Direct Replication in Experimental Communication Science: A Conceptual and Practical Exploration

Abstract

Replication is generally considered a keystone of the scientific enterprise. Unfortunately, in communication science, there is a lack of clarity on what a replication actually entails, and to what extent replicators may deviate from original studies. In order to support researchers in conducting, evaluating, and justifying the setup of replications of communication science experiments, we provide a taxonomy of replication types. We argue that researchers almost always need to adapt some elements of an original communication study to meaningfully replicate it. The extent to which deviations - ranging from mere updates to deliberate deviations and additions - are permissible, however, depends on the motivation behind conducting a replication study. We distinguish three basic motivations: verification of an original study's findings, testing the generalizability of an original study (which we further differentiate into the generalizability of study outcomes vs. theoretical claims), and extending an original study beyond the original goals. We argue that these motivations dictate what types of deviations are permissible and thereby determine the type of replication (i.e., direct, modified, and conceptual). We end with concrete recommendations for replicators: to specify the motivation to conduct a replication study and clearly label and justify any deviations from the original study for all study elements.