| Abstract: | Leadership remains a popular and heavily researched area in the social sciences. Such popularity has led to a proliferation of new constructs within the leadership domain. Here, we argue that such construct proliferation without pruning is unhealthy and violates the principle of parsimony. Our purpose was to examine construct redundancy via a comprehensive review of task-oriented, relational, passive, and inspirational leader behaviors as well as values-based and moral leadership behaviors. Our findings, as indexed via meta-analytic correlations, reveal that construct redundancy remains problematic for the leadership literature. In addition, many of the values-based and moral behavior models correlated heavily with constructs traditionally examined as outcome variables (e.g., trust, LMX, justice). Implications for future research are discussed in regards to construct redundancy and how to avoid endogeneity bias in primary studies in the leadership literature. |
