I have the following question tonight. I did a random block factorial
experiment with repeated measures and let the computer collect all the
response data and calculate the ANOVAs. Now the thesis committee is denying
that my residual degrees of freedom could be in the hundreds, which is what
the software says. All sums of squares and degrees of freedom add up
properly in the printout. My understanding of the model (though I let the
computer do the work) is that the term 'n' can quite reasonably be expected
to be in the hundreds even though only 10 people participated in the
experiment, because the repeated measures count as extra "subjects" in the
model testing for the treatments and their interactions. I'd like to do part
of the data over on a hand calculator just to show them that this is
reasonable. Am I correct?
TIA,
David Klein
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|