Research participation effects: a skeleton in the methodological cupboard

Abstract

Objective: There have been concerns about impacts of various aspects of taking part in research studies for a century. The concerns have not, however, been sufficiently well conceptualized to form traditions of study capable of defining and elaborating the nature of these problems. In this article we present a new way of thinking about a set of issues attracting long-standing attention.

Study design and setting: We briefly review existing concepts and empirical work on well-known biases in surveys and cohort studies and propose that they are connected.

Results: We offer the construct of "research participation effects" (RPE) as a vehicle for advancing multi-disciplinary understanding of biases. Empirical studies are needed to identify conditions in which RPE may be sufficiently large to warrant modifications of study design, analytic methods, or interpretation. We consider the value of adopting a more participant-centred view of the research process as a way of thinking about these issues, which may also have benefits in relation to research methodology more broadly.

Conclusion: Researchers may too readily overlook the extent to which research studies are unusual contexts, and that people may react in unexpected ways to what we invite them to do, introducing a range of biases.

Keywords: Bias; Cohort studies; Hawthorne effect; Mixed methods; Research assessment; Research methods; Research participation; Surveys.

Fig. 1

The research process. Cross-sectional surveys end with baseline assessment, cohort studies also involve follow-up assessment(s) only, RCTs involve randomization to study conditions as described previously. RCT, randomized controlled trial.