I'm pretty much reporting these results. I've bootstrapped my sample using lmersampler, which does not provide p-values for bootstrapped models. I relied on confidence interval boundaries to justify if the estimates were either significant (CI didn't contain zero) or not (it crossed zero). How would be the best way to report that?

So far, I've included a column called 'significance' and manually included sig or ns, based on the CI, but I'd like to know if there's a better approach to that.

  • $\begingroup$ Report for whom? $\endgroup$
    – Dave
    Commented Jan 27, 2023 at 14:41
  • $\begingroup$ to a journal submission, @Dave $\endgroup$ Commented Jan 27, 2023 at 15:20
  • $\begingroup$ What does the journal want? Many journals have guides about how to report statistics, either made explicit in the submission instructions or implicit in other publications in that journal. $\endgroup$
    – Dave
    Commented Jan 27, 2023 at 15:23


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.