2 · Oi‑Man Kwok3 · Guan Kung Saw4

Accepted: 25 April 2023

© The Author(s) 2023

Abstract
Despite the increasing demand for professionals in science, technology, engi-
neering, and mathematics (STEM), only a small portion of young people in the
USA pursue a postsecondary degree in STEM. To identify the major predictors
of STEM participation, this study uses a machine learning approach, a Clas-
sification and Regression Tree (CART), to analyze a wide range of individual,
family, and school factors obtained from national survey data of US high school
freshmen in fall 2009 who eventually enrolled in STEM college majors by
2016. The analytic results indicate that calculus credits, science identity, total
STEM credits, and math achievement are the most predictive factors during the
high school years of college STEM major selection. The CART-based tree also
shows how these four variables interactively predict the likelihood of students
enrolling in STEM college majors.

Keywords STEM college major · Course-taking · Motivation · Classification and

regression tree · High School Longitudinal Study of 2009–2016

Despite the increasing demand for mid- and high-skilled professionals in sci-
ence, technology, engineering, and mathematics (STEM), only a small por-
tion of young people in the USA pursue a postsecondary degree in STEM
(National Science Foundation, 2019; U.S. Bureau of Labor Statistics, 2017;
U.S. Department of Education, 2015). Prior studies employing expectancy-
value theory and social cognitive career theory have documented a variety of
individual and contextual factors—such as mathematics achievement, science

* Chi‑Ning Chang
[email protected]
1
Virginia Commonwealth University, Richmond, VA, USA
2
American Institutes for Research, Austin, TX, USA
3
Texas A&M University, College Station, TX, USA
4
Claremont Graduate University, Claremont, CA, USA

13
Vol.:(0123456789)
 Journal for STEM Education Research

self-efficacy, and financial support—that are related to STEM college major

choice and career aspirations (Mau & Li, 2018; Wang, 2013; Wille et al.,
2020). Less clear is which factor plays a relatively more significant role for
adolescents who choose a STEM college major in pursuit of a STEM-related
career. Identifying the most predictive factor(s) of STEM college major
choice among high school students has important implications for efforts
to increase STEM participation. This study addresses that critical research
gap by analyzing the US nationally representative High School Longitudinal
Study of 2009–2016 (HSLS:09–16) (National Center for Education Statistics
[NCES], 2018) to identify individual, family, and school factors in adoles-
cence that are most predictive of students entering a postsecondary STEM
degree program.
The HSLS:09–16 study began with more than 23,000 US ninth grad-
ers (high school freshmen), their parents, math and science teachers, school
administrators, and school counselors in fall 2009 (NCES, 2018). It collected a
broad range of STEM-related variables, including both intrinsic factors (such
as math interest and science identity) and extrinsic/behavioral factors (such
as STEM course-taking and afterschool program participation) (NCES, 2018).
We employed the Classification and Regression Tree (CART) algorithm,
which uses a machine learning approach that permits auto-selection and fur-
nishes the results with a tree structure, to help visualize how STEM-related
variables influence students’ decision-making related to STEM major choice
(Steinberg & Colla, 2009a, b). This study is one of the first to apply the CART
method to uncover the most predictive factors that influence pursuit of STEM
degrees based on hundreds of variables in a nationally representative, longitu-
dinal study.

Literature Review

Increasing opportunities for learners to choose STEM careers are a national

priority (National Science Foundation, 2020). Given that most STEM workers
(72.3%) have a college degree in STEM (U.S. Census Bureau, 2019), inves-
tigating the factors that influence a high school student’s choice to pursue
a STEM college major can help address this priority. Previous studies have
identified various pre-college factors that might be associated with STEM
college major choice. Students’ demographic and family backgrounds are
major factors. Specifically, female students, racial and ethnic minorities, and
economically disadvantaged students tend to show lower interest in pursu-
ing careers in STEM (Riegle-Crumb & Morton, 2017; Saw et al., 2018). Par-
ents’ occupations and involvement also influence students’ STEM learning
and career development (Howard et al., 2019; Moakler & Kim, 2014). Other
factors include students’ performance and motivation in STEM (Eccles, 1983,
2009; Lent et al., 1994; Saw & Chang, 2018; Wang, 2013), as well as their

