Sorry. I am not converting odds ratios to percentages I was talking about the percantage of the variation among the women that is attributable to smoking. The odds ratio is (p1/(1-p1))/(p2/(1-p2)). I guess you are interested in p1/(p1+p2). First remember that you are computing estimates of the parameters and not the parameters themselve. An odds ratio of 1 means p1=p2 or p1/(p1+p2)=1/2. An odds ratio of 1.8 means p1/(1-p1)= 1.8 p2/(1-p2) or p1(1-p2)=1.8 p2(1-p1) or p1-p1 p2 =1.8 p2-1.8 p1 p2. This implies p1=p2-.8 p1 p2 or p1= p2(1-.8 p1) or p1/(p1+p2)= p2(1-.8 p1)/[p2(2-.8 p1)]= (1-.8 p1)/(2-.8 p1)] This depends on p1. If p1=0 this is 1/2 and if p1=1 it is .2/1.2=1/6. for any p1>0 it is less than 1/2.
Michael R. Chernick
- 43.2k
- 28
- 85
- 159