Robert Oostenveld r.oostenveld at FCDONDERS.RU.NL
Thu Mar 22 09:49:57 CET 2007

On 22 Mar 2007, at 6:09, Harald Stoegbauer wrote:
> A short additional question, it is also possible to make a 'normal'
> ttest on a particular time window of an ordinary VEP-ERP ?

TIMELOCKSTATISTICS (idem for freq/source) allows you to do
conventional tests as well.

Instead of
  cfg.method    = 'montecarlo'       % since the p-value is a monte-
carlo estimate
  cfg.correctm  = 'cluster'
you would specify
  cfg.method    = 'analytic'         % since it is an analytic p-
value that will be computed
  cfg.correctm  = 'no'               % or bonferoni, fdr, holms

The cfg.statistic is the same, just as the cfg.design matrix. The
choises that you have for multiple-comparison correction are
different. See the help of STATISTICS_ANALYTIC for more details.


More information about the fieldtrip mailing list