13
Journal for STEM Education Research

learning experiences and context factors in high school, such as school loca-
tion (Saw & Agger, 2021), teacher quality (Althauser, 2015; Lee et al., 2015;
Park et al., 2019), extracurricular opportunities (Kitchen et al., 2018; Franco
& Patel, 2017; Means et al., 2016), and STEM course-taking (Gottfried &
Bozick, 2016).
Although previous studies have collectively identified a broad range of factors
that could potentially affect STEM college major choice, each study has only cov-
ered limited aspects due to the scope of the research. Some studies suggest that
future research should include more potential exogenous variables, such as science-
related motivational factors rather than merely math-related expectancy value con-
structs, to investigate the links between these factors and the pursuit of STEM career
pathways (Gottfried & Bozick, 2016; Wang, 2013; Wille et al., 2020). In practice,
all of these identified factors work simultaneously throughout the STEM career
development process. Therefore, it is important not only to investigate what factors
can predict the choice of a STEM college major, but also to identify how some of
the factors can play a relatively more significant role than others in predicting the
choice of college major.
To fill this literature gap, the present study includes a wider range of pre-
dictors collected by the HSLS:09–16 study. For students’ demographics, we
included predictors such as socioeconomic status (SES), gender, and race/eth-
nicity, as previous studies have shown that female students, racial and ethnic
minorities, and low-income students are less likely to pursue STEM careers
(Riegle-Crumb & Morton, 2017; Saw et al., 2018). For students’ family back-
grounds and parental involvement, we selected predictors such as parents’
occupations and their support for math and science homework, as well as in-
school and out-of-school STEM activities. These parental factors could ben-
efit students’ learning and career development in STEM by providing them
with greater exposure and opportunities (Howard et al., 2019; Moakler & Kim,
2014). For students’ career aspirations, motivation, and performance in STEM,
we selected variables including their career and education goals, math and
science self-efficacy, utility, identity, interest, and cost, as well as a range of
performance measures (e.g., math standardized scores, GPA, SAT, ACT, AP
and IB scores in STEM), based on expectancy-value theory, social cognitive
career theory, and prior research (Eccles, 1983, 2009; Lent et al., 1994; Saw &
Chang, 2018; Wang, 2013).
For teacher quality, we included unobserved factors that are critical to stu-
dents’ STEM learning achievement, such as math and science teachers’ per-
ceptions of professional learning communities, self-efficacy, expectations,
collective responsibility, and principal support (Althauser, 2015; Lee et al.,
2015; Park et al., 2019). For school location, we selected urbanicity and geo-
graphic region as predictors, given the geographic disparities in postsecondary
STEM participation (Saw & Agger, 2021). For extracurricular opportunities,
we included variables such as whether a school offers STEM-related programs
(e.g., supporting underrepresented students in STEM and informing parents

13
 Journal for STEM Education Research

about college majors and careers in STEM), which may benefit students pur-
suing careers in STEM (Kitchen et al., 2018; Franco & Patel, 2017; Means
et al., 2016). Since high school STEM course completion is positively linked to
college major choice (Gottfried & Bozick, 2016), we included a list of STEM
courses taken as predictors.
No prior studies have included such a large number of relevant variables
to explore factors that could predict the choice of a STEM college major in
high school students. The CART algorithm is a powerful tool for identify-
ing factors with the most predictive power while unveiling how the selected
factors interactively predict the STEM college major choice. This has never
been applied in prior studies due to substantially smaller numbers of vari-
ables along with traditional analytic approaches (e.g., logistic regression in
Lee’s (2015) study, multilevel logistic regression in Bottia and colleagues’
(2017) study, and Wang’s (2013) study with the use of structural equation
modeling). By examining a wide range of potential predictors and applying
this advanced technique, this study could provide educators and policymak-
ers with new perspectives and insights into which factors could be relatively
more important in predicting the choice of a STEM college major among high
school students.

Methods

Sample and Measures

The eligible sample from HSLS:09–16 is composed of 11,560 US high school

students who participated in the 2009 base year, 2012 first follow-up survey,
2013–2014 updates and high school transcripts collection, and then reported
their college majors in the 2016s follow-up survey. About 23% of these stu-
dents majored in STEM. Guided by prior studies, we selected a wide range
of 102 variables, including individual, family, and school factors. These vari-
ables are used simultaneously to predict students’ college majors as either
STEM or non-STEM. The list of variables for this study is provided in the
Appendix.

Analytic Strategy

We employed the CART algorithm, implemented using the R package rpart

(Therneau & Atkinson, 1997), to capture the complex mechanism of students’
decision-making with regard to enrolling in STEM college majors by identify-
ing a set of factors and explaining how those factors predict the students’ deci-
sions about enrolling in STEM majors. The algorithm was chosen due to its
desirable properties: (a) it does not require strong model assumptions, which

13
Journal for STEM Education Research

are typically needed when using traditional regression models; (b) it automati-
cally identifies the important predictors and their linear/nonlinear relation-
ships with outcomes (Lee et al., 2010; Steinberg & Colla, 2009a, b; Timofeev,
2004); (c) it is able to handle missing data without extra imputation proce-
dures (Deconinck et al., 2005; Feelders, 1999; Verbyla, 1987); and (d) it is an
interpretation-friendly algorithm compared with other “black-box” data-mining
techniques.
First, to build a CART-based tree, the Gini index (Breiman et al., 1984; Stein-
berg & Colla, 2009a, b) was used to automatically select the important independent
variables. The maximum depth of the tree was set at 30. Cost complexity (Breiman
et al., 1984), with complexity parameters equaling 0.1, was chosen in the pruning
process. Surrogate splitting (Feelders, 1999) was used to handle missing data for
independent variables. Through these settings, the algorithm produced a pruned tree
to predict the probability that a given student will declare a college major in STEM
based on the selected predictors.
Second, to avoid model overfitting issues and to be able to evaluate predic-
tive accuracy, the sample was split into training and testing datasets using the
80/20 rule (Anis et al., 2015; Zheng, 2004). Specifically, we used the random
sampling method without replacement to select 80% of the samples ( Ntrain =
9248) as the training data for developing the CART-based tree. The remaining
20% of the samples ( Ntest = 2312), who were not exposed to the tree develop-
ment, served as the testing data to evaluate the predictive accuracy of the tree.
In other words, we established the statistical model used to predict the out-
come using the training dataset, and we used the testing data to validate the
prediction through the established model. The measure of prediction accuracy
was examined. A sensitivity analysis using random forest analysis was con-
ducted to evaluate the consistency of the CART results. The CART algorithm
also applied the student longitudinal analytic weight provided by HSLS:09–16.
Hence, the results are weighted to represent US ninth graders in fall 2009.

