LEADERS IN THE LAB: A META-ANALYSIS OF LEADERSHIP RESEARCH GENERALIZABILITY AND ENDOGENEITY RISK

Contributing to ongoing debates regarding generalizability and rigor in leadership research, we explored whether differences existed in the results of matched research due to study in laboratory or field environments. In a series of meta-analyses including 439 effects from 265 independent studies, we considered whether similar research questions resulted in different results due to research setting effects. Laboratory and field studies were coded and matched based on theory-driven predictors (Dinh et al., 2014) and the type of outcome examined (Avolio et al., 2009). Further, we coded for the threat of endogeneity in each study design (Antonakis et al., 2010) and included that measure as a covariate moderator in order to account for methodological differences between studies and determine its influence on driving divergence in results. The overwhelming majority of relationships tested did not significantly differ as a product of research setting alone, suggesting broad generalizability between lab and field research. Endogeneity risk was found to significantly moderate meta-analytic effects in several cases, occasionally explaining enough variance in effects to make research setting a significant moderator when endogeneity was controlled for. Given the relevance of these findings to the study of leadership and social sciences more generally, additional implications are discussed.