Teaching Constructive Replications in the Behavioral and Social Sciences Using Quantitative Data

Abstract

The replication crisis in the behavioral and social sciences spawned a credibility revolution, calling for new open science research practices that ensure greater transparency, including preregistrations, open data and code, and open access. Replications of published research are an important element in this revolution as part of the self-correcting process of scientific knowledge production; however, the teaching value of replications is still underutilized thus far. Pedagogical knowledge points to the value of replication as critical to the scientific method of test and retest. Psychology has already begun mass efforts to reproduce previous experiments. Yet, we have very few examples of how analytical and reanalysis replications, after the data come in, contribute to the reproducibility crisis and can be integrated into undergraduate and graduate courses. Replications with quantitative data can be a pedagogical tool for improving student research method skills and introducing them to best research practices via learning-by-doing. This article aims to start filling this gap by offering guidance to instructors in designing and teaching replications for students at various levels and disciplines in the social and behavioral sciences, including a supplementary teaching companion.