Updating Our Understanding of Doctoral Student Persistence: Revising Models Using Structural Equation Modeling to Examine Consideration of Departure in Computing Disciplines.

Given a myriad of recent contemporary challenges graduate students are facing as well as long-standing issues with student attrition, there is a pressing need to reexamine models of doctoral student progress. While existing research commonly examines departure or failure to meet milestones as the outcome of interest, by the time students leave their programs, it is too late for faculty or departments to offer meaningful interventions. Drawing from Girves and Wemmerus's conceptual framework for doctoral student degree progress, we leverage more recent literature to propose and test an adapted conceptual framework among a recent nationwide sample of computing doctoral students in the United States. Findings illuminate the importance of psychosocial factors (e.g., sense of belonging, researcher self-efficacy), positive perceptions of departmental community, and the role of faculty advisors in reducing doctoral students' consideration of departure. Implications for departmental efforts and faculty advising practices that can honor students' agency and identity in the computing Ph.D. process and that may affect change in doctoral education are discussed. [ABSTRACT FROM AUTHOR]