Results

Figure 1 shows the output of the final CART-based tree, which predicts the
probability of a student declaring a STEM college major. Out of all the inde-
pendent variables, only four variables are deemed relatively more important and
are automatically selected to construct the final tree: credits earned in calculus
during high school, science identity in grade 11, total STEM credits earned dur-
ing high school, and math achievement in grade 11. Therefore, these four vari-
ables play relatively important roles in a student’s decision to choose a STEM
college major.
With the final four predictors, the trained samples were split into five
groups. As illustrated in Fig. 1, Group 1 is students (accounting for 81% of the
high school students) who did not earn any credits in calculus and have a low

13
 Journal for STEM Education Research

Fig. 1  The final CART-based tree. Probability indicates the chance of actually majoring in STEM.
The HSLS:09–16 provides Z scores for science identity (X2SCIID) and T scores for math achievement
(X2TXMTSCOR). To make these two standardized scores more comprehensible when interpreting the
results, while also keeping the interpretation consistent across these two measures, we present the per-
centile ranks (PR) for these two measures converted from Z score and T score

probability of majoring in STEM (prob. = 0.16). Group 2 is students (account-

ing for 7% of the high school students) who earned credits in calculus and had
a percentile rank (PR) for science identity in 11th grade < 74, and also have a
low probability of selecting a college major in STEM (prob. = 0.22). Group 3
is students (accounting for 5% of the high school students) who earned cred-
its in calculus, had a PR for science identity in 11th grade ≥ 74, earned fewer
than 9.8 STEM credits during high school, and had a PR for math achievement
scores in 11th grade < 97, and have a probability of declaring a STEM college
major (prob. = 0.37).

13
Journal for STEM Education Research

Fig. 2  STEM pipelines — predicting STEM college major choice in high school (2009–2016). The water
drops represent the probability of declaring a college major in STEM. PR stands for percentile rank. The
CART results illustrate how these four variables influenced the 2009 cohort’s college major choice in
2016. In summary, if high school students do not earn any calculus credits, the likelihood of majoring in
STEM disciplines will be only 16%. Furthermore, even if students earn calculus credit(s), their chance
of pursuing a postsecondary STEM degree will still be low (22%) if they do not exhibit a high science
identity in the 11th grade (PR < 74). On the other hand, if students earn calculus credit(s) and have a high
level of science identity in the 11th grade (PR ≥ 74), the likelihood of enrolling in a STEM college major
will increase substantially (from 16 to 37%). Interestingly, the probability of students majoring in STEM
will be boosted to 68% if students earn calculus credit(s), have a high level of science identity in the 11th
grade, and either earn at least 9.8 credits in STEM-related courses or have high math achievement in
the11th grade (PR ≥ 97)

The remaining two groups have average probabilities larger than 0.5. Group 4 is
students (accounting for 2% of the high school students) who had credits in calculus,
a PR of science identity in 11th grade ≥ 74, total STEM credits < 9.8, and a PR for
math achievement scores in 11th grade ≥ 97, and show the highest probability of
enrolling in a STEM major in college (prob. = 0.68). Group 5 is students (account-
ing for 5% of the high school students) who earned credits in calculus, had a PR for
science identity in 11th grade ≥ 74, and earned at least 9.8 credits in STEM (i.e.,

13
 Journal for STEM Education Research

X3TCREDSTEM ≥ 9.8), and also have the highest probability of selecting a college
major in STEM (prob. = 0.68).
Using the CART algorithm for prediction based on the test dataset (i.e., 20% of
the full data) led to a classification accuracy equaling 0.80. A sensitivity analysis
using random forest analysis indicated that the four selected variables in the CART
are also identified as important variables using the mean decrease accuracy method,
which further strengthened our confidence in the CART results. The CART-based
tree can also be converted into Fig. 2, a more understandable image demonstrating
how these four variables interactively predict STEM college major choice.

