Logistic Regression!!

[deleted_user]

Hello Everyone,

I need to calculate ODDS RATIO for HIP t-score = -2.5 for four country regions and then plot it.

I created the model ex:-

ODS OUTPUT parameterestimates=fparameterestimates;
ODS OUTPUT Oddsratios=foddsratio;
proc logistic data= fracx descending ;
class cntry / param=ref ref=first;
MODEL censor = BMD cntry ;
run;
quit;
ODS OUTPUT CLOSE;

But how can i calculate OR for a paticular BMD t-score = -2.5 .? Pls suggest.

Thanks for all help.

Amit

[StatDave]

I'm not quite sure what you mean when you say you want an "OR for a paticular BMD t-score = -2.5". Does this mean you want an odds ratio comparing two countries when your BMD variable is set at 2.5? If so, then that is provided by default in the Odds Ratio Estimates table. Since your model contains only main effects, the odds ratio comparing a country to the reference country at BMD=2.5 remains the same at any other value of BMD.