Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!

Supported by

To log or not to log ORs

Very basic question, I am sorry! For a standard multivariable logistic regression, the outputs include "Estimate" and "OR". The Estimate is the natural log of the OR. The 95% CI surround the Estimate not the OR. So, do I report the "Estimate" and associated CIs, or report the OR and inverse log the provided 95% CI? I'm assuming the latter but getting in a bit of a pickle.

Comments

  • edited March 10

    yes the latter (to get the OR, you take the exponential of the estimate)

    You can take the exponential of the lower and upper limits of the CIs if they're based on estimates and they'll become odds ratios

  • Under the "Statistics" menu where you checked confidence intervals, you can also check "Odds ratio scale":


  • That is awesome and super helpful thank you so much!!

  • This sounds amazing as I found it very much useful and informative to be honest. Also, I have gone through this post which definitely helped me out a lot as a new member I am looking forward for more such discussion

  • This sounds amazing as I found it very much useful and informative to be honest. Also, I have gone through this post which definitely helped me out a lot as a new member I am looking forward for more such discussion

Sign In or Register to comment.