Bob,

Thank you for your candor and alerting us to an error. But what exactly does this mean? Could you please give those of us who have been analyzing datasets prepared with 3dTshift in preprocessing some more specfics?

Namely, what was the precice effect on the 3d+time EPIRT dataset and what types of analyses will have been affected by this error?

For example, is this error so egregious that:

1) a portion of previous findings that presented data (having used the bugged 3dTshift in AFNI) may be suspect?

2) I had better re-process all my original data with the revised program code, re-do the regressions and group maps, and time series dumps?

Please advise.

Jim Bjork