Entering edit mode

10 months ago

mel22
▴
100

Hello everybody, I am trying to construct a regression model with the stepwise procedure. I wonder given the relatively high number of variables (about 50 clinical variables) I am testing if it would be better to adjust for multiple testing (Bonferroni for example ?) Do you have any idea, recommendation of paper or any other reference ?

Thank you !

Please check the discussion in the link below which is very close to your design model.

https://stats.stackexchange.com/questions/468620/correction-for-multiple-testing-in-multiple-regression-analysis

Many thanks