A latent class analysis of cultural wealth: insights into academic success in introductory biology at a Hispanic-serving institution

ABSTRACT

Success in STEM majors often depends on students’ ability to navigate gateway courses, such as introductory biology, which establishes foundational knowledge and predicts retention in the major. However, disparities in performance within these courses often reflect systemic inequities rather than differences in ability. This study explores the role of cultural wealth, as defined by Yosso’s Community Cultural Wealth (CCW) framework, in shaping academic outcomes. Using data from 518 biology majors at a Hispanic-serving institution, we employed latent class analysis to identify distinct subgroups of students based on their cultural wealth profiles. Four latent classes emerged, characterized by varying levels of engagement across the CCW dimensions. Class 1 exhibited the highest cultural wealth but did not achieve the highest grades, highlighting the potential influence of unexamined mediators. Class 2, with moderate responses across dimensions, achieved the highest grades, suggesting a potential interaction of cultural wealth and external supports. Classes 3 and 4, with lower overall cultural wealth, exhibited lower academic performance. These findings reveal the complex interplay between cultural wealth and academic success in biology education.

INTRODUCTION

Success in STEM majors often hinges on completing rigorous “gateway” courses. For biology majors, introductory courses play a dual role in laying the foundation for advanced coursework and serving as predictors of student retention in the major (1, 2). However, large class sizes and limited variation in instructional approaches often characterize these courses, potentially disadvantaging students from diverse backgrounds (3). Despite efforts to implement interventions and changes to course structure with varying degrees of success (4, 5), the need to explore the interaction between student characteristics and experiences remains of interest (6, 7).

Higher education students enter their institutions with diverse forms of capital shaped by their backgrounds and experiences. Yosso’s Community Cultural Wealth (CCW) framework (8) outlines six forms of cultural capital: familial, aspirational, navigational, social, linguistic, and resistant. Together, these forms of capital represent assets that students draw upon to navigate educational spaces. Importantly, these forms of capital can influence academic performance by shaping students’ engagement, resilience, and sense of belonging in educational environments (9). For example, aspirational capital can motivate students to persevere despite obstacles (10). Similarly, linguistic and navigational capital equip students to communicate effectively and navigate institutional systems, both of which are critical in the context of challenging STEM courses (11).

Recent studies have explored the dimensionality of CCW, highlighting how these forms of capital manifest uniquely among students (12, 13). Yet, not enough is known about how these particular dimensions interact with the institutional context to influence academic outcomes in biology courses. Understanding this interplay is particularly crucial in gateway courses such as introductory biology, where disparities in performance may often reflect systemic inequities rather than differences in ability alone. Furthermore, for many students, gateway courses are not only foundational academically but also represent initial and critical testing grounds of their ability to leverage their cultural wealth. Lacking forms of wealth, such as navigational capital, can have long-lasting repercussions in navigating this new context in the first year of a student’s higher education. In contrast, leveraging aspirational capital can keep students motivated to persist in a program despite challenges that arise, thus emphasizing the importance of cultural wealth.

This study uses latent class analysis (LCA) to identify subgroups of students based on their cultural wealth profiles and explores how these profiles relate to academic outcomes in an introductory biology course. The research questions guiding this study are

• What latent classes of students emerge based on their cultural wealth as measured by a CCW survey?

• How do these latent classes relate to academic performance as measured by final grades in an introductory biology course?

By uncovering latent class structures, we aim to provide insights into the subgroups of students that may enter institutions based on their cultural wealth, so that departments may better consider how to support them in pursuit of their degrees.

METHODS

Context

This study took place at a large R1 Hispanic-serving institution in the southeastern United States. Two-thirds of the student population at this institution identifies from a Hispanic or Latin* background, with the vast majority identifying as Cuban. In addition, about three-quarters of the first-year students commute to the university from the surrounding counties. The majority of the students who attend also identify as continuing generation, with most acknowledging knowing a family member who has previously attended an institution of higher education. With a large proportion of Hispanic and Latin* students, the institutional culture is highly reflective of this with sights of Latin*-themed events and practices being commonplace.

Data collection

Participation in the study was voluntary, and for participating, students received extra credit in their course. Informed consent was collected through the survey, and students could elect to complete an alternate assignment for the extra credit if they preferred not to complete the survey. All students in the study were biology majors enrolled in an introductory biology course. The survey was administered at the beginning of the Fall 2023 and the Spring 2024 semesters. As such, the data set utilized for the results reported here is also utilized by Eden and Dewsbury (13).

