A PROTOCOL FOR ANALYZING REPEATED MEASURES OF ONLINE GROUP BEHAVIOR

Abstract

In this article, we feature a novel protocol that enables the analysis of repeated measures of online group behavior. The protocol accounts for (1) the nested hierarchy of the data with weeks nested in persons, and persons nested in weeks, and (2) the temporal nature of the behavior at the early, mid, and late periods of each week. To manage and analyze such data in a general way, we first give an illustration of the data structure. Thereafter, we propose a five-step Courtney-Fanguy-Costley protocol that (1) considers the data structure, (2) defines the levels of data, (3) considers variable variation, timing, and necessary aggregation, (4) ensures necessary variation, and (5) specifies null and mixed-effects models. We also provide exemplary R code for readers to replicate our approach. •A general five-step protocol for analyzing repeated measures of online group behavior is offered. •A description of the complex nested data structure is offered. •Users can simulate data in R to run through the protocol.