[pymvpa] Suspicious results

[Francisco Pereira]

On Mon, Feb 28, 2011 at 11:58 AM, Nynke van der Laan
<nynkevanderlaan at gmail.com> wrote:

> To mention: all data stems from 1 scanning session (approximately 15
> minutes). Each chunk is one trial (approx 16 sec in total). Trials
> were separated by a (random duration) inter trial interval.

Varying between how much and how much?

> Because of
> the inter trial interval I would not expect that there is
> contamination between chunks/trials, would you?

Nobody expects the Spanish contam... sorry, I couldn't help myself :)

Now seriously, this was just a very simple sanity check. I've been
surprised by how much examples from trials that should have been far
enough apart could be correlated (thanks to Michael Hanke for helping
me realize this), and since then I've been more careful. See
randomizing the class labels as a very simple, cheap sanity check. If
that is not the problem, then you have an intriguing situation that is
bound to attract curiosity and effort!

cheers,
Francisco