referencepopulaatioon

A reference population is a clearly defined group of individuals used as a baseline for comparison in research and analysis. It provides a standard against which measurements, predictions, and outcomes can be evaluated, and it may also supply normative values for specific variables.

Key characteristics include explicit inclusion criteria, geographic and temporal relevance, and sufficient sample size. The reference population should resemble the target population in important aspects such as age distribution, sex, and ethnicity, or be stratified into subgroups that match study categories. High data quality and documentation are essential.

Applications span several fields. In biomedicine, reference populations establish normal ranges for laboratory tests, growth charts, and diagnostic thresholds. In epidemiology, they enable age- and sex-standardized rates and comparisons across populations. In genetics, reference panels of sequenced individuals support genotype imputation and haplotype inference.

Methods involve careful selection and pooling of data, weighting, and, when appropriate, pooling multiple sources. Standard statistical techniques used with reference populations include direct standardization, stratification by relevant variables, and the calculation of reference intervals or z-scores to enable comparison.

Limitations include potential lack of representativeness for groups not included, changes in population characteristics over time, and biases from nonresponse or selection. When possible, researchers maintain multiple reference populations or update them periodically to reflect current demographics.

Common examples are World Health Organization or national growth references, and genetic reference panels used for imputation. The term reference population can also be seen as a baseline population in modeling.