







































Rodriguez 1 

 

CALIFORNIA SOCIOLOGY FORUM VOL. 5 

Elite Course-Taking and Racial Disparities in STEM  

Eglys Rodriguez 

Department of Sociology 

 California State University, Los Angeles 

 

ABSTRACT  

African American and Latinx populations are still 

disproportionately underrepresented in science, technology, 

engineering, and math (STEM) fields. To understand why racial 

disparities persist, this article investigates African American and 

Latinx students' high school careers. Specifically, it examines 

whether students take a course in trigonometry, pre-calculus, or 

calculus students’ (elite math courses) in their senior year. Using 

the theoretical framework of categorical inequality, I examine 

whether racial disparities exist in elite math courses that often 

serve important gatekeeping functions for future STEM pathways.  

Data are from the Educational Longitudinal Study of 2002 

(ELS:02) public use data, focusing on data from the first follow-

up (when the students are in their senior year of high school) and 

postsecondary education transcripts collected in 2012. I use 

logistic regression methods to examine the odds of taking an elite 

math course across racial and ethnic subgroups. Then I analyze 

the odds of high school graduation rates, postsecondary 

enrollment across different racial/ethnic subgroups, and whether 

students obtained a bachelor’s degree in STEM, after accounting 

for elite math courses taken in high school. Surprisingly, my 

findings show that African American and Latinx students have 

similar odds as White students of taking an elite math course and 

have higher odds of enrolling in a postsecondary institution than 

their white counterparts.  

 

INTRODUCTION  

Careers in science, technology, engineering, and 

mathematics (STEM) have become one of the fastest growing 

areas in the United States; however, racial and gender disparities 

continue to persist in the STEM workforce (Chen 2013). The 

representation of African American and Latinx populations in 

STEM fields is still considerably low; where only 9% of the 

STEM workforce is African Americans and 7% is Latinx (Pew 



Rodriguez 2 

 

CALIFORNIA SOCIOLOGY FORUM VOL. 5 

Research Center 2018). Moreover, only 7% of African Americans 

and 12% of Latinx students graduated with a degree in STEM, 

lower than the national average of 15% (Fry et al. 2021; Pew 

Research Center 2018).  

Examining math course-taking in high school can help us 

understand the factors that contribute to students pursuing and 

persisting in STEM (Sanabria and Penner 2017). In doing so, my 

study highlights the disparate pathways of minority racial and 

ethnic groups in STEM. Prior research has established that the 

type of math courses taken in high school often serves a 

gatekeeping function into STEM in higher education. Students are 

differentially sorted into higher-level math courses offered in high 

schools and evidence shows that students from minority groups 

are less likely to enroll and take higher-tracked math courses, such 

as trigonometry, pre-calculus, or calculus, than White and Asian 

American students (U.S. Department of Education 2018). 

Advanced math course-taking in high school is a strong predictor 

for attending college (Long et al. 2010). Additionally, higher-level 

math courses operate as a “gatekeeper” in STEM, limiting the rate 

at which students can prepare for college coursework in STEM 

fields. Investigating the extent to which African American and 

Latinx students take these higher-tracked math courses in high 

school (which I will refer to simply as elite math courses in this 

thesis) and earn a bachelor’s degree in STEM can better illuminate 

a key mechanism in racial stratification in STEM. Thus, this thesis 

examines the racial disparities in STEM elite math courses taken 

in high school, high school graduation, postsecondary attendance, 

and STEM degree attainment.  

 

LITERATURE REVIEW 

This thesis will be drawing from previous literature which 

utilized various theoretical frameworks to demonstrate why the 

categorical inequality framework can better explain racial 

disparities in STEM. This thesis will also examine previous 

literature which focus on math course taking at the high school 

level, their findings and how racial disparities in math course 

taking has contributed to STEM attrition among African American 

and Latinx students.  

 



Rodriguez 3 

 

CALIFORNIA SOCIOLOGY FORUM VOL. 5 

Racial Disparities in Higher-Level Math Courses  

Previous researchers have focused on patterns relating to 

math course-taking and race. As previously stated, research on 

racial disparities in STEM has been an ongoing issue for decades 

with researchers focusing on several factors affecting math course 

taking. For example, Daempfle’s (2004) study on STEM attrition 

at the undergraduate level found that there was a disconnect 

between high school and college teachings that impacts for 

students’ opting out of STEM majors. While Daempfle’s research 

focused on students at the undergraduate level, others researched 

students at the high school level. Some of the earlier studies 

conducted on math course taking at the high school level found 

that elite math courses can predict postsecondary attendance 

(Riegle-Crumb 2006; Tyson 2007; Riegle-Crumb and Grodsky 

2010), math course taking across different racial/ethnic subgroups 

has increased (with some exceptions) across different cohorts 

(Dalton 2007), African American and Latinx students have lower 

odds of taking higher level courses (Archbald and Farley; Ripple 

2012), and elite math and science courses in high school play a 

significant role in STEM persistence amongst students of color 

(Lichtenberger and George-Jackson 2013).  

