Research indicates that social networks are particularly important to the university experiences of U. S. student military service members/veterans (SSM/Vs), a diverse population of adult learners. While these students bring unique experiences and skills to campus, their transitions into higher education are frequently complicated by service-related impairments, cultural conflicts between military and civilian life, and a strong sense of social isolation. Despite the importance of SSM/V social networks, few studies have investigated their characteristics nor how these characteristics may associate with military experience or beneficial outcomes. Using an analysis of survey data from 1,255 SSM/V and civilian students across four U. S. universities, this study creates an inductive typology of egocentric networks. Following a process established by previous researchers, we first use unsupervised Random Forest to develop several foundational egocentric characteristics—from over fifty compositional and structural variables—that iteratively delineate nine university student network types fitting nearly all sampled networks. Second, we test relationships between student network types and background characteristics, including military experience, to present a richer portrait of how SSM/V egocentric networks differ from those of civilian students as well as to gauge the typology’s face validity. Third, we test whether network types associate with feelings of campus belonging, an important predictor of student university fit and success. Results are valuable to an emerging body of scholarship focused on the development and influence of marginalized college student social networks, as well as to literature centered on the sociocultural experiences of SSM/Vs in institutions of higher education.

Scholars studying postsecondary education are often interested in the antecedents, nature, and consequences of social relationships between agentic actors, such as students, faculty, staff, and administrators, in higher education. Decades of research have documented how sociodemographic characteristics, such as race/ethnicity, sex/gender, and socioeconomic status, inform students’ experiences in the classroom and broader campus community. Moreover, theorists often posit that psychosocial elements, such as students’ self-efficacy beliefs and their sense of belonging in the classroom and university community, are intertwined with sociodemographic influences on their learning experiences in higher education.

Recently, scholars have critiqued the ways that sociodemographic characteristics are conceptualized and measured in quantitative research broadly, and social networks research specifically. For example, some scholars have noted that racial groups are often conflated, such as studies wherein Black, Latino/a/x, and Indigenous students are categorized as “underrepresented” or “historically excluded” in STEM education research. Others have noted that race is often conceptualized using broad, binary categories that assume race is immutable. These critiques note that such categorization schemes risk essentializing within and across racial groups, where important social distinctions shape students’ lived experiences. Moreover, other scholars posit that such categorization schemes are ill-suited for social networks research because they are often used as proxies for the very social processes that are at the center of social networks research.

Researchers have responded to these critiques in various ways. Notably, scholars frequently call for the disaggregation of broad racial categories into finer grained racial categories. However, conceptualizing race in network studies, which are fundamentally about modeling social processes, is about more than disaggregating racial categories. Instead, researchers must ensure that models reflect context-sensitive, socially meaningful representations of the social context under examination.

The purpose of this research was to examine the relationship between students’ socio-cognitive beliefs, such as their local sense of belonging in specific classrooms and/or course-specific self-efficacy beliefs, on their formation of socio-academic relationships in large-lecture courses in STEM. We also modeled sociodemographic characteristics, such as students’ race/ethnicity, gender, and international student status, to understand the role of sociodemographic characteristics on students’ socio-academic relationships. We found that sociocognitive elements, such as students’ local sense of belonging to the classroom, were statistically significant predictors of students’ formation of socioacademic ties in large lecture classrooms.

However, the central focus of this talk will be to discuss the theoretical foundations of our context-sensitive racial categorization scheme. Namely, we drew on qualitative data from a research study at the institution, as well as existing frameworks that reflect diasporic discourses to understand how elements of students’ various racial, ethnic, and cultural backgrounds shape the ways students come to and experience higher education. We find that, without careful conceptualization of racial categorization schemes, our models failed to capture statistically significant racial homophily effects on students’ formations of socioacademic relationships. We will discuss the implications of this for social networks research in higher education.

Using Network Ecology (Mcfarland, et al, 2014) and Campus Ecology Frameworks (Brown & Smith, 2024), I illustrate how students in informal socio-academic relationships make changes to configurational and categorical norms to facilitate learning in undergraduate education. Drawing on case studies (n=34 ego-networks) of undergraduate science classrooms, I conceptualize individuals engaged in significant bridging and bonding work to create space and time for learning interactions as central figures in ecological context. I draw on ecological theory to characterize these individuals as keystone agents. Keystone species in biological and cultural ecologies play an essential role in maintaining health and function of ecological arrangements. In campus ecologies, keystone agents maintain teaching and learning relationships among campus actors through emergent social organization strategies. These actors are different from Stanton-Salazar's (2011) conceptualization of institutional agents as they have informal roles in network ecologies, most often students supporting other students to cultivate time, space, and relationships through which learning and development interactions can occur. As a corollary, these students often possess significant institutional knowledge that they share with their peers to help with institutional navigation (including course selection, advice on how to engage with faculty, and strategies for coursework management). When these individuals exit an ecological context, the local network ecology often suffers (or even collapses) as a result. For example, in community colleges, when a keystone agent transfer, their role in the ecology changes and if other individuals do not step into the vacant space, relationships and routines often are abandoned. I present some defining features of what makes a keystone agent in learning ecologies and offer strategies for how institutions might cultivate and support these students, as well as approaches for supporting local ecologies after students depart.

Despite efforts to support first generation (FG) students at large research universities in the United States, FG retention rates still do not keep up with the rates of continuing generation (CG) students. A FG student is one with neither primary parent/guardian earning at least a bachelor’s degree. In 2021, it was reported that 20% of FG students completed at least a bachelor’s degree compared to 60% of CG students. As prior research shows, students who come from families with prior university experience are more likely to attend a university and graduate. Part of this disparity stems from the lack of access FG students have to people who hold knowledge about higher education. As of academic year 2015-16, 56% of undergraduates nationally were FG college students, of which 25% are Latine students. With these statistics, it is critical for higher education to better understands FG Latine (FGL) student university experiences through a variety of analyses.

This paper shares results from a longitudinal (4-year) study that sought to broadly answer, how do first generation students’ personal university networks impact their retention and success? More specifically, this paper reviews an important result from a year three sample of FGL students (n=19). Analysis from this sample’s data revealed high student affairs staff presence and details into the relationships they maintain with FGL students. Specifically, behaviors and support types that are necessary for supporting FGL students to and through graduation. By highlighting the important and different role student affairs staff achieve in retaining and supporting FGL students, we can better inform higher education institutions on how to enhance their student programming to include the FGL experience and voice; effectively leading efforts to decrease persistent inequities amongst the FGL student population.