Multilevel Modeling: Research-Based Lessons for Substantive Researchers

Abstract

Organizations are multilevel systems. Most organizational phenomena are multilevel in nature, and their understanding involves variables (e.g., antecedents and consequences) that reside at different levels. The investigation of these phenomena requires appropriate analytical methods: multilevel modeling. These techniques are becoming increasingly popular among organizational psychology and organizational behavior (OPOB) researchers. In this article we review the literature that has evaluated the performance of multilevel modeling techniques to test multilevel direct and indirect effects and cross-level interactions. We also provide guidelines for OPOB researchers about the appropriate use of these techniques, and we suggest ways these techniques can contribute to theoretical advancement and research development in OPOB.