Multiple comparisons

The performance of multiple analyses on the same data. Multiple statistical comparisons increase the probability of making a type I error, i.e. attributing a difference to an intervention when chance is the more likely explanation.

Page last edited: 01 March 2007