On Tue, 5 Dec 1995, Jo Robinson wrote:
>>>CUT>>>>>>
> That would work. THe problem is, my "study" is far too small (10 mice
> in all) to come up with significant data.
>>>>CUT>>>
I disagree. If the effect is strong it will show up. If the effect is
weak it will require thousands of mice to see it. Such a weak effect has
little practicial importance. So small sample sizes should be preferred
to pick up important effects. Ron Blue