Sample bias

from Wikipedia, the free encyclopedia

A sampling bias ( English selection bias , selection effect , or sampling bias ) is a statistical bias in the selection of sampling units.

One way to deal with sample bias is to use Heckman's correction .

Sample distortions occur, for example, due to the non-random difference in the willingness of people to participate in empirical social research (see distortion of silence , self-selection ). Another example is survivorship bias , which occurs when only the successful are considered in a success / failure study.

See also

literature

  • James Heckman (1979): Sample selection bias as a specification error. Econometrica, 1979, pp. 153-161.

Applications

  • Geddes, Barbara: How the Cases You Choose Affect the Answers You Get: Selection Bias in Comparative Politics, in: Political Analysis 2 (1990), 1, pp. 131-150.