Dear experts,
Hi, I’m trying to analyze my date based on the factorial design using TDT.
My experimental design includes 2 X 2 factors, such as A1B1, A1B2, A2B1, and A2B2.
My main interest is the effect of A1B1 condition.
Thus, is it correct to use the interaction contrast ((A1B1 - A1B2) - (A2B1 - A2B2))
and describe as follows,
cfg = decoding_describe_data
(cfg,{’A1B1’,’A1B2’, ‘A2B1’, ‘A2B2’},[1 -1 -1 1], regressor_names, beta_loc);
For univariate analyses, I think such value setting is common.
Is it also appropriate for MVPA?
If it’s not, are there any other methods to clarify an interaction contrast?
Thank you for your help.
Rue