Discussion

Consistent with prior studies (Gottfried & Bozick, 2016; Riegle-Crumb et al.,
2012), our findings suggest that completing at least one calculus class during
high school is highly predictive of entering STEM fields. More importantly,
our CART analysis is the first to demonstrate that calculus course completion
is the most predictive factor among 102 examined variables, including indi-
vidual, family, and school factors. Specifically, the probability of selecting
a STEM college major is only 16% for students who do not earn any calcu-
lus credits during high school. This set of findings underscores the impor-
tance of offering and supporting the completion of advanced math course-
work for high school students, particularly the study of calculus. Alarmingly,
only about 50% of high schools in the USA offer calculus (U.S. Department
of Education, 2016). Although our study does not include the school-level
course offering variables, we could still speculate that students from the cal-
culus-excluded schools might be more likely to have a lower rate of STEM
participation.
Our study also uncovers that science identity is the second most predictive
variable for enrolling in a postsecondary STEM degree program. It is important
to note that science identity is relatively more significant when compared to
other STEM motivational factors, including math self-efficacy, science interest,
and STEM career aspiration. Science identity reflects how students act to con-
vince themselves and others that they are science students, which is a powerful
source of persistence in science (Robinson et al., 2019; Stets et al., 2017). Our
CART study indicates that if students earn calculus credit(s) and report a high
level of science identity (PR ≥ 74), the likelihood of choosing a STEM college
major will increase substantially, to 37% and higher. School administrators and
policymakers might consider developing or adopting programs or curricula that
can help students cultivate a science identity in high school or at an earlier
stage.
Predictably, students who earn more credits in STEM-related courses and
have excellent math achievement in high school are more likely to enroll in
STEM majors in college. However, these two factors (3rd and 4th most

13
Journal for STEM Education Research

predictive variables) are “conditional” on the first two. In other words, only if
students earn 9.8 or more STEM credits or demonstrate excellent math achieve-
ment, in combination with earning calculus credit(s) and having a high level
of science identity, will the probability of declaring a STEM college major
increase from 37 to 68%. This “conditional” implication, uncovered by the
CART method, is a novel finding and addition to the literature on STEM educa-
tion and career development.
There are four limitations to our study. First, our study relies on pub-
lic-use secondary data. Other important predictors that are not released
(e.g., school-level course offerings) or collected (e.g., neighborhood STEM
resources) might be omitted. For example, the inclusion of school- and dis-
trict-level data could provide insight into how related policies, resources,
and programs contribute to students’ STEM learning and pursuit of STEM
careers. Due to this limitation, we are unable to determine the relative
importance of the four selected variables compared to the variables omit-
ted from this study. Although we could not include all possible predictors
in our model, our study covers a broader range of aspects for prediction
than previous studies. Second, we restrict the initial sample (23,000 + ninth
graders from 944 schools in 2009) to those students (n = 11,560) who par-
ticipated in the follow-up surveys from 2009 to 2016. We acknowledge that
attrition bias might be a threat to internal and external validity. Therefore,
to reduce the threat, our analysis has applied the student longitudinal ana-
lytic weight provided by the NCES. Third, our study results could only
represent the findings for the 2009 ninth-grade cohort. Nonetheless, this
cohort is the latest nationally representative, longitudinal high school sam-
ple for STEM education research conducted by the NCES. Fourth, some
of our measures (e.g., parental involvement and STEM motivational fac-
tors) involve items with repeated measures. However, each of these items
is measured only twice (grades 9 and 11) in high school. Due to the lim-
ited number of repeated measures, we use the regular CART for this study.
Future longitudinal studies with repeated measures at multiple time points
could consider employing the promising longitudinal CART algorithm
(Kundu & Harezlak, 2019).
Despite these limitations, the findings of this study contribute to the current
STEM literature in the following important ways: (a) identifying the relatively
important variables among a rich set of predictors associated with STEM col-
lege major choice, (b) presenting how these four most predictive variables
interactively predict the likelihood of choosing a STEM college major, and
(c) demonstrating the potential of using the CART algorithm to uncover pre-
viously unexamined nuances of STEM educational and career pathways.
Well-developed and effectively implemented programs could increase STEM
participation and motivation (Hudson et al., 2020; Pike & Robbins, 2019).
Our findings provide educators and policymakers with new perspectives and
insights on which relatively important factors could be intervened among
young students.

13
 Appendix 1
Table 1
Table 1  List of variables for the study

13
Variable name Description Data type

Dependent variable
X4RFDGMJSTEM X4 If first or second/double major is STEM Categorical
Independent variables
Demographics
X1SES X1 Student SES (composite score) Continuous
X1SEX X1 Student Gender Categorical
X1RACE X1 Student Race/Ethnicity Categorical
Family backgrounds
X1DADOCC_STEM1 X1 Father/male guardian’s current/most recent occupation: STEM code 1 (sub-domain) Categorical
X1MOMOCC_STEM1 X1 Mother/female guardian’s current/most recent occupation: STEM code 1 (sub-domain) Categorical
X2DADOCC_STEM1 X2 Father/male guardian’s current/most recent occupation: STEM code 1 (sub-domain) Categorical
X2MOMOCC_STEM1 X2 Mother/female guardian’s current/most recent occupation: STEM code 1 (sub-domain) Categorical
Parental involvement
P1MTHHWEFF P1 Confidence in helping with ­9th-grade math homework Categorical
P1SCIHWEFF P1 Confidence in helping with ­9th-grade science homework Categorical
P1CAMPMS P1 Participated in math or science camp outside of school in last year Categorical
P1STEMDISC P1 Discussed STEM program or article with 9th grader in last year Categorical
P2MTHHWEFF P2 Confidence in helping with math homework 2011–2012/when last enrolled Categorical
P2SCIHWEFF P2 Confidence in helping with science homework 2011–2012/when last enrolled Categorical
P2STEMDISC P2 Discussed STEM program or article with teenager in last year Categorical
STEM career aspiration
X1STU30OCC_STEM1 X1 Student occupation at age 30: STEM code 1 (sub-domain) Categorical
X2STU30OCC_STEM1 X2 Student occupation at age 30: STEM code 1 (sub-domain) Categorical
Journal for STEM Education Research
 Table 1  (continued)
Variable name Description Data type
X1STUEDEXPCT X1 How far in school 9th grader thinks he/she will get Categorical
STEM motivation
X1MTHEFF X1 Math self-efficacy (composite score) Continuous
X1MTHUTI X1 Math utility (composite score) Continuous
X1MTHID X1 Math identity (composite score) Continuous
X1MTHINT X1 Math interest (composite score) Continuous
Journal for STEM Education Research

