On 10 July 2011 15:07, Alana James <[log in to unmask]> wrote:
> Hi Laura,
>
> something that occurred to me was that you could potentially reduce the 70
> food items into food groups e.g. fruits, vegetables etc. You could then
> create a score for each participant for each food group, and perform the
> appropriate ANOVA to see if participants' judgements significantly
> differed across food groups.
>
> Of course, whether this will give you useful information will depend on
> your research question(s).
>
That's a good idea. If you also calculated Cronbach's alpha (and if
it was OK) that would give you more evidence that this was a sensible
thing to do.
Post hoc tests in repeated measures anova are also hard, but at least
your number of variables will be easier to handle.
Jeremy
|