representativeness

Representativeness refers to the degree to which a sample, study, or model reflects the characteristics of the population it is intended to describe. In statistics, representativeness affects external validity or generalizability: conclusions drawn from a sample should apply to the population of interest. A representative sample typically mirrors key attributes such as age, gender, ethnicity, geographic distribution, and other variables relevant to the research question. Achieving representativeness commonly involves probability sampling methods such as simple random sampling, stratified sampling, or cluster sampling. Non-representative samples can introduce bias and distort estimates, and may arise from nonresponse, self-selection, measurement error, or frame coverage issues. Weighting and design adjustments can mitigate some discrepancies, but cannot fully substitute for a truly representative sample.

In psychology and behavioral science, representativeness also refers to the representativeness heuristic: the tendency to judge the probability of events or hypotheses by how closely they resemble known cases or stereotypes, rather than by base rates. This can lead to systematic errors such as base-rate neglect or the conjunction fallacy.

Limitations: perfect representativeness is rarely achievable, and researchers must balance practicality with validity. Transparency about sample characteristics and the use of sensitivity analyses help contextualize findings. See also sampling bias, external validity, and weighting.