X1SCIEFF X1 Science self-efficacy (composite score) Continuous

X1SCIUTI X1 Science utility (composite score) Continuous
X1SCIID X1 Science identity (composite score) Continuous
X1SCIINT X1 Science interest (composite score) Continuous
X2STUEDEXPCT X2 How far in school 9th grader thinks he/she will get Categorical
X2MTHEFF X2 Math self-efficacy (composite score) Continuous
X2MTHUTI X2 Math utility (composite score) Continuous
X2MTHID X2 Math identity (composite score) Continuous
X2MTHINT X2 Math interest (composite score) Continuous
X2SCIEFF X2 Science self-efficacy (composite score) Continuous
X2SCIUTI X2 Science utility (composite score) Continuous
X2SCIID X2 Science identity (composite score) Continuous
X2SCIINT X2 Science interest (composite score) Continuous
X2BEHAVEIN X2 Scale of school motivation (composite score) Continuous
X2MEFFORT X2 Scale of math class effort (composite score) Continuous
X2SEFFORT X2 Scale of science class effort (composite score) Continuous
S1TEFRNDS S1 Time/effort in math/science means not enough time with friends Categorical
S1TEACTIV S1 Time/effort in math/science means not enough time for extracurriculars Categorical

13
 Table 1  (continued)
S1TEPOPULAR S1 Time/effort in math/science means 9th grader won’t be popular Categorical
S1TEMAKEFUN S1 Time/effort in math/science means people will make fun of 9th grader Categorical

13
Student academic performance and preparation
X1TXMTSCOR X1 Mathematics standardized theta score Continuous
X1SCHOOLBEL X1 Scale of student’s sense of school belonging (composite score) Continuous
X1SCHOOLENG X1 Scale of student’s school engagement (composite score) Continuous
X2TXMTSCOR X2 Mathematics standardized theta score Continuous
X2EVERDROP X2 Ever drop out Categorical
X2PROBLEM X2 Scale of problems at high school (composite score) Continuous
X3EVERDROP X3 Ever drop out Categorical
X3TGPAMAT X3 GPA: mathematics Continuous
X3TGPAHIMTH X3 GPA: highest level mathematics course taken Continuous
X3TGPASCI X3 GPA: science Continuous
X3TGPAHISCI X3 GPA: highest level science course taken Continuous
X3TGPAENGIN X3 GPA: engineering/engineering tech Continuous
X3TGPASTEM X3 GPA: STEM courses Continuous
X3TGPATOT X3 Overall GPA computed Continuous
X3TGPAMTHAP X3 GPA: AP/IB math courses Continuous
X3TGPASCIAP X3 GPA: AP/IB science courses Continuous
S1HRMHOMEWK S1 Hours spent on math homework/studying on typical school day Categorical
S1HRSHOMEWK S1 Hours spent on science homework/studying on typical school day Categorical
C2AVGSATMATH C2 Average SAT mathematics score Continuous
C2AVGACTMATH C2 Average ACT mathematics score Continuous
C2AVGACTSCI C2 Average ACT science score Continuous
Math and science teacher quality
Journal for STEM Education Research
 Table 1  (continued)
X1TMCOMM X1 Scale of math teacher’s perceptions of math professional learning community (composite score) Continuous
X1TMEFF X1 Scale of math teacher’s self-efficacy (composite score) Continuous
X1TMEXP X1 Scale of math teacher’s perceptions of math teachers’ expectations (composite score) Continuous
X1TMPRINC X1 Scale of math teacher’s perceptions of principal support (composite score) Continuous
X1TMRESP X1 Scale of math teacher’s perceptions of collective responsibility (composite score) Continuous
X1TSCOMM X1 Scale of science teacher’s perceptions of science professional learning community (composite score) Continuous
X1TSEFF X1 Scale of science teacher’s self-efficacy (composite score) Continuous
X1TSEXP X1 Scale of science teacher’s perceptions of science teachers’ expectations (composite score) Continuous
Journal for STEM Education Research

