This package contains a set of functions for inferiential statistics
using resampling methods. Data are organized into arrays so multiple tests can be run at once (up to one million test per second). The package supports bootstrap, permutation and parametric statistics using paired or unpaired t-test, or repeated measure ANOVA (one way and two ways).
Arnaud Delorme (2019). Resampling statistical toolkit (https://www.mathworks.com/matlabcentral/fileexchange/27960-resampling-statistical-toolkit), MATLAB Central File Exchange. Retrieved .
I am having some trouble believing some of the results. When computing a repeated-measures ANOVA (1X3) for example, I obtain very small F values <2, but very significant p-values when using the perm and bootstrap modes. How can this be? Is it really the case that I can have a teeny F and still beat the surrogate distribution?
Agree with Jens, the first example does not work unless you change line 150.
I think statcond.m line 150 should assign g.mode='param'