While these studies contributed greatly towards the 

understanding of STEM attrition and persistence among students 

of color, they are outdated. Some of the studies used data from the 

National Education Longitudinal study of 1988, Beginning 

Postsecondary Students Longitudinal Study (BPS:96/01), and 

Adolescent Health and Academic Achievement (AHAA) which 

was conducted in the 1990’s. Although the data used in these 

studies were instrumental in research and policy, newer data 

would help researchers today in understanding why STEM 

attrition continues to persist.  

 

Categorical Inequality Framework  

Plenty of research on racial disparities in STEM discusses 

the relationship between elite math course taking and racial 

disparities at different intervals of students education (Riegle-

Crumb 2006; Dalton et al. 2007; Kelly 2009; Kokkelenberg and 

Sinha 2010; Riegle-Crumb 2010; Archbald and Farley-Ripple 

2012; Domina and Saldana 2012; Xie et al. 2015; Champion 2016; 



Rodriguez 4 

 

CALIFORNIA SOCIOLOGY FORUM VOL. 5 

Shi 2017; Fong 2020; Sanabria 2020; McEachin et al. 2020; 

Irizarry 2021). This thesis examines the impact of math course 

taking at the high school level and how this affects students’ 

STEM pathways after they transition to a postsecondary 

institution. More recent studies (2016 onward) analyzed similar 

data to that of this article but used different theoretical frameworks 

to explain racial disparities. Much like this thesis, previous 

research focused on students starting from high school to 

understand how math courses might affect STEM persistence at 

the postsecondary level. For example, Fong and Kremer (2020) 

used the expectancy value theory to examine math 

underachievement at the high school level and found that ‘math 

motivation’ was a significant predictor in underachievement 

which played a role in the students' transition to college. Riegle-

Crumb et al. (2019) used the opportunity hoarding framework in 

their study of underrepresentation of racial/ethnic groups in 

STEM and found STEM is a high value degree and due to 

opportunity hoarding, racial disparities continue to persist. The 

theoretical frameworks used in these studies were instrumental in 

providing relevant and significant findings, the categorical 

inequality framework will contribute to understanding why racial 

disparities continue to persist in STEM.  

Previous researchers used a myriad of theoretical 

frameworks to investigate and further understand why racial 

disparities exist in STEM (Fong and Kremer 2020; Andersen and 

Ward 2013; Domina and Saldana 2012; Archibald Farley-Ripple 

2012; Daempfle 2004; Crisp 2009; and Riegle-Crumb et al. 2019). 

However, this article uses the categorical inequality framework 

(Domina et al. 2018) to explore racial inequality in elite math 

courses in high school. In their review, Domina, Penner and 

Penner (2018) discuss and apply the theoretical framework of 

categorical inequality to education. Unfortunately, within the 

institution of education, schools have been sorting and 

categorizing students, placing them on track to higher education, 

technical schools, or elsewhere, leading them on a specific 

pathway. For working class students, female students and people 

of color, education is sometimes the only clear path towards a 

successful life.  



Rodriguez 5 

 

CALIFORNIA SOCIOLOGY FORUM VOL. 5 

By creating categories and sorting youth among 

them, schools develop templates that influence 

the contours of inequality throughout 

contemporary societies…The categorical 

inequality perspective draws attention toward the 

organizational processes through which schools 

create categories and sort individuals into them, 

and how, in doing so, they generate and reinforce 

social inequalities. (Domina, Penner, and Penner 

2018)  

By sorting students, educational institutions are creating 

and perpetuating inequalities; students who are sorted into lower-

level courses are at a disadvantage compared to students sorted 

into higher level courses. Applying the categorical inequality 

framework towards elite math course-taking and racial disparities 

in STEM gives insight as to the important role sorting students 

based on race and status has on the persistence of racial disparities.  

 

The Current Study  

This project examines the relationship between elite math 

courses and STEM pathways, as well as the role of sorting in math 

courses on racial disparities. In sum, the literature suggests that 

math course taking plays an important role in continuing on the 

STEM pathway. Thus, this thesis asks (1) What are the racial 

disparities in elite math course-taking in high school? After 

accounting for elite math courses taken, I ask the following 

questions: (2) What is the high school graduation rate across racial 

and ethnic groups? (3) What are the odds of ever attending a 

postsecondary institution across ethnic/racial subgroups? And 

after students have enrolled in a postsecondary institution, I ask 

(4) what are the odds of obtaining a degree in STEM across 

ethnic/racial subgroups, after accounting for elite math courses 

completed in high school?  

I expect that my findings for research question 1 will 

support previous research on math course taking and racial 

disparities (Tyson et al. 2007; Kelly 2009; Riegle-Crumb 2010; 

Irizarry 2021). I hypothesize African American and Latinx 

students would have lower odds of taking elite math courses in 

high school compared to White students. I hypothesize that 



Rodriguez 6 

 

CALIFORNIA SOCIOLOGY FORUM VOL. 5 

African American and Latinx students would be less likely to 

complete high school compared to their White counterparts even 

after accounting for elite math courses taken in high school.  

