1
$\begingroup$

As title suggest: Is it possible to have confidence intervals crossing 0 but p = 0.042?

This happened to me in glm model with gamma family and link identity. The conf. intervals were generated by tidy function in R.

How should I interpret it? Or is my model wrong?


Actualization:

The intervals from tidy functions seems to be based on likelihood ratio. However when I compute the Wald's CI the the values now doesn't cross 0. So is tidy function using different aproache towards CI and p value?

$\endgroup$
1
  • $\begingroup$ Yes, this sounds like the libraries have different methods of calculating the critical value: Wald vs likelihood ratio. $\endgroup$ – Dave Mar 15 at 13:04

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.