Re: Within-subjects proportional data

Janet wrote:
> [...]
> I assume that if the omnibus F test is valid, then post hoc or planned
> comparisons for specific contrasts can also be performed with ipsative
> data?

Yes. In fact, if the Bonferroni problem is handled properly then the
contrasts do not require the overall F to be done, let alone be
significant.

.

Relevant Pages