Previous research has demonstrated evidence of math 

course taking at the high school level playing a significant role in 

students’ STEM pathways after high school (Lichtenberger and 

George-Jackson 2013; Shi 2017; Riegle-Crumb et al. 2019; Fong 

2020; Sanabria 2020; Irizarry 2021). As previous studies suggest 

that taking elite math courses in high school affects students’ 

decision in attending a postsecondary institution and obtaining a 

degree in STEM. Therefore, this project proposes two more 

research questions: For my third research question, I hypothesize 

African American and Latinx students are less likely to attend a 

postsecondary institution compared to their White counterparts.  

Previous research has found that students sorted into 

different levels of math courses affect students’ decisions in 

choosing a STEM major (Irizarry 2021). Therefore, for my final 

research question, I hypothesize that among those who attend a 

postsecondary institution, African American and Latinx students 

are less likely to obtain a bachelor’s degree in STEM even after 

accounting for elite math courses taken in high school. As students 

enter postsecondary institutions, required courses for STEM 

majors affect African American and Latinx students persisting in 

STEM (Chen 2009).  

 

METHODS 

Data  

Data is drawn from the Educational Longitudinal Study 

of 2002 (ELS:02) by the National Center for Education Statistics 

(NCES) public-use data, specifically focusing on high school 

seniors graduating in spring 2004. The ELS (2002) is a 

longitudinal study that followed a baseline representative sample 

of approximately 17,500 tenth-grade students over ten years, 

beginning in 2002. The ELS data contains a nationally 

representative sample of high school seniors graduating in 2004 

and follows students’ trajectories into enrollment and degree 

completion at postsecondary institutions. As a longitudinal panel 

study, ELS experienced sample attrition and non-response bias. To 

adjust for the sampling frame, the ELS:02 replenished the sample 



Rodriguez 7 

 

CALIFORNIA SOCIOLOGY FORUM VOL. 5 

with additional respondents. I adjust for attrition by using a 

sampling weight and only including students who are non-missing 

on key outcome, predictor, and control variables in my analyses.  

ELS (2002) collected a wide range of individual-level 

details from students, including race, gender, socioeconomic 

backgrounds, family housing composition, and family educational 

background. ELS surveyed students through four waves: base 

year (BY) during the students’ tenth-grade year in 2002, first 

follow up (F1) during the student's senior year in 2004, second 

follow-up (F2) in 2006 two years after high school, and third 

follow up (F3) occurred in 2012 (eight years after high school). 

High school transcripts were collected in the first follow-up (F1) 

and postsecondary transcripts were collected in 2012 in the fourth 

follow-up (F4).  

 

Measures  

I examine whether racial disparities in elite math course-

taking and persist across a range of academic outcomes: high 

school postsecondary attainment, and STEM degree completion. 

The key independent variable is race and ethnicity. The categories 

under race and ethnicity in the ELS (2002) data were American 

Indian/Alaska Native, non-Hispanic; Asian, Hawaii/Pacific 

Islander, non-Hispanic; Black or African American, non-

Hispanic; Hispanic, no race specified; Hispanic, race specified; 

More than one race, non-Hispanic, and White, non-Hispanic. Due 

to the small sample of American Indian/Alaska Native and More 

than one race, these categories were combined into one category 

and categorized as “other.” For this study, White, non-Hispanic, is 

the reference category due to their overrepresentation in the 

STEM fields and workforce (Pew Research Center 2021).  

Given that this paper investigates the role of elite math 

course taking on future student outcomes, the four dependent 

variables used in this thesis are whether the student has taken and 

elite math course (pre-calculus or higher), high School diploma or 

GED equivalent, enrollment at a postsecondary institution, and a 

bachelor’s degree in STEM.  

Demographic variables will be student-level covariates, 

which include gender, socioeconomic status, family composition, 

math item response theory (IRT) scores in senior year, and parents’ 



Rodriguez 8 

 

CALIFORNIA SOCIOLOGY FORUM VOL. 5 

highest level of education. For Research Questions 2, 3 and 4, elite 

math course is added to the complex models as a control. In 

addition, models for Research Questions 3 and 4 only include 

students who have obtained a high school diploma (or 

equivalent).  

 

Sample  

As shown in Table 1 (n = 17,500), the sample was fairly 

evenly split by gender consisting of 48.85% male and 50.15% 

female students, with male students being the reference category. 

The racial composition consisted of White (62.22%), Hispanic, 

race specified (8.46%), Hispanic, no race specified (6.6%), Black 

(13.34%), Asian (4.51%), and Other (4.86%) students. 

Socioeconomic status (SES) is measured in quartiles with students 

fairly evenly distributed across quartiles. Family composition was 

categorized as: ‘mother and father’ (60.08%), ‘female guardian 

only’ (1.22%), with ‘mother only’ (17.66%), and the following 

categories were combined: ‘mother and male guardian’, ‘father 

and female guardian’, ‘two guardians’, ‘lives with student less 

than half time’, and ‘father only, male guardian only’ (17.73%). 

