From Micro to Meso: Critical Steps in Conceptualizing and Conducting Multilevel Research

Although interest in multilevel organizational theory, research, and methods has been on the rise in recent years, vigorous debates in the literature regarding appropriate ways to conceptualize and measure multilevel constructs, justify aggregation, and analyze multilevel models have contributed to confusion. New investigators interested in testing multilevel theory are intrigued, but wary. The goal of this article is to cut through the confusion, identifying the critical choices and issues a researcher may confront as he or she shifts from a single level to a multilevel perspective. The authors address four primary choices-construct and measurement issues, model specification, research design and sampling, and data analyses-describing critical steps in conceptualizing and conducting multilevel research.