0

how you doing?

I am doing regression analysis, and I have a question about selecting variables in multiple regression analysis.

enter image description here

There was multicollinearity in my model.

So, I used stepwise variable selection method.

I got a new model and here is the result.

enter image description here

I wanna use the variables which p-value are less than 0.5.

So, I eliminated "hfood" at first.

Because, among the vairables(hfood, beauty, furniture, fashion, tsucall)

hfood has the highest p-value.

I did one more analysis and I got the result.

And I did same process until the p-value of all variables is smaller than 0.5.

And finally, I got this model, and here is the result.

enter image description here

Was the precess I did right?

Is there more good process or something more I should do?

I am really appreciated with your help.

Have a nice day. Thank you~:)

mkt
  • 18,245
  • 11
  • 73
  • 172
Doyeong Park
  • 47
  • If you achieved your goal then the process worked for this data set. – James Phillips Mar 29 '18 at 13:48
  • When you perform stepwise exclusion of variables, the eventual $p$-values can not be interpreted without correction. See: https://stats.stackexchange.com/tags/stepwise-regression/info – Frans Rodenburg Aug 04 '18 at 03:01

0 Answers0