Calculating odds ratios for interaction variables that are binary in logistic regression

[Aleksandar]

Hi everyone,

I have run the logistic regression for a binary dependent variable and a set of independent variables (numeric variables). However, I could not obtain the odds ratios for the variables used in the interaction term and the interaction term odds ratio. These variables are 

I have run the following code:

proc logistic data=First.Model descending;
model TTNIRS = age Gender D_Year_1 D_Years_2 D_Years_5 D_Year_10 Ascending_group High_amount Regular_saver
D_Low_litearcy D_High_litearcy D_Highest_litearcy Ascending_group*High_amount / ctable pprob=0.5;
output out=lpred predicted=plogit;
run;

However, the output did not provide the odds ratios for the Ascending_group, High_amount, and Ascending_group*High_amount.

I got the following results for odds ratios:

Odds Ratio EstimatesEffect Point Estimate 95% WaldConfidence LimitsageGenderD_Year_1D_Years_2D_Years_5D_Year_10Regular_saverD_Low_litearcyD_High_litearcyD_Highest_litearcy

0.997	0.988	1.007
0.847	0.682	1.051
0.868	0.634	1.189
0.502	0.363	0.692
0.435	0.314	0.602
0.364	0.261	0.507
1.680	1.296	2.178
0.847	0.480	1.495
0.717	0.420	1.224
0.927	0.555	1.549

How to obtain odds ratios for Ascending_group, High_amount, and Ascending_group*High_amount? 

Thank you in advance.

Aleksandar

[ballardw]

What did your log say after running the procedure? You might consider posting the code with all messages from the log.

Without data there really isn't any way to tell what you need if the issue is data related.

[StatDave]

See this note.