help
help > RE: second-level analysis in Conn
Jun 16, 2017 01:06 AM | Alfonso Nieto-Castanon - Boston University
RE: second-level analysis in Conn
Dear Marta,
Yes, just as with multiple seeds/ROIs, in the case of multiple ICA components you can also simply select all of the ones you are interested in and leave the default 'eye(n)' contrast in order to perform an F-test across any of the selected ICA components. That said, you do not typically want to select too many components simultaneously (because that reduces very drastically the degrees of freedom in your analyses) so in ICA you typically first use the 'Summary display' to identify one or a few networks that you are interested in and then focus the rest of your analyses only on that or those few networks.
Hope this helps
Alfonso
Originally posted by marta555:
Yes, just as with multiple seeds/ROIs, in the case of multiple ICA components you can also simply select all of the ones you are interested in and leave the default 'eye(n)' contrast in order to perform an F-test across any of the selected ICA components. That said, you do not typically want to select too many components simultaneously (because that reduces very drastically the degrees of freedom in your analyses) so in ICA you typically first use the 'Summary display' to identify one or a few networks that you are interested in and then focus the rest of your analyses only on that or those few networks.
Hope this helps
Alfonso
Originally posted by marta555:
Dear Alfonso,
Does this method apply to ICA studies? (e.g. do I have to select all ICA I want to test simultaneously or one by one?)
Many thanks,
Marta
Does this method apply to ICA studies? (e.g. do I have to select all ICA I want to test simultaneously or one by one?)
Many thanks,
Marta