unrepresentative

Unrepresentative is an adjective used to describe something that does not adequately reflect the characteristics of a broader group or population to which it is related. In research and data analysis, an unrepresentative sample or dataset fails to mirror the target population, which can lead to biased conclusions and misleading inferences.

In statistics, unrepresentativeness is often the result of sampling bias or nonresponse bias. Causes include convenience sampling, voluntary response, self-selection, undercoverage of segments of the population, or differential response rates. When a sample overrepresents or underrepresents certain subgroups, estimates such as means, proportions, or model parameters may be biased or not generalizable.

Examples include political polls that over-sample supporters of a particular party, customer surveys that favor urban shoppers over rural ones, or clinical studies that exclude certain demographic groups. Such unrepresentativeness can distort findings and affect decision making, policy, or resource allocation.

Mitigating unrepresentativeness involves designing studies to improve representativeness. Methods include random sampling, stratified sampling, quota sampling, and post-stratification weighting. Ensuring a representative sampling frame, maximizing response rates, and adjusting analyses to reflect population characteristics are common practices.

The term can also describe individuals or actions that do not typify a group, used to note that a particular case is not typical of broader patterns. Overall, representativeness is central to the credibility and applicability of empirical conclusions.