Current Research Projects
-
Academic momentum, financial aid dynamics, and stop-out risk: Evidence from discrete-time hazard models
This study examines how academic momentum (measured by the credits attempted in the fall terms) and financial aid dynamics (Millenium Scholarship Status) shape student persistence and completion trajectories at a large public research university. Using a person-term panel of first-time, full-time (FTFT) cohorts, we estimate discrete-time hazard models of graduation timing and stop-out (defined as a two-consecutive-term enrollment gap requiring re-admission).
-
Engineering Pathway Project
This project reconceptualizes engineering persistence as a set of competing educational pathways and modeling these pathways using a discrete-time competing risks framework. Rather than focusing on a single persistence or graduation outcome, the analysis models transitions among multiple outcomes, including Graduate UNLV engineering; Graduate UNLV non-engineering; switching to non-engineering majors; transferring out to other institutions; stopping out without external enrollment, and multiple forms of continuous UNLV enrollment (Enrolled at six years, Engineering, no degree; Enrolled at six years, non-Engineering, no degree; and Enrolled at six years, unknown majors, no degree). Using longitudinal administrative data, the study tracks students over six academic years and examines how academic momentum, gateway course performance, major-specific academic performance, and financial aid dynamics jointly shape these transitions.
-
Diagnosing structural DFWI risk in gateway courses: A faculty-embedded visual analytics
Gateway courses are high-leverage points in the undergraduate curriculum because student performance in introductory sequences is closely tied to persistence, progression, and degree completion. Yet institutional course outcome reports are often static and aggregated, making it difficult for faculty to see where risk is concentrated or how to respond. This project presents a faculty-embedded visual analytics framework that converts administrative DFWI data into a diagnostic heatmap for instructional review. Using multi-term student-level enrollment and outcome data from a large public research university, the framework applies enrollment thresholds, rolling averages, and subject-level benchmarks to stabilize interpretation and distinguish persistent structural risk from cohort-specific fluctuation. The resulting interface surfaces risk patterns by subject, sequence position, and modality, enabling faculty and curriculum leaders to identify actionable bottlenecks and prioritize course-level interventions.
-
Early engagement indicators and student outcomes
This project examines how early patterns of student involvement and engagement are associated with key academic outcomes, including persistence, course performance, and longer-term success. Using available student data, such as advanced days of application submission, participation of Rebel-ready week, and Canvas login information during the first four weeks of a semester, the study seeks to identify which forms of engagement are most predictive of positive outcomes and which early indicators may signal risk. The goal is to generate actionable insights that can inform timely outreach, improve student support strategies, and help institutions better understand how to foster success from the start of the academic journey.
-
From silos to solutions: Unifying data systems to promote student success
Collaborating with colleagues from the Division of Student Affairs, this project focuses on integrating fragmented Student Affairs data systems into a more unified and usable infrastructure. By connecting data across units and reducing information silos, the project aims to create a more complete view of student experiences, needs, and progress. This improved data integration is intended to strengthen early identification of students who may need support, enhance coordination among campus partners, and enable more timely, informed, and effective student success interventions.
-
Leveraging Canvas data to optimize student engagement and academic performance
This project uses Canvas learning management system (LMS) data to examine patterns of student and faculty engagement and their relationship to academic performance. By analyzing behavioral indicators such as login frequency, assignment access, submission timing, page views, and participation in course activities, the study seeks to identify meaningful signals of student persistence, difficulty, and success. The findings are intended to support the development of actionable indicators that can help instructors and institutional leaders detect students who may benefit from timely support, improve course design, and strengthen data-informed strategies for promoting student achievement.
-
Modeling course evaluation completion behaviors over student careers and estimating the policy impacts
This project examines how students' course evaluation completion behaviors evolve across their academic careers and how institutional policy changes may influence participation. Using longitudinal student-level data, the study explores variation in completion patterns over time, including how engagement differs by class standing, academic context, and prior response history. It also evaluates the extent to which policy design features such as timing, reminders, incentives, or access requirements shape response rates and overall participation. The goal is to generate evidence that can inform more effective course evaluation policies, improve the quality and representativeness of feedback, and strengthen institutional decision-making related to teaching and learning.
-
Grading leniency and reliability and student evaluation of teaching
This project examines the relationship between faculty grading leniency, grading reliability, and student evaluations of teaching. It investigates whether patterns in grading practices are associated with changes in evaluation outcomes and whether those patterns systematically shape how students rate instructional quality. The work also considers the implications of grading behavior for the interpretation of student evaluations, with the goal of improving how evaluation data are used in faculty assessment, personnel review, and instructional improvement.
-
Integrating Canvas engagement data and student evaluations of teaching to predict academic performance
This project combines Canvas engagement data with student evaluations of teaching to examine whether these measures together improve prediction of academic performance. The study aims to identify stronger, more informative indicators of student outcomes.