Addressing common method variance and endogeneity in vocational behavior research: A review of the literature and suggestions for future research

Abstract

For various reasons, research on vocational behavior often has to rely on self-report measures. Although it is well known that the exclusive use of self-report measures in a given study can pose major threats to the validity of that study, for example via common method variance (CMV) or endogeneity, we still have to witness a systematic literature review of study designs particularly prone to CMV and endogeneity, and of procedural and statistical remedies used in the field of vocational behavior. To determine the vulnerability for CMV and endogeneity, and the extent to which researchers have already taken action and eased concerns over CMV and endogeneity in vocational behavior research through their research design/data collection (procedural remedies) and data analysis (statistical remedies), we review articles published from 2015 to 2018 in the Journal of Vocational Behavior. We found that 81% of all quantitative studies in this period relied exclusively on self-report measures gathered from the same respondent. Of these studies, the majority used procedural remedies to ease concerns of CMV (68%), mainly through the temporal separation of the focal variables (53%). However, statistical tests to detect and/or control for CMV (13%) are still rarely used, and we found no examples of techniques such as instrumental variable estimation to address potential endogeneity. To encourage and guide vocational behavior researchers to further minimize concerns over CMV and endogeneity arising from the exclusive use of self-report measures in future research, we summarize existing recommendations from the methodological literature and provide an updated discussion, specific to vocational behavior, on how to design and conduct impactful vocational behavior research.