Does not seem to be solid list in any of the tutorials or online help. Is the test still robust if inputs are not normally distributed?
You should read the original references at the end of this page rather than relying on a help file comment
How Hot Spot Analysis (Getis-Ord Gi*) works—Help | ArcGIS for Desktop
plus there are some good resources cited here
Spatial Statistics Resources | ArcGIS Blog
And to interpret statistical results for tests where the underlying assumption is complete spatial randomness, then inferences can be made from the discussion of z-scores and p0values
What is a z-score? What is a p-value?—Help | ArcGIS for Desktop
