I have some rather exciting findings in my experiments. What's a good way
to describe an approach that uses every subject as their own baseline and
then applies ANOVA using the intrasubject change as the DV? I find I've
greatly amplified effect sizes and p levels in this way compared to the
traditional method of post-hoc testing of pooled data. I've found stuff
totally different from what is in the literature. So, what does all this
imply? That intersubject variability is too high for the pooled data
approach to be useful? Any other population parameters, such as deviations
from normality, that I should look at in presenting my findings? For
instance, there are large differences of skew between the distributions of
response times for the two groups (normal and schizophrenic) in the
control, baseline condition - an order of magnitude different. How
important is this?
I want to document and present these findings, and at this point I'm
wondering what to say, and what parameters to measure to document the case.
David B. Klein
M.Ps.
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|