To measure the cultural wealth of the participants, an instrument validated by the research team was utilized (13). This instrument contained several 6-point Likert-scale questions aimed at measuring 10 dimensions of cultural wealth. These dimensions, as well as examples of how they might be displayed, are listed below with the original form of capital from which they derive listed in Table 1. The instrument also contained additional demographic questions aimed at capturing Latin* identity, first-generation status, gender identity, and sexual orientation identity. This instrument was utilized for this study as it was already validated with the same population. In turn, this allowed for responses to be trustworthy without the need to be re-validated. Finally, final grades in the introductory biology course were captured for all student participants. Responses were gathered during the fall 2023 and the spring 2024 semesters.

TABLE 1.

Cultural wealth dimensions

Cultural wealth dimension	Original form of cultural wealth based on Yosso (8)	Example
Familial Heritage and Support (FHS)	Familial capital	A student’s parents or extended family offering emotional reassurance and encouragement to persist through challenging biology coursework
Future Vision and Encouragement (FVE)	Aspirational capital	Students motivated by clear goals for future careers, such as aspiring to medical school, which encourages persistence in difficult courses
Navigational Mastery (NM)	Navigational capital	A student who proactively seeks out tutoring or leverages office hours to address gaps in understanding course content
Social Justice Conviction (SJC)	Resistant capital	A student who actively engages in campus initiatives promoting equitable representation and inclusivity within STEM education
Networked Support System (NSS)	Social capital	Students forming study groups or joining student organizations to share resources, support each other academically, and foster a sense of belonging
External Inspirations (EI)	Aspirational capital	A student inspired by a scientist or mentor whose success story motivates their persistence and engagement in challenging biology courses
Empowerment through Representation (ER)	Resistant capital	Students feeling empowered by instructors or curriculum materials that include scientists who share their racial or ethnic identities
Artistic and Creative Expression (ACE)	Linguistic capital	Students utilizing drawing, creative writing, or multimedia presentations to demonstrate understanding of complex biological concepts
Advocacy for Equity (AE)	Resistant capital	Students participating in advocacy efforts to address inequities in education, such as campaigns for curriculum reform to incorporate diverse perspectives
Adaptable Communication (ACM)	Linguistic capital	Students adept at switching between academic language used in coursework and colloquial language used in peer interactions, enabling them to build effective relationships across diverse settings

Quantitative analysis

Statistical analyses for this study were completed using JASP and the snowRMM package in jamovi software (14, 15). Following the collection of data, incomplete responses were removed. To explore if any latent groups existed among the participants, an LCA was conducted. An LCA allows researchers to capture subgroups that may exist among the data based on different characteristics or categorical patterns (16, 17).

To conduct the LCA, the survey items aimed at measuring cultural wealth were input into the model. Seven models were run with varying numbers of classes (from 2 to 8) until a model of best fit was identified. To determine the best fit, Bayesian information criterion (BIC) values were prioritized as they balance model fit with complexity (18); entropy values were also monitored to ensure separation between classes (19). Once a final model was selected, the demographics and cultural wealth of each class were explored to identify trends between the classes.

To explore how demographics grouped together into various classes, an analysis utilizing contingency tables was conducted (20). Specifically, the different demographics captured in this study were compared across classes to determine if there were any significant trends. To elaborate, for the analysis, students would be labeled as either “1” or “0,” denoting membership in a specific class. For example, if exploring gender identity distribution in Class 2 versus all other classes, that contingency table would have labeled members of Class 2 as “1,” and members of Class 1, 3, and 4 as “0.” Finally, to explore how class membership may be linked to final grade in their introductory biology course, a Kruskal-Wallis Test was conducted to compare the classes (21). A Kruskal-Wallis test was chosen due to the non-parametric distribution of some items and the ordinal nature of the Likert-scale responses.

RESULTS

The survey instrument was administered to six sections of introductory biology in the fall 2023 semester as well as three sections in the spring 2024 semester. The fall sections were each taught by a different instructor, with three of these instructors teaching the spring sections as well. Following the collection of data, a total of 531 responses were received. With the removal of any incomplete responses, the final number of participants was brought to 518. The total number of students enrolled in these sections was 1,086 for the fall semester and 777 for the spring semester. The overall response rate was approximately 28% of students enrolled. General demographics of the participants are shown in Table 2.

TABLE 2.

