Abstract
Undergraduate research experiences (URE) and course-based UREs (CURE) vary in terms of what students do, and it is unclear whether CUREs can substitute as research experiences for UREs. We used ecological momentary assessment (EMA), or in-the-moment data collection, to identify and compare research tasks of CURE and URE students at nine universities. We texted surveys to two cohorts, receiving 3537 responses from 538 students in Cohort 1 and 10,862 responses from 711 students in Cohort 2. Qualitative analysis revealed that students' tasks fit 11 themes (planning, managing research, experimenting, working with data, iterating, learning, reading, writing, presenting, finalizing work, interacting with others), which mapped onto four topics (experimentation, data collection, data analysis, communication) derived using natural language processing and machine learning. CURE and URE students completed similar tasks in similar proportions and described engaging in their tasks actively and constructively, but not interactively. Our results suggest CUREs may substitute for UREs for engaging students in research tasks in university contexts and that both CURE and URE students may benefit from greater interactive engagement. Our study also demonstrates the feasibility of EMA for studying research experiences and the utility of topic modeling for analyzing large text datasets.