Common Method Bias in Survey Research: Warning Signs, Impacts and Statistical Remedies
S. Sathyanarayana
MPBIM, Karnataka, India.
*Author to whom correspondence should be addressed.
Abstract
Common Method Bias (CMB) is a measurement error that occurs when correlations among variables are inflated due to the use of the same data collection method rather than the actual relationships between the constructs. The current paper offers an in-depth examination of Common Method Bias (CMB), an important problem encountered in behavioural and social sciences. The paper begins by defining the concept of CMB and explaining its implications in empirical studies, particularly in survey-based research. It further identifies major red flags that indicate the possible presence of CMB in research design and data collection procedures. The study also discusses various sources of common method bias, including respondent-related factors, measurement context, item characteristics, and data collection methods. In addition, the paper examines the consequences of CMB, such as inflated relationships among variables, misleading statistical inferences, and reduced construct validity. To address these concerns, both procedural and statistical remedies are discussed in detail. Procedural remedies include questionnaire design improvements and temporal separation, while statistical remedies include Harman’s single-factor test, marker variable techniques, and common latent factor approaches. The research paper ends with a discussion section that emphasizes the importance of recognizing and handling CMB in order to maintain scientific validity in scholarly research. The paper also recommends that future empirical studies addressing the proposed framework incorporate procedural and statistical remedies to minimize common method bias and improve the validity of findings.
Keywords: Self-reported survey, common method bias, Harman’s single factor, scale design, common latent factor