Parent’s highest level of education was divided into three 

categories: ‘Did not finish high school’ and ‘graduated from high 

school or GED’ (25.03%); ‘Attended 2-year, no degree,’ 

‘Graduated from 2-year school,’ attended college no 4-year 

degree’ (34.21%); ‘graduated from college,’ ‘completed Master’s 

degree or equivalent,’ ‘completed Ph.D., MD, other advanced 

degree’ (34.78%). Students’ math scores are on a continuous scale 

ranging from 13.74 to 83.03 (std=13.84). This variable measures 

the probability of students correctly answering, “each of the items 

in the pool” (NCES 2014).  

The key dependent variables: elite math course, high 

school attainment, postsecondary enrollment, and bachelor’s 

degree in STEM were collapsed into dichotomous variables. The 

categories for ‘elite math course’ were ‘no math course or math 

course is other,’ ‘pre-algebra, general or consumer math,’ ‘Algebra 

I,’ ‘Geometry,’ and ‘Algebra II’ which were combined (54.03%) 

and my reference category and trigonometry, pre-calculus, or 

calculus as the reference category (45.08%). High school 

attainment was categorized as: ‘successful graduate (HS diploma 



Rodriguez 9 

 

CALIFORNIA SOCIOLOGY FORUM VOL. 5 

recipient),’ ‘marginal graduate (HS diploma recipient),’ 

‘Completer (GED/equivalency/certificate of attendance),’ and 

‘non-completer.’ The categories were collapsed into ‘completer’ 

(90.24%) and ‘non-completer’ (2.6%) with non-completer as the 

reference category. The categories for postsecondary enrollment 

were already dichotomous ‘has some postsecondary enrollment’ 

(87.9%) and ‘no postsecondary enrollment’ (12.1%) which was 

the reference category. The final key dependent variable 

‘Bachelor’s degree in STEM’ was collapsed into ‘STEM’ 

(10.25%) and ‘non-STEM’ (31.99%) categories (57.76% were 

missing or NA); non-STEM was the reference category.  

 

Analytical Strategy  

Given that the key dependent variables are dichotomous 

(has taken an elite math course or not), I use logistic regression to 

examine (1) racial disparities in elite math course taking by senior 

year of high school; after accounting for elite math course taking: 

I examine the odds of (2) high school attainment (3) 

postsecondary enrollment and (4) STEM degree attainment. With 

each research question and corresponding key dependent variable 

I examine the students’ transition into a postsecondary institution 

and whether students earned degrees in STEM. For ease of 

interpretation, logit coefficients from the analyses are 

exponentiated into odd ratios that are then interpreted as either an 

increase (>1) or decrease in odds (<1) of the outcome variable 

occurring. The analyses start with a baseline model to estimate the 

association between elite math course-taking and race: 

 

 
 

 (𝐸𝐿𝐼𝑇𝐸)𝑖 estimates the odds of taking an elite math 

course by senior year of high school for every student. I first 

examine the odds of taking an elite math course across racial and 

ethnic subgroups. In the second model, I include the following 

student-level covariates: gender, socio-economic status, family 

composition, parent’s highest level of education, and prior math 

achievement, represented as 𝑋𝑖. The second model with added 

controls examines whether elite math courses have a distinct 



Rodriguez 10 

 

CALIFORNIA SOCIOLOGY FORUM VOL. 5 

relationship with racial disparities in STEM relative to other 

significant predictors. 

  

 

 

Equation 2 predicts the odds of high school attainment 

across racial/ethnic groups, controlling for whether the student 

had taken an elite math course by senior year. Then, the second 

model includes controls. Analyses for Research Questions 3 and 4 

follow a similar equation as Equation 2 to predict the odds of 

attending postsecondary education and STEM degree attainment.  

 

RESULTS  

Table 2 reports the odds ratios for taking an elite math 

course by the students’ senior year of high school. Model 1 (n = 

13, 300) reports the baseline model without controls showing the 

odds of taking an elite math course across racial and ethnic 

categories. We can see that Asian, Hawaii/Pac. Islander students 

had twice the odds (β = 2.0; p < 0.001) of taking an elite math 

course compared to their white counterparts. Black, non-Hispanic 

(β = 0.6; p < 0.001), Hispanic, race specified (β = 0.5; p < 0.001), 

Hispanic, no race specified (β = 0.4; p < 0.001), and Other (β = 

0.7; p < 0.001) students had significantly lower odds than White 

students of taking an elite math course by their senior year.  

Model 2 in Table 2 (n = 13,000) shows the odds of taking 

an elite level math course by senior of high school with the 

following covariates: gender, SES, Family composition, prior 

math achievement, and parents’ education. We see in Model 2 that 

Black, non-Hispanic (β = 2.2; p < 0.001) and Asian, Hawaii/Pac. 

Islander (β = 2.4; p < 0.001) students have twice the odds of taking 

an elite math course compared to White students. Hispanic, no 

race specified, was no longer statistically significant after 

introducing controls. However, Hispanic, race specified, and 

Other were only marginally significant (β = 1.2; p < 0.10; β = 0.7; 

p < 0.10 respectively).  

