Using Single and Multiple Informant Research Designs to Address Common Method Variance Problem in Work Team Research

Abstract

A multiple-informant approach is typically recommended to address the problem of common method variance (CMV) in cross-sectional and single-level designs, but this approach is seldom empirically tested in work team research. Therefore, this study aims to assess the relative benefits of single- versus multiple-informant designs in terms of resolving CMV bias in work team research. Data from 1,991 participants in 176 bank branch teams in Northeast Thailand were analyzed using the confirmatory factor analysis-multitrait-multimethod model. The results indicate that when employing a different array of measurement scales, a multiple-informant survey (design 1) provides little incremental value compared to a single-informant survey in terms of reducing the threat of CMV bias. This paper concludes by offering guidelines to assist work team researchers in deciding whether and how to apply a multiple-informant survey approach.