Hi, I am using Bayesian estimations to run logit regression models in Stata with odds ratios. However, i seem to have run into some problems with Stata's command for measuring intervals: bayestest interval.
For one categorical variable called edu_years, the odds ratio is above 1. However, Stata does not report the interval test. Regardless of whether I change the upper and lower bounds, add other variables to the logit regression model, or whether I change the burn in, thinning or MCMC size, Stata always report the interval tests as 0 or 1.
These are some of the outputs for the interval test I did after the logit model.
Would anybody have any advice on what I could be doing wrong? Thanks!
These are the interval tests:
Even when I change to using 1 for the upper and lower bounds it doesn't seem to solve the problem.
For one categorical variable called edu_years, the odds ratio is above 1. However, Stata does not report the interval test. Regardless of whether I change the upper and lower bounds, add other variables to the logit regression model, or whether I change the burn in, thinning or MCMC size, Stata always report the interval tests as 0 or 1.
These are some of the outputs for the interval test I did after the logit model.
Would anybody have any advice on what I could be doing wrong? Thanks!
These are the interval tests:
Even when I change to using 1 for the upper and lower bounds it doesn't seem to solve the problem.
Attached Files