X1TSPRINC X1 Scale of science teacher’s perceptions of principal support (composite score) Continuous
X1TSRESP X1 Scale of science teacher’s perceptions of collective responsibility (composite score) Continuous
School location
X1LOCALE X1 School locale (urbanicity) Categorical
X1REGION X1 School geographic region Categorical
X2LOCALE X2 School locale (urbanicity) Categorical
X2REGION X2 School geographic region Categorical
Extracurricular programs
C1PURSUE C1 School has program to encourage underrepresented student in math/science Categorical
C1INFORM C1 School has program to inform parent about math/science higher ed/careers Categorical
C1ENCCLG C1 School has program to encourage students not considering college to do so Categorical
C2ENCSTEM C2 School has program to encourage underrepresented student in STEM Categorical
C2INFSTEM C2 School has program to inform parent about STEM higher ed/careers Categorical
C2ENCCLG C2 School has program to encourage students not considering college to do so Categorical
STEM course-taking
X3T1CREDALG1 X3 At least one credit earned in: algebra 1 Categorical
X3T1CREDALG2 X3 At least one credit earned in: algebra 2 Categorical
X3T1CREDINTM X3 At least one credit earned in: integrated math Categorical

13
 Table 1  (continued)

X3T1CREDPREC X3 At least one credit earned in: analysis/pre-calculus Categorical

X3TCREDAPMTH X3 Credits earned in: AP/IB mathematics courses Continuous

13
X3T1CREDCALC X3 At least one credit earned in: calculus Categorical
X3T1CREDGEO X3 At least one credit earned in: geometry Categorical
X3T1CREDSTAT​ X3 At least one credit earned in: statistics/probability Categorical
X3T1CREDTRIG X3 At least one credit earned in: trigonometry Categorical
X3TCREDMAT X3 Credits earned in: mathematics Continuous

X3TCREDAPSCI X3 Credits earned in: AP/IB science courses Continuous

X3T1CREDBIOL X3 At least one credit earned in: biology Categorical
X3T1CREDCHEM X3 At least one credit earned in: chemistry Categorical
X3T1CREDESCI X3 At least one credit earned in: geology/earth science Categorical
X3T1CREDPHYS X3 At least one credit earned in: physics Categorical
X3TCREDSCI X3 Credits earned in: science Continuous
X3TCREDENGIN X3 Credits earned in: engineering/engineering tech Continuous
X3TCREDSTEM X3 Credits earned in: STEM Continuous
X3TCREDAPIB X3 Credits earned in: AP/IB combined Continuous
X3TCREDMTSC X3 Credits earned in: combined mathematics and science Continuous
Analytic weight
W4W1STU Student longitudinal analytic weight Continuous
A search for the variable name in the codebook (http://​nces.​ed.​gov/​onlin​ecode​book) will reveal more details on value labels, variable description, frequency, and percentage
X1 composite variables/fall term of the 9th grade (2009), X2 composite variables/spring term of the 11th grade (2012), X3 composite variables/beyond high school gradu-
ation (2013–2014). X4 composite variables/second follow-up (2016), S1 student survey/fall term of the 9th grade (2009), C1 counselor survey/fall term of the 9th grade
(2009), C2 counselor survey/spring term of the 11th grade (2012), P1 parent survey/fall term of the 9th grade (2009), P2 parent survey/spring term of the 11th grade (2012)
Journal for STEM Education Research
Journal for STEM Education Research

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License,
which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as
you give appropriate credit to the original author(s) and the source, provide a link to the Creative Com-
mons licence, and indicate if changes were made. The images or other third party material in this article
are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the
material. If material is not included in the article’s Creative Commons licence and your intended use is
not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission
directly from the copyright holder. To view a copy of this licence, visit http://​creat​iveco​mmons.​org/​licen​
ses/​by/4.​0/.

