>As to the anxiety about overinterpretation of unthresholded maps, I'm
>not sure there is much else to say. I refer to you Tom Nichols email
>in this thread, or to almost any behavioural paper, where
>not-signficant effects are routinely and appropriately reported.
But in behavioral results it is typical to report the mean and std.....not
just a p-value. So equivalence would require two sets of images, right? and
if you want the p-value, you then need three....seems a bit much to ask in
published articles (although i'm all for requiring these things to be
stored online).
vince
|