Group Sequential Designs Applied in Psychological Research

Authors

  • Klemens Weigl
    Department of Psychology, Catholic University of Eichstätt-Ingolstadt, Eichstätt, Germany; Human-Computer Interaction Group, Technische Hochschule Ingolstadt, Ingolstadt, Germany; Institute for Health Services Research and Clinical Epidemiology, School of Medicine, Philipps-Universitaet Marburg, Marburg, Germany
  • Ivo Ponocny
    Department for Sustainability, Governance, and Methods, MODUL University Vienna, Vienna, Austria

Abstract

Psychological research is confronted with ever-increasing demands to save resources such as time and money while assuring high ethical standards. In medical and pharmaceutical research, group sequential designs have fundamentally changed traditional statistical testing approaches featuring only one analysis at the end of a single-stage study. They enable early stopping at an interim stage, after a group of observations, for efficacy or futility in case of an overwhelmingly large or small effect, respectively. Otherwise, the trial is continued to the next stage. On average over many studies time and money are saved and more ethical trials are facilitated by diminishing the risk of patients' exposure to inferior treatments. We provide an easy-to-use tutorial for psychological research replete with easily understandable figures highlighting the core idea of different group sequential designs, a workflow chart, an empirical real-world data set, and the annotated R code. Finally, we demonstrate the application of early stopping for efficacy.