Exploratory graph modeling of longitudinal change in students undergoing biomedical research training’s characteristics

Understanding the experiences and development of undergraduate students undergoing biomedical research training is foundational for promoting equity and inclusion for underrepresented students. Many underrepresented students do not report the same levels of major indicators (i.e., science identity, research self-efficacy, and academic self-concept) for research and academic success in science, technology, engineering, mathematics, and medical (STEMM) fields, and this affects their retention to continue in the major and advancement to higher-level graduate degrees in STEMM. However, the development and changes underlying the key indicators for undergraduates involved in biomedical research training are complex and poorly understood. In this presentation, we will discuss the structural characteristics of students undergoing research training in biomedical sciences at both the onset and conclusion of the program. Making use of exploratory graph analysis (EGA) framework, which is a suitable approach to use for estimating the number of dimensions present in psychological multivariate data, we will summarize the changes over time in three key indicators: science identity, research self-efficacy, and academic self-concept. The results will be reported for the overall population and subpopulations defined by gender and ethnicity.

Our study analyzes survey responses of 235 students participating in the BUILDing Scholars program. The BUILDing Scholars program aims to increase the representation of Hispanics in the biomedical research workforce to mirror the nation’s demographics. As part of the BUILDing Scholars program, students received tuition, a stipend, and the opportunity to conduct research with a faculty mentor during the academic year and at various summer research programs. The majority of the participant students self-identified as Hispanics. Matching pairs of student scores are compared across time points two or more years apart. Results indicate structural differences in the development and growth among the three indicators by gender and by ethnicity. EGA models estimate symmetric graphical networks with dimensions present in the data to help visualize the relationships among the survey items. Comparing the early-stage and later-stage networks side by side allows us to see the changes in associations among survey questions over time. The results of this study provide insight into how these three key indicators co-regulate and develop over the course of a biomedical research-training program focused on the inclusion of Hispanic students.
Attendees of this presentation will learn how to better understand underrepresented minorities and their interest in STEMM degrees and undergraduate research. This presentation will help researchers, change agents, and administrators better understand underrepresented students in undergraduate STEMM research. We hope this presentation will open the conversation about how undergraduate research affects students’ science identity, research self-efficacy, and academic self-concept, and inform approaches for institutional change to promote underrepresented students in STEMM research.