Statistical demonstrations

Sometimes it's easier to demonstrate something than to explain it. Other times people don't believe us when we tell them something patently and verifiably true. For both these reasons, we've put together the following demonstrations of simple statistical principles that are relevant in the context of modeling your fMRI data.

• Mean centering demo — what happens you you mean center your covariates (or not), with or without an intercept
• Paired t-test demo — how do you code a paired t-test in a GLM
• Orthogonalization demo — what's the effect of orthogonalizing partially correlated regressors
• Linear contrast demo — how can you test for a linear effect