Skip to main content
The Library

MASH Guide

Bonferroni adjustment (manual)

Why do you need Bonferroni adjustment? 

When you perform multiple tests (pairwise comparisons) on the same data set, your chances of obtaining a false positive result increase. The Bonferroni adjustment is used to minimise this problem, by changing the significance threshold, alpha.

For some tests, SPSS has built-in menus to run pairwise comparisons. The p values generated through this procedure have already been Bonferroni-adjusted. However, when running multiple tests manually, you have to Bonferroni adjust the significance level yourself. One way to do this is to divide the .05 alpha threshold by the number of tests you are intending to run. Report the Bonferroni-adjusted level of alpha, as well as the calculation it comes from.

The example below is with three Pearson correlation tests.

examples of three Pearson correlation tests