Alternative sequential methods in statistical testing: A reply to Lakens (2021) and Erdfelder and Schnuerch (2021).

Abstract

We recently developed a simple and general sequential sampling method for testing null hypotheses, the independent segments procedure (ISP; Miller & Ulrich, 2021). In this reply, we discuss the comments of Erdfelder and Schnuerch (2021) and Lakens (2021), who consider alternative methods such as the sequential probability ratio test (SPRT) and the group sequential design (GSD), respectively. We evaluate the pros and cons of these alternatives and conclude that the ISP does have several advantages over these other methods, especially for psychological research. All of these sequential methods can save research resources because smaller sample sizes are required compared to standard nonsequential methods, so it seems appropriate for researchers to choose from a variety of sequential methods based on the practical requirements of their research. (PsycInfo Database Record (c) 2021 APA, all rights reserved).