I've read several pieces of documentation that say that the Geisser-Greenhouse correction for sphericity always increases the p value.
I've noticed in SPSS that if a test is not significant (before any correction), sometimes the Geisser-Greenhouse corrected value is actually less than the...