1

I have been using model averaging and model selection bases on AIC and BIC for a while. I have recently discover the stepwise regression technique and I found a lots of people critize this methods.

Why stepwise regression is to avoid ? Is it just only because you may not capture interaction between variable and therefore loose informations ?

Thank you

Nico Coallier
  • 301
  • 3
  • 11
  • 4
    There have been a number of Q&A already on this site: https://stats.stackexchange.com/questions/9171/aic-or-p-value-which-one-to-choose-for-model-selection?rq=1, https://stats.stackexchange.com/questions/20836/algorithms-for-automatic-model-selection?noredirect=1&lq=1, https://stats.stackexchange.com/questions/218208/what-are-the-advantages-of-stepwise-regression?rq=1 – mdewey Apr 27 '17 at 15:47

0 Answers0