Test noise ceiling against chance
HeikoSchuett opened this issue · comments
When discussing the tests in our inference paper, we noticed that it might be informative to test whether there is any reliable information in the measured RDMs. Given that we have variance estimates for the noise ceiling from bootstrapping, this should be easy to implement.
Describe the solution you'd like
Implement the same types of tests we use to test whether the models are significantly above chance for the noise ceiling.
Describe alternatives you've considered
There are other tests for the presence of information, but here it seems most sensible to just do the simple consistent test for the completely flexible model.