ANOVA should be a subclass of statistical hypothesis test to be consistent with STATO

[cmungall]

OBI:

note that ANOVA and 'statistical hypothesis test' are siblings

STATO:

STATO injects its own intermediate "between group comparison statistical test" between two OBI classes. I think injection is problematic, but that is for STATO and OBI to work out.

However, STATO also injects an is-a between two OBI classes, ANOVA and 'statistical hypothesis test' - I am not sure why that was injected into STATO rather than requested directly of OBI

[bpeters42]

Good catch! We should change the placement of ANOVA to be a statistical test in OBI. I did a quick review of other data transformation processes in OBI, and there are no obvious other examples of this happening.