Design and Implementation of Reduced Longitudinal Experiments

• In Event: AERA Poster Session Friday 9:45am
  In Poster Session: Design and Implementation of Reduced Longitudinal Experiments - Longitudinal Studies Poster Session

Abstract

Longitudinal experiments are essential for investigating the impacts of interventions across time, but are often constrained by significant resource demands which limits their implementation. This study introduces and evaluates reduced-resource designs to enhance the feasibility of longitudinal experiments. Using Monte Carlo simulations, we assessed the statistical trade-offs of reducing sample size, measurement occasions, and the proportion of experimental units assigned to intervention groups to provide simplified but actionable recommendations for non-experts. We further assessed the robustness of these reductions to attrition, non-sampling errors, and violations to fundamental model assumptions. These findings aim to support pursuit of longitudinal methods and enable exploratory studies in resource-limited contexts such as teacher-led action research or student-driven preliminary scientific investigations.

Authors

• Le Grande Dolino, Indiana University

• Justin Philip Tuazon, University of the Philippines

• Leslie Rutkowski, Indiana University