st: Rescaling odds ratios

[BARTHRILEY@comcast.net]

Dear Statalist

 

I am performing a logistic regression using xtlogit. Some of my predictors are measured in days of treatment. Is there a way to recalculate the OR for a different unit of days, i.e., 30 days of treatment? Or would I have to transform the variable (e.g., dividing the per day variable by 30)?

 

Thanks

 

Barth