Rik,
In ANOVAs and SPM (revised ed., July 12, 2005), a process is detailed
whereby the "contrasts for each experimental effect can be obtained by
the
Kronecker products of Cm's and Dm's for each factor..."
What's the reasoning underlying this construction? Is it merely one
systematic way of constructing the main and interaction effects? Or is
it intended to additionally preserve covariance structure in the error
terms as a precursor to estimating nonsphericity?
Best,
Stephen J. Fromm, PhD
Contractor, NIMH/MAP
(301) 451-9265
|