References
Althauser, K. (2015). Job-embedded professional development: Its impact on teacher self-efficacy and
student performance. Teacher Development, 19(2), 210–225. https://​doi.​org/​10.​1080/​13664​530.​
2015.​10113​46
Anis, M., Ali, M., & Yadav, A. (2015). A comparative study of decision tree algorithms for class imbal-
anced learning in credit card fraud detection. International Journal of Economics, Commerce and
Management, 3(12), 86–102.
Bottia, M. C., Mickelson, R. A., Giersch, J., Stearns, E., & Moller, S. (2018). The role of high school
racial composition and opportunities to learn in students’ STEM college participation. Journal of
Research in Science Teaching, 55(3), 446–476. https://​doi.​org/​10.​1002/​tea.​21426
Breiman, L., Friedman, J., Stone, C. J., & Olshen, R. A. (1984). Classification and regression trees. CRC
Press.
Deconinck, E., Hancock, T., Coomans, D., Massart, D. L., & Heyden, Y. V. (2005). Classification of
drugs in absorption classes using the classification and regression trees (CART) methodology. Jour-
nal of Pharmaceutical and Biomedical Analysis, 39(1), 91–103. https://​doi.​org/​10.​1016/j.​jpba.​2005.​
03.​008
Eccles, J. (1983). Female achievement patterns: Attributions, expectancies, values, and choice. Journal of
Social Issues, 1, 1–26.
Eccles, J. (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/​00461​52090​
28323​68
Feelders, A. (1999). Handling missing data in trees: Surrogate splits or statistical imputation? European
Conference on Principles of Data Mining and Knowledge Discovery (pp. 329–334). Springer.
Franco, M. S., & Patel, N. H. (2017). Exploring student engagement in STEM education: An examination
of STEM schools, STEM programs, and traditional schools. Research in the Schools, 24(1), 10–30.
Gottfried, M. A., & Bozick, R. (2016). Supporting the STEM pipeline: Linking applied STEM course-
taking in high school to declaring a STEM major in college. Education Finance and Policy, 11(2),
177–202. https://​doi.​org/​10.​1162/​EDFP_a_​00185
Howard, N. R., Howard, K. E., Busse, R. T., & Hunt, C. (2019). Let’s talk: An examination of parental
involvement as a predictor of STEM achievement in math for high school girls. Urban Education,
58(4), 586–613. https://​doi.​org/​10.​1177/​00420​85919​877933
Hudson, M. A., Baek, Y., Ching, Y. H., & Rice, K. (2020). Using a multifaceted robotics-based inter-
vention to increase student interest in STEM subjects and careers. Journal for STEM Education
Research, 3(3), 295–316. https://​doi.​org/​10.​1007/​s41979-​020-​00032-0
Kitchen, J. A., Sonnert, G., & Sadler, P. M. (2018). The impact of college-and university-run high school
summer programs on students’ end of high school STEM career aspirations. Science Education,
102(3), 529–547. https://​doi.​org/​10.​1002/​sce.​21332
Kundu, M. G., & Harezlak, J. (2019). Regression trees for longitudinal data with baseline covariates. Bio-
statistics & Epidemiology, 3(1), 1–22. https://​doi.​org/​10.​1080/​24709​360.​2018.​15577​97
Lee, A. (2015). Determining the effects of computer science education at the secondary level on STEM
major choices in postsecondary institutions in the United States. Computers & Education, 88, 241–
255. https://​doi.​org/​10.​1016/j.​compe​du.​2015.​04.​019

13
 Journal for STEM Education Research

Lee, B. K., Lessler, J., & Stuart, E. A. (2010). Improving propensity score weighting using machine
learning. Statistics in Medicine, 29(3), 337–346. https://​doi.​org/​10.​1002/​sim.​3782
Lee, S. W., Min, S., & Mamerow, G. P. (2015). Pygmalion in the classroom and the home: Expectation’s
role in the pipeline to STEMM. Teachers College Record, 117(9), 1–40. https://​doi.​org/​10.​1177/​
01614​68115​11700​907
Lent, R. W., Brown, S. D., & Hackett, G. (1994). Toward a unifying social cognitive theory of career and
academic interest, choice, and performance. Journal of Vocational Behavior, 45(1), 79–122. https://​
doi.​org/​10.​1006/​jvbe.​1994.​1027
Matusovich, H. M., Streveler, R. A., & Miller, R. L. (2010). Why do students choose engineering? A
qualitative, longitudinal investigation of students’ motivational values. Journal of Engineering Edu-
cation, 99(4), 289–303. https://​doi.​org/​10.​1002/j.​2168-​9830.​2010.​tb010​64.x
Mau, W. C. J., & Li, J. (2018). Factors influencing STEM career aspirations of underrepresented high
school students. The Career Development Quarterly, 66(3), 246–258. https://​doi.​org/​10.​1002/​cdq.​
12146
Means, B., Wang, H., Young, V., Peters, V. L., & Lynch, S. J. (2016). STEM-focused high schools as a
strategy for enhancing readiness for postsecondary STEM programs. Journal of Research in Science
Teaching, 53(5), 709–736. https://​doi.​org/​10.​1002/​tea.​21313
Moakler, M. W., Jr., & Kim, M. M. (2014). College major choice in STEM: Revisiting confidence and
demographic factors. The Career Development Quarterly, 62(2), 128–142. https://​doi.​org/​10.​1002/j.​
2161-​0045.​2014.​00075.x
National Center for Education Statistics (2018). High school longitudinal study of 2009 (HSLS:09): User
manuals. Retrieved December 15, 2020, from https://​nces.​ed.​gov/​surve​ys/​hsls09/​userm​anuals.​asp
National Science Foundation (2019). Science and engineering indicators: post–high school transitions.
Retrieved January 24, 2021, from https://​ncses.​nsf.​gov/​pubs/​nsb20​196/​post-​high-​school-​trans​itions
National Science Foundation (2020). STEM education for the future: a visioning report. Retrieved June
15, 2022, from https://​www.​nsf.​gov/​ehr/​Mater​ials/​STEM%​20Edu​cation%​20for%​20the%​20Fut​ure%​
20-%​202020%​20Vis​ioning%​20Rep​ort.​pdf
Park, J. H., Lee, I. H., & Cooc, N. (2019). The role of school-level mechanisms: How principal support,
professional learning communities, collective responsibility, and group-level teacher expectations
affect student achievement. Educational Administration Quarterly, 55(5), 742–780. https://​doi.​org/​
10.​1177/​00131​61X18​821355
Pike, G. R., & Robbins, K. (2019). Expanding the pipeline: The effect of participating in project Lead the
way on majoring in a STEM discipline. Journal for STEM Education Research, 2(1), 14–34. https://​
doi.​org/​10.​1007/​s41979-​019-​00013-y
Riegle-Crumb, C., King, B., Grodsky, E., & Muller, C. (2012). The more things change, the more
they stay the same? Prior achievement fails to explain gender inequality in entry into STEM
college majors over time. American Educational Research Journal, 49(6), 1048–1073.
10.3102%2F0002831211435229
Riegle-Crumb, C., & Morton, K. (2017). Gendered expectations: Examining how peers shape female stu-
dents’ intent to pursue STEM fields. Frontiers in Psychology, 8, 329. https://​doi.​org/​10.​3389/​fpsyg.​
2017.​00329
Robinson, K. A., Perez, T., Carmel, J. H., & Linnenbrink-Garcia, L. (2019). Science identity develop-
ment trajectories in a gateway college chemistry course: Predictors and relations to achievement
and STEM pursuit. Contemporary Educational Psychology, 56, 180–192. https://​doi.​org/​10.​1016/j.​
cedps​ych.​2019.​01.​004
Saw, G. K., & Agger, C. A. (2021). STEM pathways of rural and small-town students: Opportunities
to learn, aspirations, preparation, and college enrollment. Educational Researcher, 50(9), 595–606.
https://​doi.​org/​10.​3102/​00131​89X21​10275​28
Saw, G., & Chang, C. N. (2018). Cross-lagged models of mathematics achievement and motivational fac-
tors among Hispanic and non-Hispanic high school students. Hispanic Journal of Behavioral Sci-
ences, 40(2), 240–256. https://​doi.​org/​10.​1177/​07399​86318​766511
Saw, G., Chang, C. N., & Chan, H. Y. (2018). Cross-sectional and longitudinal disparities in STEM
career aspirations at the intersection of gender, race/ethnicity, and socioeconomic status. Educa-
tional Researcher, 47(8), 525–531. https://​doi.​org/​10.​3102/​00131​89X18​787818
Steinberg, D., & Colla, P. (2009a). CART: classification and regression trees. The top ten algorithms in
data mining. X. Wu & V. Kumar (Ed.) 179–201. New York: CRC Press.
Steinberg, D., & Colla, P. (2009b). CART: Classification and regression trees. In X. Wu & V. Kumar
(Eds.), The top ten algorithms in data mining (pp. 179–201). CRC Press.