Demographics of participants in the study

Group	Subgroup	No. of participants (N = 518)
Gender	Female	384
	Male	125
	Non-binary	7
	Other	1
	Prefer not to say	1
Latin* identity	Latin*	407
	Non-Latin*	111
First-generation status	First generation	215
	Continuing generation	303
LGBTQ+ identity	Asexual	0
	Bisexual	18
	Gay	8
	Heterosexual	163
	Pansexual	0
	Queer	6
	Questioning	4
	Other	0
	Prefer not to specify	14

For the latent class analysis, seven different class models (two to eight classes) were evaluated to identify a model of best fit. After fitting based on BIC value, a four-class model was deemed to fit best. Respective fit indices for each of the models that were evaluated are shown in Table 3. The proportions of the sample that make up each class in the chosen model are 0.2597 (n = 135), 0.3951 (n = 205), 0.0695 (n = 36), and 0.2757 (n = 142), respectively. Figure 1 showcases characteristics of each class, including demographics, grade distribution, and the CCW dimension that was highest and lowest in each class.

TABLE 3.

Model comparison based on number of classes and fit indices

Class number	AIC	BIC	SABIC	Entropy	Log likelihood
2	75,187	77,571	75,790	0.965	−37,032
3	73,233	77,301	74,139	0.981	−35,775
4	72,683	77,177	73,891	0.982	−35,219
5	71,758	77,725	73,269	0.987	−34,475
6	71,727	78,888	73,539	0.991	−34,178
7	71,858	80,213	73,973	0.992	−33,963
8	72,782	82,332	75,200	0.990	−34,144

Fig 1.

(A) Characteristics of each class, including average final grade in their introductory biology course, the CCW dimensions that each class scored highest and lowest in, and some basic demographics. (B) Bubble chart highlighting the letter grade distribution by class. The number of students (N) in the respective letter grouping is shown in the bubble, with the size of the bubble highlighting the proportion within the class that receive that respective grade.

The contingency table analysis for demographics revealed significant results for Latin* membership in Class 2 (P = 0.009; Cramer’s V = 0.115) and Class 4 (P < 0.001; Cramer’s V = 0.164), sexual orientation identity for membership in Class 2 (P = 0.002; Cramer’s V = 0.199) and Class 4 (P < 0.001; Cramer’s V = 0.339), and for gender identity for membership in Class 4 (P = 0.030; Cramer’s V = 0.144). The contingency table revealed marginal significance (P = 0.076) for the first-generation demographic in Class 3, which was the only class to have more students classified as first generation than continuing generation. To explore final grades, a Kruskal-Wallis test was conducted to identify significant differences among the classes. Results from the Kruskal-Wallis test and Dunn’s post hoc analysis showed a significant difference in final grade between Class 2 (85.015) and Class 4 (80.437; P = 0.016), as well as Class 2 (85.015) and Class 3 (81.756; P = 0.027) (Fig. 2). While these averages may appear close, the individual breakdown of letter grade received highlights the contrasting outcomes within each class (Fig. 1B). For example, while Class 4 had an average grade in the B− range, the class contained a higher proportion of F’s and D’s compared to the other groups, while Class 2 had the highest proportion of A’s and B’s received.

Fig 2.

Class comparison of final grades based on Kruskal-Wallis results. Solid line denotes significance, while dashed line denotes non-significant results, with the respective P-value shown.

To explore the trends within each CCW dimension, average response choice by each class to each dimension was captured (Fig. 3). Class 1 consisted of students with the highest average response on 9 of the 10 dimensions, thus being a class with a tendency for “higher capital” responses. The lowest average response for each dimension was mostly split between Class 3 and Class 4. Class 2 tended to average between the highest and lowest averages except for Advocacy for Equity (AE), a form of resistant capital. Class 4’s highest dimension was AE. This class also happened to contain the largest in-class proportion of LGBTQ+ students, a historically marginalized group. Finally, across all four classes, Artistic and Creative Expression (ACE) had the lowest average for each respective class.

Fig 3.

Average response on the survey to each CCW dimension for each class.

While the demographic and grade comparison among the groups mostly revealed significant findings among Class 2 and Class 4, a comparison of the cultural wealth dimensions highlights several significant results. Specifically, all classes appeared to be significantly different across most of the dimensions when compared to each other. Table 4 shares the mean differences between the classes, with all significant results noted.

TABLE 4.

