My question is concerned about two-way repteated anova on source data.
My experiment is a 2 factorial design: Factor A (3 levels), Factor B (2 levels). I want to see the main effects of the two factors. In the statis analysis, I set
effects = 'A+B' to test the main effect for both factor A and factor B. Then, we got
f_thresh = f_threshold_mway_rm(n_subjects, factor_levels, effects, pthresh) = [3.1995817058519864, 4.279344309144648], which is a list type, containing two elements. However,
threshold = float(threshold) in the function
spatio_temporal_cluster_test requires that the input cannot be a list (as the error attached).
If I set
effects = 'A' or
effects = 'B' or
effects = 'A:B' , the error goes away. So, testing the two factors one by one is an option.
I’d like to ask: statistically speaking, does it means the same thing to test effect A and B separately and to test them together (‘A+B’)?
Below are the statis scripts that define the statis parameters:
factor_levels = [3, 2] effects = 'A' # 'A','B','A+B':main effect of A, main effect of B, main effect of A and B # Tell the ANOVA not to compute p-values which we don't need for clustering return_pvals = False # a few more convenient bindings n_times = X.shape n_conditions = 6 def stat_fun(*args): # get f-values only. return f_mway_rm(np.swapaxes(args, 1, 0), factor_levels=factor_levels, effects=effects, return_pvals=return_pvals) pthresh = 0.05 f_thresh = f_threshold_mway_rm(24, factor_levels, effects, pthresh) print(f_thresh)