Pitfalls in Pathways: Some Perspectives on Competing Risks Event History Analysis in Education Research

Abstract

A set of discrete-time methods for competing risks event history analysis is presented. The approach used is accessible to the practitioner and the article describes the strengths, weaknesses, and interpretation of both exploratory and model-based tools. These techniques are applied to the impact of “nontraditional” enrollment features (working, stopout, and delayed enrollment) on competing outcomes of sub-baccalaureate enrollment (degree attainment, transfer, and dropout). In this setting, model selection, estimation, and comparative inference are discussed and more general guidelines for each stage of analysis are provided. Substantively, the study finds that nontraditional enrollment features are associated with dramatic changes in expected outcomes.