Mean differences between classes based on responses to CCW dimension questions^a

Class comparison	1–2	1–3	1–4	2–3	2–4	3–4
Familial Heritage and Support	0.482***	1.623***	1.532***	1.140***	1.050***	−0.091
Future Vision and Encouragement	0.044***	1.039***	0.869***	0.995***	0.825***	−0.170
Navigational Mastery	0.801***	1.494***	1.640***	0.693***	0.838***	0.146***
Social Justice Conviction	0.657***	1.382***	0.379***	0.725***	−0.278**	−1.003***
Networked Support System	0.777***	1.343***	1.554***	0.566***	0.777***	0.211
External Inspirations	0.654***	0.626**	1.340***	−0.028	0.686***	0.714**
Empowerment through Representation	0.866***	1.331***	1.012***	0.465**	0.146	−0.319*
Artistic & Creative Expression	0.164	−0.090	0.306*	−0.254	0.143	0.396*
Advocacy for Equity	0.866***	0.570	0.863***	−0.296*	−0.003	0.293*
Adaptable Communication	0.314***	1.103***	0.823***	0.789***	0.509***	−0.280*

^a ***P < 0.001; **P < 0.01; and *P < 0.05.

DISCUSSION

When pursuing a higher education, students enter this environment with the cultural wealth they acquired throughout their lives. This cultural wealth influences how they interact with the institutional context and can lead to varied academic outcomes. The LCA in this study identified four distinct subgroups among the student respondents based on their cultural wealth profiles, highlighting the complex relationship between cultural wealth and academic performance. Findings suggest that potential interactions with external factors may affect academic outcomes, as students with similar profiles experienced differing academic outcomes. For example, Class 1 emerged as the “high capital” group, scoring the highest average response on 9 of 10 dimensions of cultural wealth. However, this group did not achieve the highest final grades, suggesting that possessing cultural wealth alone may not be enough to find success. In other words, external factors or barriers (institutional policies, classroom environments, structural inequities, etc.) may mediate or inhibit the translation of cultural wealth into academic success. Future work may seek to identify these mediating factors to further understand and address barriers faced by this subgroup. For instance, qualitative explorations may allow students to share experiences that may be missed by quantitative data alone, thus providing direct evidence of systemic or institutional practices that may influence the translation of cultural wealth into success.

Class 2 achieved the highest final grade average and demonstrated moderate responses across most cultural wealth dimensions, except for AE, where they scored lower than Class 4. The “balanced” profile of Class 2 further highlights the potential interplay between cultural wealth and some form of external support or influence, which may have contributed to their academic success. For example, supportive institutional practices or support networks may assist students in accessing resources necessary for success, such as social capital honed through connections with certain institutional offices and administrators. This finding aligns with previous research emphasizing the role of institutional support in fostering STEM success (5, 22). In turn, this underscores the importance of identifying institutional scaffolding that complements and activates the cultural wealth of students in meaningful ways. Exploring the specific actions and experiences that these students take while pursuing their education may identify the support that most complements students’ cultural wealth.

In contrast to Class 2, Classes 3 and 4, characterized by lower average responses across most dimensions, showed poorer academic outcomes. Notably, Class 4, with a higher proportion of LGBTQ+ students, had the lowest average grade and a disproportionately high number of D’s and F’s. Additionally, Class 3 had a majority of first-generation students (56%), and thus their lower average responses further emphasize the disparity in how cultural wealth is acquired and the variations that exist between populations. Similarly, an inconsistency was observed between Classes 2 and 3, which share similar highest and lowest cultural wealth dimensions yet differ significantly in academic outcomes. This discrepancy again suggests that the mere presence of specific cultural wealth dimensions does not uniformly translate into academic success. Potential factors such as differential access to mentoring, academic support services, classroom climate, or inclusive pedagogies may enhance the ability of some students to leverage their cultural wealth while limiting others. For instance, even students with similar cultural wealth profiles might experience varying degrees of recognition, validation, or resource accessibility within their classrooms or broader institutional environment. To elaborate, a Hispanic-serving institution with a critical mass of Hispanic or Latin*-identifying students might have structures in place that aim to validate the culture of their student population. However, without intentional reflection and action, they may inadvertently validate a portion of their population more than others, such as by catering to Cuban experiences and inadvertently ignoring a smaller subset of students from other backgrounds. Future research should explicitly explore these potential discrepancies and institutional mediators, investigating the precise institutional practices or supports that enable certain students to translate their cultural assets into academic achievement while posing barriers for others.

