SPM users,
Although the issue of orthogonalization of conditions (particularly in
conjunction analyses) has been address many time in the list, including
questions by me, I continue to have trouble understanding this issue.
I study sensory processing that is based on animal and human
electrophysiology studies. My experiments typically have 3 conditions as
well as a rest period. Unfortunately, the stimulus conditions share the
same rest period. I believe that much of my results are due to exogenous
attention switching and require that the stimuli be presented in
pseudorandom order. If I were to design an experiment with 3 different
trials, one for each condition, I do not expect to observe the results
that I am interested in. So, is there any way to design a study with
multiple conditions that share the same baseline such that the columns
in the design matrix are orthogonal? I continue to have analysis
problems due to the non-orthogonal nature of my contrasts and am in the
process of starting an event-related study. I want to ensure that I
collect the data in a form that that will allow for comparisons between
conditions in a second level analysis or using conjunctions.
Any help will be appreciated.
Paul Laurienti
|