Log in

View Full Version : What is the probability of type I error with multiple comparisons


oncotype
May 8, 2010, 01:34 AM
I would appreciate help with the following:
There are 512 contiguous points of comparison.
I perform a t-test for each point.
The probability of getting an erroneous result for each comparison is 0.05.
What is the probability of getting an erroneous result in 5 contiguous points (at any 5 points alonf these 512 points).
Thanks.

Parlaydog
Jun 24, 2010, 07:49 PM
0,05 ^ 5