Table 3 reports the odds ratio of completing high school 

by race/ethnicity after accounting for elite math course taking. 

Model 1 (n = 12,800) represents the baseline model without 



Rodriguez 11 

 

CALIFORNIA SOCIOLOGY FORUM VOL. 5 

controls, showing that the odds of completing High School were 

significantly lower for Black non-Hispanic (β = 0.4; p < 0.001), 

Hispanic no race specified (β = 0.4; p < 0.001), Hispanic race 

specified (β = 0.3; p < 0.001), and Other (β = 0.3; p < 0.001) race 

compared to White students; Asian, Hawaii/Pac. Island students 

were just as likely to complete High School as white students and 

was not statistically significant.  

After accounting for control variables in Model 2 (n = 

12,000), the odds of Asian students completing High school 

lowered slightly and were not significant (β = 0.9; p > 0.10). For 

Black (β = 0.6; p < 0.001), Hispanic, race specified (β = 0.5; p < 

0.001), and Other (β = 0.4; p < 0.001) students, there wasn’t much 

change from Model 1, the odds increased slightly and remained 

statistically significant. However, interestingly Hispanic, race 

specified, students were slightly more likely to complete High 

School than White students but lost significance in Model 2. The 

elite math course covariate was marginally significant (β = 0.7; p 

< 0,10).  

Table 4 reports odds ratios predicting enrollment at a 

postsecondary institution. Model 1 (n=10,400) shows the baseline 

model without controls, showing the odds of students ever 

attending a postsecondary institution by race. Hispanic, no race 

specified (β = 0.6; p < 0.001) and Other students (β = 0.6; p < 

0.001) had lower odds of ever attending a postsecondary 

institution; Black and Hispanic, race specified were not 

statistically significant. Asian students were more than twice as 

likely to attend a postsecondary institution than White students (β 

= 2.3; p < 0.00).  

After accounting for controls in Model 2 (n = 10,000), the 

odds of ever attending a postsecondary institution increased 

significantly for each category of students with the exception of 

Other students who were just as likely as white students to attend 

a postsecondary institution although this was not significant (β = 

1.0; p > 0.10). Model 2 shows Asian students have three times the 

odds (β = 3.0; p < 0.00) than White students to attend a 

postsecondary institution. Black (β = 2.1; p < 0.001), Hispanic, 

race specified (β = 1.8; p < 0.001), and Hispanic, no race specified 

(β = 1.8; p < 0.001), students were about twice as likely as White 

students to ever attend a postsecondary institution. The control, 



Rodriguez 12 

 

CALIFORNIA SOCIOLOGY FORUM VOL. 5 

family household composition was only marginally significant, 

and the parent's highest level of education was only statistically 

significant for the parents in the category High School/GED or 

less. Highest elite math course taken was statistically significant 

(β = 0.4; p < 0.00).  

Table 5 reports the odds ratios of attaining a bachelor’s 

degree in STEM. Model 1 (n = 4,700) shows the baselines without 

controls with White students as the reference category. Asian 

students had twice the odds of earning a STEM degree (β = 1.7; p 

< 0.001) compared to White students; all other categories were not 

statistically significant. After accounting for controls in Model 2 

(n = 4,500), the odds for earning a STEM degree compared to 

White students lowered slightly for Asian students but remained 

statistically significant (β = 1.5; p < 0.001). However, even after 

accounting for controls, Black, Hispanic, race specified, Hispanic, 

no race specified, and Other students were still not significant. 

Students who did not take an elite math course were less likely 

than students who took pre-calculus, calculus, and/or 

trigonometry to earn a STEM degree (β = 0.5; p < 0.00).  

When examining elite math course taking, I hypothesized 

that African American and Latinx students were less likely to take 

an elite math course in high school compared to their White 

counterparts, the results presented in table 2 shows that Model 2 

does not support my hypothesis. As previously stated, African 

American and Latinx students were twice as likely or just as likely 

to take an elite math course by their senior year of high school. 

This finding shows that students of color are slowly narrowing the 

gap in elite math course taking. For the second hypothesis: African 

American and Latinx students are less likely to complete high 

school compared to their counterparts (after accounting for elite 

math), Table 3 supports my second hypothesis. Apart from 

Hispanic, no race specified, African American and Latinx (when 

race is specified) students are less likely to complete high school 

compared to White students.  

Regarding postsecondary enrollment, I hypothesized that 

African American and Latinx students were less likely to ever 

attend a postsecondary institution compared to White students 

(after accounting for elite math course). Table 4 shows that the 

results do not support my hypothesis. Results show African 



Rodriguez 13 

 

CALIFORNIA SOCIOLOGY FORUM VOL. 5 

American and Latinx students are about twice as likely to ever 

attend a postsecondary institution compared to White students. 

Lastly, when looking at the first known bachelor’s degree, I 

hypothesized that African American and Latinx students are less 

likely to obtain a bachelor’s degree in STEM, after accounting for 

elite math course taking. Unfortunately, results were not 

statistically significant therefore I was not able to conclude if 

African American and Latinx students were more or less likely to 

obtain a bachelor’s in STEM.  

