Maximizing the Value of Learning Assistants for Students' Belonging, Mattering, and Confidence in Gateway STEM Courses.

Learning Assistants (LAs) can reduce psychosocial barriers in gateway science courses. In this study, we specify the impact on students from different demographic groups. An inductive analysis of group interviews with 33 students from introductory biology and general chemistry courses revealed six narratives that capture students' experiences with course structures, people, and personal connections. All students valued varied, flexible resources, but other narratives differed across groups. Underrepresented racial/ethnic and first-generation students described active and collaborative learning as promoting belonging, confidence, and mattering. Instructor interactions were helpful, with underrepresented and first-generation students highlighting the importance of establishing course climate. LAs were crucial for belonging among underrepresented students, first-generation students, and continuing-generation women, but not for continuing-generation men. Shared academic identities were universally important, while shared gender and racial identities held greater significance for underrepresented and first-generation students. Experiences with peers varied, with negative peer interactions and fear of judgment having less impact on well-represented, continuing-generation students. These findings extend Dewsbury's Deep Teaching model by demonstrating how LA integration into STEM courses can support student experiences across diverse populations. The results also reveal how instructors and LA programs can adapt their practices to support all students, ultimately enhancing belonging, confidence, and mattering.