An Examination of Perceptions and Usage of Regression Diagnostics in Organization Studies

Abstract
Few studies have evaluated the use of perhaps the most dominant analytic technique in organization studies-linear regression models. In this research, we investigate the extent to which researchers understand the assumptions and verify the appropriateness of linear regression modeling techniques prior to using them. After examining regression usage in leading journals and two samples of researchers in organization studies, we find that the assumptions and requirements of regression modeling are commonly neglected. We discuss consequences of disregarding these assumptions, the role of researchers' beliefs and understanding in not using diagnostics, and various diagnostic tools and statistical software which may be employed to assess regression assumptions. We conclude with a call for greater recognition of the importance of diagnostic checking and an increased application of diagnostics in organization studies.