I know that a useful method is to see the AIC or nested F-test but... is it correct to select variables with the stepwise methodology, selecting the variables with the lowest p-value from the t-test and discarding the variable with the highest p-value? thanks for any help
Asked
Active
Viewed 15 times
1
-
This is a FAQ: see https://stats.stackexchange.com/search?q=stepwise+regression. Be aware, too, that your characterization of "stepwise methodology" differs from the standard implementations, which take some steps to do better than the procedure you describe. Your methodology reads simply like "retain the most significant variables." There doesn't seem to be any iteration. – whuber Mar 03 '22 at 15:56