Hi,

Please I have a small query regarding calculation of adjusting odds ratio for stunting. I am working on publication of paper with DHS data and I am confused regarding unadjusted and adjusted odds ratio. If all the confounding variables are in the model, then as per my understanding, this is adjusted odds ratio. If the model just contains one independent variable, then it is unadjusted.

Please help me with this.

Thank you

Sajama]]>

You are correct. The unadjusted odds ratio comes from a logit regression with only one independent or predictor variable. If control variables are included then the odds ratio for the variable of interest is described as an "adjusted" odds ratio. Of course, the adjusted odds ratio will be different for different choices of control variable(s).]]>