Across all groups, ACE consistently scored as the lowest dimension, suggesting a potential undervaluation of this form of linguistic capital in biology education. While our study interprets this as indicative of systemic undervaluation, this interpretation remains speculative. Future research should explicitly examine whether biology classrooms actively provide or deny opportunities for students to leverage ACE. Recognizing best practices for integrating this form of expression into biology may allow students who struggle in traditional formats to succeed by leveraging creative expression. For example, past studies have explored how integration of creative expression, such as through poetry, can positively impact student experiences in science (23, 24). Further exploration of other forms of creative expression can highlight the most effective strategies to implement in pedagogical practices. As we strengthen our understanding of ACE, discussions regarding the undervaluation of other forms of cultural wealth can be deepened. Additionally, by promoting creative expression, STEM fields such as biology may be able to attract new students who tend to see the sciences as limiting when it comes to how they can express themselves, especially when compared to disciplines in the humanities and arts.

Ultimately, one overarching theme emerges from the data: possessing cultural wealth alone is insufficient to find academic success in the higher education context. Institutions must go beyond acknowledging diversity to intentionally creating structures and practices that appreciate and activate student cultural wealth. Identifying the ways to most effectively activate cultural wealth will vary based on context and student population. Specifically, understanding the makeup of a specific student population is critical to effectively create culturally sustaining practices. This includes integrating inclusive pedagogies and fostering spaces for belonging and identity validation (25, 26). Additionally, incorporating culturally responsive strategies that align with students’ lived experiences can bridge the gap between cultural wealth and academic success.

Overall, this study is a reminder of the critical role institutions hold in creating equitable pathways for all students, especially in STEM fields such as biology. Institutions must intentionally adopt practices and create spaces that activate all dimensions of cultural wealth for students. Thus, future research should integrate qualitative approaches to enrich our understanding of the student experience, as it pertains to cultural wealth, by amplifying their voices and perspectives within specific contexts.

Limitations

Our study’s observational design limits causal inferences regarding the relationship between cultural wealth and academic outcomes. Furthermore, the relatively small sample size for each, especially for Class 3, constrains the generalizability and robustness of some class comparisons. Additionally, the absence of direct evidence regarding the study’s specific institutional practices’ alignment with specific dimensions of cultural wealth represents another significant limitation. As such, the interpretations made here are seen as hypotheses to lay the foundation for future work, especially those that seek to explore how specific institutional practices mediate the activation of cultural wealth among students.

Conclusion and future directions

This study highlights the complex relationship between cultural wealth and academic outcomes in introductory biology courses at a Hispanic-serving institution. While the findings illustrate distinct patterns of cultural wealth among student subgroups, these patterns did not consistently translate directly into academic success. As such, while cultural wealth often correlates with positive outcomes (27), it is not universally predictive of academic success, highlighting the nuanced interplay between students’ cultural wealth and institutional contexts. Our findings suggest that regardless of cultural wealth possession, institutional structures and practices, as well as other external factors, may have a varied impact on students, even within the same environment. The varied academic outcomes among classes that shared similar cultural wealth dimensions point to the possible existence of these mediating factors. Therefore, future research should explicitly investigate these mediators through qualitative methods like classroom observations or student interviews. Such approaches could provide richer insight into how specific institutional factors interact with students’ cultural wealth, potentially reinforcing or limiting academic success.

Additionally, the low responses in the ACE dimension indicate a potential systemic undervaluation of this dimension within the biology classrooms. This suggests an opportunity for educators to intentionally integrate pedagogical approaches that recognize and value diverse forms of cultural expression, particularly artistic and creative modes, to create more inclusive learning environments.

Future research should address the limitations inherent in this study’s observational design and smaller subgroup sample sizes. Specifically, longitudinal studies with larger samples, qualitative explorations, and classroom observations could offer richer, causal insights into how cultural wealth dimensions interact with institutional contexts to influence student outcomes. In conclusion, this study highlights the need for biology educators and institutional leaders at Hispanic-serving institutions—and beyond—to critically evaluate classroom and institutional practices, actively foster inclusive learning environments, and leverage the full range of students’ cultural wealth to enhance equity and academic achievement in STEM education.

ETHICS APPROVAL

This study was approved by the Florida International University Institutional Review Board (IRB-23-0428), with data collected used first to validate the quantitative instrument and later used for quantitative analysis.