Primary Submission Category: Mediation Analysis, Mechanisms
Causal Mediation and Functional Outcome Analysis with Process Data
Authors: Youmi Suk, Chan Park,
Presenting Author: Youmi Suk*
Over the past two decades, there has been growing interest in analyzing the effects of educational programs on outcomes using process data from computer-based testing and learning environments. However, most analyses focus on final outcomes measured at the end of a test or session, overlooking their functional nature over time. Such analyses fail to capture the dynamic causal mechanisms associated with functional outcomes. To address this limitation, this paper proposes a novel causal framework for identifying and estimating functional average treatment effects, functional natural direct effects, and functional natural indirect effects, along with their subgroup effects. Building on the literature on causal mediation and moderation, we define these effects using potential outcomes and provide nonparametric identification strategies. We then develop estimation methods using generalized additive models, a flexible and robust tool for analyzing functional data. The proposed approach is applied to examine the effects of extended time accommodations (ETA) on two functional outcomes—test scores and item access—in large-scale educational process data. In this analysis, students’ disability status serves as a moderator. This application uncovers the dynamic causal mechanisms underlying the effects of ETA on outcomes and highlights when and for whom each effect works during the testing period.