Multilevel random coefficient analyses of event- and interval-contingent data in social and personality psychology research

Increasingly, social and personality psychologists are conducting studies in which data are collected simultaneously at multiple levels, with hypotheses concerning effects that involve multiple levels of analysis. In studies of naturally occurring social interaction, data describing people and their social interactions are collected simultaneously. This article discuses how to analyze such data using random coefficient modeling. Analyzing data describing day-to-day social interaction is used to illustrate the analysis of event-contingent data (when specific events trigger or organize data collection), and analyzing data describing reactions to daily events is used to illustrate the analysis of interval-contingent data (when data are collected at intervals). Different analytic strategies are presented, the shortcomings of ordinary least squares analyses are described, and the use of multilevel random coefficient modeling is discussed in detail. Different modeling techniques, the specifics of formulating and testing hypotheses, and the differences between fixed and random effects are also considered.