Selection Protocol for Replication in Sports and Exercise Science

Abstract

Introduction - To improve the rigor of science, experimental evidence for scientific claims ideally needs to be replicated repeatedly with sufficiently similar procedures to increase the collective confidence in the veracity of those claims. Large replication projects in psychology, cancer biology and social science have evaluated the replicability of their fields but no collaborative effort has been undertaken in sports and exercise science. We propose to undertake such an effort here. As this is the first large replication project in this field, there is no agreed-upon protocol for selecting studies to replicate. Criticism of the previous selection protocols include claims they were non-randomized and non-representative, and alleged to be biased. Any selection protocol in sports and exercise science must be unbiased and representative to provide an accurate estimate of replicability of the field. The aim of this document is to produce a protocol for selecting studies to replicate for inclusion in a large replication project in sports and exercise science. Methods - The proposed selection protocol uses multiple inclusion and exclusion criteria for replication study selection, including the year of publication and citation rankings, research disciplines, study types, the research question and key dependent variable, study methods and feasibility. Studies selected for replication will be stratified into pools based on instrumentation and expertise required and will then be allocated to volunteer laboratories for replication. Replication outcomes will be assessed using a multiple inferential strategy and descriptive information will be reported regarding the final number of included and excluded studies, and original author contact.