Key points are not available for this paper at this time.
The goal of the current article is to conduct a widescale empirical investigation on the (in)efficacy of Harman's single-factor test by showing that the approach is insensitive to aspects of research design known to influence common method bias (CMB). Our systematic literature review of 1,619 sources demonstrates that the amount of variance explained by the first factor of Harman's single-factor test does not differ between cross-sectional and multi-wave, single-source and multi-source, or mono-method and multi-method studies. We instead find that extraneous aspects of studies influence the amount of variance explained, including the number of studied indicators and retained factors. These results therefore suggest that Harman's single-factor test is not a reliable assessment of CMB, and we hope these results prevent future researchers from applying the analysis. Our discussion concludes with alternative suggestions for identifying and addressing CMB, such as the application of sophisticated research designs and marker variable techniques.
Howard et al. (Tue,) studied this question.