Bonferroni correction

Quick Reference

In statistics, a procedure whereby the significance level of a statistical test is adjusted in order to protect against Type I errors when multiple comparisons are being made. Thus if 10 t tests need to be performed on a set of scores, a researcher may use a Bonferroni-corrected significance level of .05/10 = .005 instead of the conventional .05. This procedure is extremely conservative inasmuch as it may lead to Type II errors, but it is safe, easy to understand, and can be used with a wide variety of tests. A Bonferroni-corrected t test is sometimes called simply a Bonferroni t test or a modified LSD test. A Bonferroni-corrected Mann-Whitney U test is sometimes called a Bonferroni Mann-Whitney U test, and so on. Compare Duncan's multiple range test, least-significant difference test, Newman-Keuls test, Scheffé test, Tukey-HSD test. [Named after the Italian mathematician Carlo Emilio Bonferroni (1892–1960) who, in 1935 and 1936, published inequalities on which the procedure is based]

Subjects: Psychology.

Reference entries