Although the topic of measurement equivalence/invariance (ME/I) is an ongoing concern in organizational research, we have little understanding of how accurate the analytical procedures involved in tests of ME/I really are. In particular, there is a need to investigate the susceptibility of ME/I tests to such contextual influences as common method variance and construct reliability. The present study addresses this need by using a Monte Carlo simulation to manipulate configural and metric invariance across two groups along with varying levels of method bias and reliability, and we assess how successful ME/I tests are at detecting group differences under such conditions. Based on these results, the likelihood of drawing false inferences from ME/I tests is evaluated. Finally, recommendations are presented for dealing with potential common method bias in organizational research, and avenues for future research are suggested.
|State||Published - 2006|
|Event||66th Annual Meeting of the Academy of Management, AOM 2006 - Atlanta, GA, United States|
Duration: Aug 11 2006 → Aug 16 2006
|Conference||66th Annual Meeting of the Academy of Management, AOM 2006|
|Period||08/11/06 → 08/16/06|
- Method Variance