 

DISCUSSION & CONCLUSION 

This paper examined the effects of taking an elite math 

course in a nationally represented dataset and provided a better 

understanding of how elite course taking plays a role in persisting 

on the STEM pathway. This study shows that while students of 

color are enrolling in elite math courses in high school at similar 

or increased odds as White students, African American and Latinx 

students are shown to have lower odds of attaining a high school 

diploma. However, according to data, African American and 

Latinx students are more likely to attend a postsecondary 

institution but results on STEM degree attainment could not 

provide evidence of the odds of which students of color earned a 

bachelor’s in STEM. Unsurprisingly and consistent with prior 

research, Asian students are more likely to earn a bachelor’s 

degree in STEM compared to their White counterparts (Chen and 

Weko 2009; Kokkelenberg and Sinha 2010; Ma and Liu 2015).  

This project contributes to literature on racial disparities 

in STEM education and persisting on the STEM pathway, despite 

some of the thesis’ limitations. The first limitation was sample 

attrition; as with longitudinal designs, one of the issues is losing 

participants over time. Furthermore, another limitation was from 

analyzing data from ELS:02 public use data instead of their 

restricted data; unfortunately, some variables of interest were 

restricted. For example, students’ first choice of major and 

institution type (two-year vs. four-year) was inaccessible. In 

addition, access to students’ first choice of major would have 

provided important information on STEM attrition across 

racial/ethnic subgroups.  



Rodriguez 14 

 

CALIFORNIA SOCIOLOGY FORUM VOL. 5 

Previous research shows there is a relationship between 

math courses taken and STEM persistence (Chen 2013). While 

some of my findings aligned with prior research, others 

contradicted it. Findings suggest that African American and 

Latinx students were just as likely or more likely to take an elite 

math course by their senior year. This finding supports previous 

literature comparing students from different cohorts (1982, 1992, 

and 2004). Students in 2004 earned more math credits than the 

other two cohorts with the exception of Latinx students who did 

not show much difference from 1992 to 2004 (Dalton et al. 2007; 

Domina and Saldana 2012). This finding also supports Xie and co-

authors (2015) study who found that the racial gap in elite course 

taking in high school has narrowed.  

Future research should focus on why students of color are 

more likely to be sorted in lower-level math courses compared to 

White students (Xie et al. 2015). Furthermore, even though this 

thesis did not find any statistically significant results on African 

American and Latinx students’ odds of obtaining a bachelor’s in 

STEM, previous research found evidence of racial disparities in 

obtaining a degree in STEM (Chen and Weko 2009; Ma and Liu 

2015). More research should be conducted to these seemingly 

contradictory results. 

It is important for researchers and policymakers to 

understand why African American and Latinx students do not 

persist in STEM once they’ve enrolled in a postsecondary 

institution. Perhaps developing pre-college programs aimed at 

creating a smooth transition from high school to postsecondary 

institutions could lower attrition rates among these students. 

Sorting students also plays a significant role in STEM persistence 

as students of color, African American female students in 

particular were found to be less likely to have been recommended 

for elite math courses (Francis et al. 2019). Policies need to be 

developed to prevent bias against students of color who qualify 

for enrollment in elite math courses. On a structural level, perhaps 

school districts can consider making elite math courses 

mandatory. In most public schools in the United States math 

requirements usually go up to Algebra II or students need to fulfil 

only three years of math by making elite a requirement it might 



Rodriguez 15 

 

CALIFORNIA SOCIOLOGY FORUM VOL. 5 

work to motivate students or introduce them to math courses that 

are relevant and lay a foundation to a STEM education.  

Although some of the hypotheses were not supported by 

the results, the findings from this thesis contributes to literature 

and future research. For example, students of color were found to 

have taken elite math courses at equal and or double the odds as 

White students. Future research should continue research on elite 

math course taking and its relationship with STEM persistence. 

As previously stated, African American and Latinx students have 

higher odds of enrolling in a postsecondary institution, future 

research should be conducted to determine whether postsecondary 

enrollment in two-year vs. four-year institutions would be 

significant in persisting in STEM across racial/ethnic subgroups.  

The theoretical framework of categorical inequality helps 

explain students being sorted in elite level courses. Previous 

research utilized different theoretical frameworks when 

researching racial disparities in course taking and STEM. For 

example, Fong and Kremer (2020) used the expectancy value 

theory which focuses on the students, their perceptions of reality 

and how this affects their academic performance. Categorical 

inequality shifts the focus from the student to the educational 

institution by stating that schools create categories based on status 

and sort students accordingly. It focuses on sorting at the 

institutional and organizational levels and frames racial status as 

more salient in elite math courses.  

When analyzing data from ELS:02 I found that students 

are taking elite math courses at similar rates. However, racial 

disparities continue to persist in STEM education and workforce. 

Future researchers should utilize this framework to understand 

how sorting students in elite math courses versus non elite math 

courses at the secondary level and its effects of STEM persistence. 

