[Eeglablist] ask again about bootstrap and group analysis. please help
wu xiang
rwfwu at ustc.edu
Tue Feb 21 23:50:20 PST 2006
Hi MICHAEL F NEELON, thanks for your help. But my problem is still not
be resolved.
In paper you send to me,Francesco Di Nocera et.al, 'Resampling approach
to statistical inference: Bootstrapping from event-related potentials
data', the authors done bootstrap for each subject (each subject had a P
value showed that if the result is reliable). They then reported how
many subjects had reliable results in paper. This is one alternative
method to show the bootstrap results.
However, as I mentioned previously, it seems that a overall bootstrap P
value across subjects can be get, as that in 'Perception's shadow:
long-distance synchronization of human brain activity',Rodriguez, E,
nature,397,430-433.(I still do not receive replies form the authors).
This is a more convenient way if many results need to be report (such as
many electrodes, or many electrodes pairs). Could someone further tell
me how get the bootstrap P value across subjects?
(btw: In methodological papers that introduce bootstrap method, there is
no problem of multiple subjects. While in experimental papers with
multiple subjects, most authors did not tell how get the overall
bootstrap P value across subjects in detail,although they showed it)
MICHAEL F NEELON wrote:
> Hi,
>
> Sorry noone's replied to your message. I'm attaching a couple of papers
> that address bootstrapping and ERPs, and the Di Nocera paper explicitly
> compares group statistics using bootstrapping with traditional ANOVA
> measures.
>
> Hope these help,
> Mike Neelon
> U. Wisconsin - Madison USA
>
More information about the eeglablist
mailing list