I have tested spatial auto-correlation in a generalised linear model using Moran's I test (DHARMa R package). I got results for observed, expected, sd and p-value and the result is there's no auto-correlation. How do one report the result of Moran's I in a scientific paper? Is the p-value itself enough or should I report any of the other results?


  • $\begingroup$ Presumably this refers to spatial autocorrelation of (some flavour of) residuals. I can't see that there is a protocol for this uniform across journals or papers. Depending on the project, full details on this might be considered anywhere between essential and not especially interesting or useful. Use the most similar published paper you can find as your guide and expect that reviewers will squawk if they don't like what you do or don't do. Personally I advise against just a P-value. Moran $I$ is complicated because -- despite a common myth -- it doesn't have simple bounds like $[-1, 1]$. $\endgroup$ – Nick Cox Apr 21 at 12:28

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.