UC Irvine

First-generation, low-income, underrepresented minority (URM) and female undergraduates are matriculating into science, technology, engineering, and math (STEM) majors at unprecedented levels. However, a disproportionate number of these students end up graduating in non-STEM disciplines. Attrition rates have been observed to spike in conjunction with introductory STEM courses in chemistry, biology, and physics. These “gateway” courses tend to be housed in large, impersonal lecture halls. First-generation and URM students struggle in this environment, possibly because of instructors’ reliance on lecture-based content delivery and rote memorization. Recent social psychological studies suggest the problem may be related to cultural mismatch, or misalignment between independent learning norms typical of American universities and interdependent learning expectancies for first-generation and URM students. Value-affirming and utility-value interventions yield impressive academic achievement gains for these students. These findings overlap with a second body of literature on culturally responsive instruction. Active gateway learning practices that emphasize interactive instruction, frequent assessment, and epistemological instruction can be successful because of their propensity to incorporate values affirming and utility-value techniques. The present study observed instruction for gateway STEM courses over a three-year period at the University of California, Irvine (N = 13,856 undergraduates in 168 courses). Exploratory polychoric factor analysis was used to identify latent variables for observational data on gateway STEM instructional practices. Variables were regressed on institutional student data. Practices implemented in large lecture halls fall into three general categories: Faculty-Student Interaction, Epistemological Instruction, and Peer Interaction. The present study found that Faculty-Student Interaction was negatively associated with student outcomes for female and first-generation students; and Epistemological Instruction was negatively associated with student outcomes for Hispanic students. More importantly, Peer Interaction was positively associated with student outcomes for female, first-generation, and Hispanic students. Study implications and limitations are discussed with reference to the research literature.