proc logistic - odds ratio for interaction terms

ChuksManuel — Mon, 05 Nov 2018 07:05:56 GMT

Hello guys, i am a little confused and a simple explanation will help here.

I ran these two codes. The second code is for the interaction model.

Variable names are "mi" and "sprob"

Proc logistic data = new descending; Model health = mi sprob; run;

Proc logistic data = new descending; Model health = mi sprob mi*sprob ; run;

I discovered that the Odds ratios was not displayed for 'mi' or 'sprob' in the interaction model when i ran it and i am wondering the reason why. The first code gave me an odds ratio but the second just didn't. Is there any theory for this? Either statistical or otherwise?

Thanks

Re: proc logistic - odds ratio for interaction terms

StatDave — Mon, 05 Nov 2018 13:52:08 GMT

See this note.

topic Re: proc logistic - odds ratio for interaction terms in Statistical Procedures

proc logistic - odds ratio for interaction terms

Re: proc logistic - odds ratio for interaction terms