Modeling directional and non-directional discrepancy between dependent variables

Bednall, Timothy Colin and Zhang, Yucheng Eason (2015) Modeling directional and non-directional discrepancy between dependent variables. Taneja, Sonia (ed.) In Academy of Management Proceedings. vol. 2015, Academy of Management.. (doi:10.5465/ambpp.2015.13998abstract).

Abstract

This paper proposes a new approach for modeling the antecedents of congruence in two dependent variables. In this approach, the difference between the dependent variables is decomposed into directional discrepancy and non-directional discrepancy, which are analogous to the concepts of accuracy and precision in measurement. The paper reviews previous approaches to congruence, including the models of Edwards (1995), McArdle (2009) and Cheung (2009). It then outlines how directional and non-directional discrepancy may be analyzed within a latent variable framework, and explains how this approach may be applied to these models. We explain how this approach resolves many of the shortcomings of past methods, and provides a more informative analysis of the nature of difference. We provide an illustrative example of our approach, and show how discrepancy in self and supervisor ratings of leadership effectiveness is influenced by each leader’s level of transformational leadership. Finally, the strengths and limitations of our approach are discussed, and we propose several directions for elaborating the basic latent directional / non-directional discrepancy (LDND) approach.

This record has no associated files available for download.

Contributors

Author: Yucheng Eason Zhang

Download statistics