13
Journal for STEM Education Research

Stets, J. E., Brenner, P. S., Burke, P. J., & Serpe, R. T. (2017). The science identity and entering a science
occupation. Social Science Research, 64, 1–14. https://​doi.​org/​10.​1016/j.​ssres​earch.​2016.​10.​016
Therneau, T. M., & Atkinson, E. J. (1997). An introduction to recursive partitioning using the rpart rou-
tine. (Vol. 61, p. 452) Mayo Foundation: Technical report.
Timofeev, R. (2004). Classification and regression trees (CART) theory and applications (Unpublished
doctoral dissertation). Humboldt University.
U.S. Bureau of Labor Statistics. (2017). STEM occupations: past, present, and future. Retrieved Jun. 1,
2021, from https://​www.​bls.​gov/​spotl​ight/​2017/​scien​ce-​techn​ology-​engin​eering-​and-​mathe​matics-​
stem-​occup​ations-​past-​prese​nt-​and-​future/​home.​htm
U.S. Census Bureau. (2019). 2019 American community survey, 1-year estimates. Retrieved Jun. 23,
2022, https://​www.​census.​gov/​libra​ry/​stori​es/​2021/​06/​does-​major​ing-​in-​stem-​lead-​to-​stem-​job-​
after-​gradu​ation.​html
U.S. Department of Education. (2015). Science, technology, engineering and math: education for global
leadership. Retrieved Jun. 1, 2021, https://​www.​ed.​gov/​sites/​defau​lt/​files/​stem-​overv​iew.​pdf
U.S. Department of Education. (2016). STEM 2026: a vision for innovation in STEM education. Retrieved
Jun. 1, 2021, from https://​www.​air.​org/​system/​files/​downl​oads/​report/​STEM-​2026-​Vision-​for-​Innov​
ation-​Septe​mber-​2016.​pdf
Verbyla, D. L. (1987). Classification trees: A new discrimination tool. Canadian Journal of Forest
Research, 17(9), 1150–1152. https://​doi.​org/​10.​1139/​x87-​177
Wang, X. (2013). Why students choose STEM majors: Motivation, high school learning, and postsecond-
ary context of support. American Educational Research Journal, 50(5), 1081–1121. https://​doi.​org/​
10.​3102/​00028​31213​488622
Wille, E., Stoll, G., Gfrörer, T., Cambria, J., Nagengast, B., & Trautwein, U. (2020). It takes two: expec-
tancy-value constructs and vocational interests jointly predict STEM major choices. Contemporary
Educational Psychology, 61, 101858. https://​doi.​org/​10.​1016/j.​cedps​ych.​2020.​101858
Zheng, J. (2004). Study on the relationship of training data size to error rate and the performance com-
parison for two decision tree algorithms. [Doctoral dissertation, Texas Tech University].

Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps
and institutional affiliations.

13