Controlling for Common Method Variance in PLS Analysis: The Measured Latent Marker Variable Approach
Common method variance (CMV) continues to be an important issue for social scientists. To date, methodologists have yet to agree upon a best practice for detecting and controlling for CMV. In a recent paper, the unmeasured latent marker variable approach, a frequently employed technique, was shown to be incapable of detecting or controlling CMV in PLS analyses. Unfortunately, this was the only method to date suggested for handling CMV in PLS models. To fill this gap, we introduce a measured latent marker variable (MLMV) approach and demonstrate how it is able to both detect and correct for CMV when using Partial Least Squares.
Key wordsCommon method variance (CMV) Unmeasured latent marker variable Measured latent marker variable (MLMV)
- R. P. Bagozzi, “Measurement and meaning in information systems and organizational research: Methodological and philosophical foundations,” MIS Quarterly, 35, 261–292, (2011)Google Scholar
- A. Burton-Jones, “Minimizing method bias through programmatic research,” MIS Quarterly,33, 445–471, (2009).Google Scholar
- W. Chin, J. B. Thatcher, and R. T. Wright, “Assessing common method variance: Assessing the UMLC approach,” MIS Quarterly, 36, 1003–1019, (2012).Google Scholar
- R. Sharma, P. Yetton, and J. Crawford, “Estimating the effect of common method variance: The method pair technique with an illustration from TAM research,” MIS Quarterly, 33, 491–512, (2009).Google Scholar