Addressing Common Method Bias in Survey Datasets: A Literature Review and Future Research Directions

Abstract

This paper provides a thorough examination of techniques for detecting and mitigatingcommon method bias (CMB) in research studies. A thorough review of literature from1990 to 2024 using Scopus and Google Scholar as primary search engines, revealedmultiple methods of dealing with CMB. Despite significant contributions from seminalstudies, methodological limitations remain, as does the need for innovativemeasurement methodologies and statistical solutions. This study, which draws onfindings from key studies from the literature, addresses the need for novel strategiesto effectively combat CMB.The study investigates a variety of methodological techniques, including blinding,counterbalancing, longitudinal designs, and multimethod approaches, and proposesstrategies for reducing bias in data collection procedures. Confirmatory factor analysis,structural equation modelling, and multilevel modelling are some of the statisticaltechniques used to evaluate measurement validity and control for CMB. However,gaps still exist in the literature, particularly relating to accurately identifying andaddressing CMB across multiple datasets and research scenarios. Existingtechniques may fail to capture the complexities of method bias or provide reliablesolutions in all contexts. These limitations highlight the need for a new technique thatoffers a systematic and parametric approach to assessing and mitigating CMB,providing researchers with a comprehensive tool for increasing the validity andreliability of their findings. This study aims to imparts valuable insights for researchersseeking to improve the reliability and validity of their findings through a nuancedexamination of each technique's strengths, weaknesses, and practical implications.The proposed parametric mathematical method (Stacey-Qangule Model) provides asystematic approach for detecting and addressing Common Method Bias (CMB) insurvey data, with the goal of improving research validity. The method aims to identifylatent variables free of method bias across various datasets and scenarios byestimating bias in manifest ratings and applying mathematical transformations.Future research should focus on refining and validating the proposed statistical model,collecting diverse and high-quality data, conducting rigorous data analysis, effectivelyinterpreting and communicating findings, disseminating research results, and pursuingnew research directions. Implementing these recommendations allows researchers tocontribute to the advancement of knowledge in organizational behaviour andperformance evaluation, ultimately leading to positive change and impact in theidentified phenomena or problem areas