What does it mean to be "representative"?

TL;DR

This paper clarifies the concept of 'representative' in health research, defining it comprehensively and emphasizing the importance of clear generalization assumptions across different study types.

Contribution

It provides a formal definition of representativeness, distinguishing between generalizable estimates and interpretative generalization, with practical examples from COVID-19 studies.

Findings

Defined two types of representativeness: estimate-based and interpretation-based.

Applied the definition to four COVID-19 studies to illustrate its use.

Highlighted the importance of explicit assumptions for valid generalization.

Abstract

Medical and population health science researchers frequently make ambiguous statements about whether they believe their study sample or results are "representative" of some (implicit or explicit) target population. Here, we provide a comprehensive definition of representativeness, with the goal of capturing the different ways in which a study can be representative of a target population. We propose that a study is representative if the estimate obtained in the study sample is generalizable to the target population (either due to representative sampling, estimation of stratum specific effects, or quantitative methods to generalize or transport estimates) or the interpretation of the results is generalizable to the target population (based on fundamental scientific premises and substantive background knowledge). We explore this definition in the context of four COVID-19 studies, ranging from laboratory science to descriptive epidemiology. All statements regarding representativeness should make clear the way in which the study results generalize, the target population the results are being generalized to, and the assumptions that must hold for that generalization to be scientifically or statistically justifiable.