View article

The potential inflation of correlations between measures assessed via the same method (eg, selfreport) is well known. This study applied CFA models to 24 multitrait-multimethod correlation matrices in order to assess the extent of common methods bias (CMB). While not trivial, CMB is often minor in magnitude

The influence of common methods variance (CMV) has been a pervasively cited concern in organizational research (Podsakoff, MacKenzie, Lee, & Podsakoff, 2003). As outgoing editor of Journal of Applied Psychology, John Campbell explicitly cited exclusive use of self report measures as an indication that a study contributes little to the literature (Campbell, 1982). Additionally, reviewers continue to cite CMV as a source of concern, particularly for research involving self-report measures (Spector, 2006). As a result, several researchers have investigated the extent to which CMV has biased correlations in psychological and organizational research (eg, Crampton & Wagner, 1994; Doty & Glick, 1998). While these studies prove insightful as to the nature and extent of the effects of common methods of measurement, they differ considerably in methodology and scope. In this study, we examined the extent to which CMV is pervasive in organizational research as well as the extent to which these methods are likely to have biased trait correlations.