[FieldTrip] ancova for sources analysis

Eelke Spaak eelke.spaak at donders.ru.nl
Wed Jul 31 09:58:39 CEST 2013

Dear Gabriel,

To do a full ANCOVA, you would have to write your own 'statfun' to use
in the statistics routines. This is not super-hard to do, but is also
far from trivial.

However, there might be a much easier solution. There is a function
called ft_regressconfound which linearly regresses out a 'confound' or
covariate from raw, freq, or source data. Would it be an option to use
this function first, and then do ft_sourcestatistics on the cleaned
data? Note that ft_regressconfound only supports one value per trial,
per confound; so the covariates cannot be fully time-resolved.


On 31 July 2013 09:05, Gabriel Gonzalez Escamilla <ggonesc at upo.es> wrote:
> Dear fieldtrip experts,
> I'm wondering if it is possible to perform an ancova or to remove the effect
> of a covariate in a source analysis?
> Many thanks in advanced,
> Gabriel
> _______________________________________________
> fieldtrip mailing list
> fieldtrip at donders.ru.nl
> http://mailman.science.ru.nl/mailman/listinfo/fieldtrip

More information about the fieldtrip mailing list