Derby, Lilly; Connell Bohlen, Lauren; Michie, Susan; Johnston, Marie; Birk, Jeffrey L; Rothman, Alexander J; Cornelius, Talea; (2024) Linking measures to mechanisms of action in behavior change: A qualitative analysis of expert views. Social Science & Medicine , 352 , Article 117023. 10.1016/j.socscimed.2024.117023.

Text Michie_Expert Consensus Qualitative.pdf Access restricted to UCL open access staff until 31 May 2025. Download (230kB)

Abstract

Rationale: Testing mechanisms of action (MoAs) hypothesized to underlie behavior change can enhance intervention effectiveness. Rigorous measurement of putative mechanisms is critical to this effort, but measures are rarely validated with respect to target MoAs. // Objective: This study aimed to elucidate challenges of linking measures to putative MoAs and to identify priorities for future research. // Method: This study was a systematic exploration of written comments by experts in behavioral intervention research and theories of behavior change (N = 20) capturing their opinions about a task querying whether self-report measures from the Science Of Behavior Change (SOBC) Measures Repository were related to a set of MoAs identified by the Human Behaviour Change Project (HBCP). // Results: Six themes were identified: 1) Study Value, 2) Measure Properties, 3) Mechanism Properties, 4) Miscellaneous Measure Concerns, 5) Conceptual Challenges, and 6) Approaches to Developing Measure-Mechanism Links. Experts noted challenges such as lack of measure validation, poor measure properties (e.g., double-barreled items), overly broad MoA definitions that limited their utility, lack of clarity around the term “related,” and more. Nonetheless, experts expressed the importance of the exercise. Suggestions included developing and refining measures that are validate for assessing MoAs, clarifying and elaborating MoA definitions, and conducting further, more granular research. // Conclusion: This systematic examination of expert comments highlights issues that need further investigation to advance behavioral science, specifically pertaining to identifying valid measures of MoAs in behavioral and process research. This study highlights the challenges and opportunities for future research on linking measures and MoA in behavioral science and subsequently enhancing the efficacy of behavioral interventions.

Download activity - last month

Export as XMLCSVJSON

Download activity - last 12 months

Export as JSONCSVXML

Downloads by country - last 12 months

Export as JSONCSVXML

Archive Staff Only

View Item