 

REFERENCES 

Andersen, Lori, and Thomas J. Ward. 2012. “Expectancy-Value 

Models for the STEM Persistence Plans of Ninth-Grade, 

High-Ability Students” A Comparison Between Black, 

Hispanic, and White Students.” Science Education 

98(2):216-242.  



Rodriguez 16 

 

CALIFORNIA SOCIOLOGY FORUM VOL. 5 

Adelman, Clifford, Bruce Daniel, Ilona Berkovits, and Jeffrey 

Owings. 2003. “Postsecondary Attainment, Attendance, 

Curriculum, and Performance: Selected Results from the 

NELS: 88/2000 Postsecondary Education Transcript 

Study (PETS), 2000.” Washington D.C.: National Center 

for Education Statistics.  

Archbald, Doug, and Elizabeth N. Farley-Ripple. 2012. 

“Predictors of Placement in Lower-Level High School 

Mathematics.” The High School Journal 96(1):33-59.  

Ballon, Estela Godinez. 2008. “Racial Differences in High School 

Math Track Assignment.” Journal of Latinos and 

Education 7(4):272-287.  

Champion, Joe, and Vilma Mesa. 2016. “Factors Affecting 

Calculus Completion among U.S. High School Students.” 

Pp. 9-25 in The Role of Calculus in the Transition from 

High School to College Mathematics, edited by David M. 

Bressoud. The Mathematics Association of America.  

Codiroli, Natasha. 2015. “Inequalities in students’ Choice of 

STEM subjects: An exploration of intersectional 

relationships.” Centre for Longitudinal Studies. Institute 

of Education: University College London.  

Chen, Xianglei. 2013. STEM Attrition: College Students’ Paths 

into and out of STEM Fields. Statistical Analysis Report. 

National Center for Education Statistics.  

Chen, Xianglei, and Thomas Weko. 2009. “Students Who Study 

Science, Technology, Engineering, and Mathematics 

(STEM) in Postsecondary Education.” Stats in Brief. 

National Center for Education Statistics.  

Crisp, Gloria, Amaury Nora, and Amanda Taggart. 2009. "Student 

characteristics, pre-college, college, and environmental 

factors as predictors of majoring in and earning a STEM 

degree: An analysis of students attending a Hispanic 

serving institution." American Educational Research 

Journal 46:924–42.  

Daempfle, Peter A. 2004. “An Analysis of the High Attrition Rates 

Among First Year College Science, Math, and 

Engineering Majors.” Journal of College Student 

Retention 5(1):37- 52.  



Rodriguez 17 

 

CALIFORNIA SOCIOLOGY FORUM VOL. 5 

Dalton, Ben, Steven J. Ingels, Jane Downing, and Robert Bozick. 

2007. Advanced Mathematics and Science Coursetaking 

in the Spring High School Senior Classes of 1982, 1992, 

and 2004. Washington, DC: National Center for 

Education Statistics.  

Domina, Thurston, Andrew Penner, and Emily Penner. 2018. 

“Categorical Inequality: Schools as Sorting Machines.” 

Annual Review of Sociology 43:311-330.  

Domina, Thurston, Andrew M. Penner, and Emily K. Penner. 

2016. “‘Membership Has Its Privileges’: Status 

Incentives and Categorical Inequality in Education.” 

Sociological Science 3:264-295.  

Domina, Thurston, Joshua Saldana. 2012. “Does Raising the Bar 

Level the Playing Field? Mathematics Curricular 

Intensification and Inequality in American High Schools, 

1982-2004.” American Educational Research Journal 

49(4): 685-708.  

Educational Longitudinal Study (ELS): 2002-2012. Washington 

D.C.: National Center for Education Statistics.  

Fong, Carlton J., and Kristen P. Kremer. 2020. “An Expectancy-

Value Approach to Math Underachievement: Examining 

High School Achievement, College Attendance, and 

STEM Interest.” Gifted Child Quarterly 64(2):67-84.  

Francis, Dania, Angela C. M. de Oliveira, and Carey Dimmitt. 

2019. “Do School Counselors Exhibit Bias in 

Recommending Students for Advanced Coursework?” 

Journal of Economic Analysis and Policy 19(4):1-17.  

Fry, Rick, Brian Kennedy, and Cary Funk. 2021. “STEM Jobs See 

Uneven Progress in Increasing Gender, Racial and Ethnic 

Diversity.” Washington D.C.: Pew Research Center.  

Funk, Cary, and Kim Parker. 2018. “Diversity in the STEM 

Workforce Varies Widely Across Jobs.” Washington D.C.: 

Pew Research Center. 

Irizarry, Yasmiyn. 2021. “On Track or Derailed? Race, Advanced 

Math, and the Transition to High School.” Socius: 

Sociological Research for a Dynamic World 7:1-21.  

Kelly, Sean. 2009. “The Black-White Gap in Mathematics Course 

Taking.” Sociology of Education 82:47-69.  



Rodriguez 18 

 

CALIFORNIA SOCIOLOGY FORUM VOL. 5 

Kokkelenberg, Edward C., and Esha Sinha. 2010. Who Succeeds 

in STEM Studies? An Analysis of Binghamton University 

Undergraduate Students. Economics of Education Review 

29:935-46.  

