Hi
It will summarise the different temporal responses in the same way as in tensor ICA, only difference being that this is just a single post-hoc summary (i.e. it is not part of the estimation itself as in TICA but just a plotting thing at the end). For effects where there is no temporal consistency these can be validly ignored, if you do happen to have some consistent temporal responses, however, this gives you the info close to what you;d be getting from TICA.
hth
Christian
On 24 May 2012, at 08:44, Gabor Perlaki wrote:
> Sorry for reposting my question, but I'm really interested in the following issue:
>
> Why melodic draw a boxplot on the report.html page of a multi-session temporal concatenation ica? It is a simple 2D ica, and not a tensor ica, so where the subjects/session modes comes from?
>
> Thanks,
> Gabor
|