Bonferonni-Holm correction
You've seen that comparing many different datasets, even randomly generated ones, can result in "statistically significant relationships" that are anything but! One way around this is to apply a correction to the alpha of your confidence level. In this exercise you'll explore why you should apply this correction and how to do so.
The 1000 p-values you calculated in the previous exercise have been loaded for you in a NumPy array p_values
, as has the package NumPy as np
.
This exercise is part of the course
Foundations of Inference in Python
Hands-on interactive exercise
Have a go at this exercise by completing this sample code.
# Compute the Bonferonni-corrected alpha
bonf_alpha = ____
# Check how many p-values were significant at this level
____