The term 'research participation effects' (RPEs) is intended to capture features and artefacts of study design that may affect measured outcomes in ways that introduce bias into research findings, impacting inference and outcome validity. This systematic review aims to identify which RPEs have been studied in the context of alcohol research and provide an overview of estimates of RPEs on self-reported alcohol consumption. This systematic review summarises the available evidence on research participation effects in alcohol research. Twenty-seven reports were included in the review. The reports included randomized controlled trials, studies-within-a-trial, between-subjects experiments, and qualitative investigations. A range of RPEs were addressed: assessment reactivity (N=15), being randomized to a waiting list control group (N=3), the impact of obtaining informed consent (N=2), experimentally induced social desirability (N=3), and the Hawthorne effect either specifically by name (N=2, one quantitative, one qualitative) or described as general RPE presence (N=2). The literature provided proportionally stronger evidence in favour of assessment reactivity and waiting list designs affecting alcohol outcomes contra obtaining informed consent or inducing social desirability. Variation in study quality, terminology, and outcome measures hinder comprehensive understanding and discussion of RPEs at present. Improved knowledge of RPEs and their potential long-term consequences in alcohol research, including a unified lexicon, would enhance trial design and improve on the certainty of evidence in alcohol research.
Read full abstract