Lichtenberger, Eric, and Casey George-Jackson. 2013. 

“Predicting High School Students’ Interest in Majoring in 

a STEM Field: Insight into High School Students’ 

Postsecondary Plans.” Journal of Career and Technical 

Education 28(1):19-38.  

Long, Mark C., Dylan Conger, and Patrice Iatarola. 2010. “Effects 

of High School Course-Taking on Secondary and 

Postsecondary Success.” American Educational Research 

Journal 49(2):285-322.  

Ma, Yingyi, and Yan Liu. 2015. “Race and STEM Degree 

Attainment.” Sociology Compass 9(7):609-618.  

McEaching, Andrew, Thurston Domina, and Andrew Penner. 

2020. “Heterogeneous Effects of Early Algebra across 

California Middle Schools.” Journal Of Policy Analysis 

and Management 39(3):1-29.  

Peralta, Claudia, Melissa Caspary, and Diane Booth. 2013. 

“Success Factors Impacting Latina/o Persistence in 

Higher Education to STEM Opportunities.” Cultural 

Studies of Science Education 8:905-918.  

Pew Research Center. 2018. “Diversity in the STEM Workforce 

Varies Widely Across Jobs.” Washington D.C.: Pew 

Research Center.  

Price, Joshua. 2010. “The Effect of Instructor Race and Gender on 

Student Persistence in STEM Fields.” Economics of 

Education Review 29:901-910.  

Riegle-Crumb, Catherine. 2006. “The Path through Math: Course 

Sequences and Academic Performance at the Intersection 

of Race-Ethnicity and Gender.” American Journal of 

Education 113(1):101-122.  

Riegle-Crumb, Catherine, Barbara King, and Yasmiyn Irizarry. 

2019. “Does STEM Stand Out? Examining Racial/Ethnic 

Gaps in Persistence Across Postsecondary Fields.” 

Educational Researcher 48(3):133-144.  

Riegle-Crumb, Catherine, and Eric Grodsky. 2010. “Racial-Ethnic 

Differences at the Intersection of Math and Course-taking 



Rodriguez 19 

 

CALIFORNIA SOCIOLOGY FORUM VOL. 5 

and Achievement.” Sociology of Education 83(3):248-

270.  

Sanabria, Tanya, and Andrew Penner. 2017. “Weeded Out? 

Gendered Responses to Failing Calculus.” Social 

Sciences 6(2):47-60.  

Sanabria, Tanya, Andrew Penner and Thurston Domina. 2020. 

“Failing at Remediation? College Remedial 

Coursetaking, Failure and Long-Term Student 

Outcomes.” Research in Higher Education 61:459–484.  

Seo, Eunjin, Yishan Shen, and Edna C. Alfaro. 2018. 

“Adolescents’ Beliefs about Math Ability and Their 

Relations to STEM Career Attainment: Joint 

Consideration of Race/ethnicity and Gender.” Journal of 

Youth and Adolescence 48:306-325.  

Shi, Qi. 2017. “English Language Learners’ (ELLs) Science, 

Technology, Engineering, Math (STEM) Course-Taking, 

Achievement and Attainment in College.” Journal of 

College Access 3(2):45-61.  

Sovero, Veronica, Moshe Buchinsky, and Matthew D. Baird. 

2021. “Playing Catch Up: A term level Investigation of 

the Racial Gap in STEM Retention.” Economics of 

Education Review 83:1-18. 

Tyson, Will, Reginald Lee, and Kathryn M. Borman. 2007. 

“Science, Technology, Engineering, and Mathematics 

(STEM) Pathways: High School Science and Math 

Coursework and Postsecondary Degree Attainment.” 

Journal of Education for Students Placed at Risk 

12(3):243-270.  

Wesley, Jeffrey. 2019. “Crossing the Finish Line? A Review of 

College Completion Inequality in the United States by 

Race and Class.” Sociology Compass. 14(5):1-19.  

U.S Department of Education. 2018. “Paths Through Mathematics 

and Science: Patterns and Relationships in High School 

Coursetaking.” Washington D.C.: National Center for 

Education Statistics.  

U.S Department of Education. 2018. “2015-16 Civil Rights Data 

Collection: STEM Course Taking.” Washington D.C.: 

Office for Civil Rights.  



Rodriguez 20 

 

CALIFORNIA SOCIOLOGY FORUM VOL. 5 

Xie, Yu, Michael Fang, and Kimberlee Shauman. 2015. “STEM 

Education.” Annual Review of Sociology 41:331-357  

 

APPENDIX 

 

 
 



Rodriguez 21 

 

CALIFORNIA SOCIOLOGY FORUM VOL. 5 

 

 
 

 

 

 



Rodriguez 22 

 

CALIFORNIA SOCIOLOGY FORUM VOL. 5 

 

 
 

 



Rodriguez 23 

 

CALIFORNIA SOCIOLOGY FORUM VOL. 5 

 



Rodriguez 24 

 

CALIFORNIA SOCIOLOGY FORUM VOL. 5 

 

 
 


