From the output of a logistic regression in JMP, I read about two binary variables:
Var1 estimate -0.1007384 Var2 estimate 0.21528927
Odds ratio for Var1 lev1/lev2 1.2232078 reciprocal 0.8175225 Odds ratio for Var2 lev1/lev2 0.6501329 reciprocal 1.5381471
Now I obtain
exp(2*0.1007384), and similarly for the other odds ratio.
So, my question is: why I have to multiply by two? Is the relation between a coefficient c and its odds ratio r equal to r=exp(c)? Isn't it?