help
help > RE: flexible ANOVA spreading interaction
Sep 28, 2016 02:09 AM | Alfonso Nieto-Castanon - Boston University
RE: flexible ANOVA spreading interaction
Dear Joyce,
Not explicitly. The simplest way to do that in CONN would be to perform a standard ANOVA interaction test (which will look at any differences with treatment which are different in patients vs. controls), and then perform post-hoc analyses (or simply evaluate the individual pre/post patients/controls effects for the resulting pathways) in order to filter out and/or interpret the found effects.
Hope this helps
Alfonso
Originally posted by Xi Zhu:
Not explicitly. The simplest way to do that in CONN would be to perform a standard ANOVA interaction test (which will look at any differences with treatment which are different in patients vs. controls), and then perform post-hoc analyses (or simply evaluate the individual pre/post patients/controls effects for the resulting pathways) in order to filter out and/or interpret the found effects.
Hope this helps
Alfonso
Originally posted by Xi Zhu:
Dear All:
In our study, we'd like to test the treatment effects in patient group, as compared with control group. Our hypothesis is the treatment will normalize Amygdala-prefrontal connectivity (more similar to control). Thus we have two groups (Patient and control) and two conditions (pre-treatment, post-treatment), and we are particularly interested in "spreading" (ordinal) interactions (i.e. pathways that patients improved over treatment, but controls remain the same), as opposed to "cross-over" (disordinal) interaction (for example, pathways that patients improved over treatment, but controls decreased over treatment). I wonder if CONN is able to do this type of flexible ANOVA test? and How? Many thanks
Joyce
In our study, we'd like to test the treatment effects in patient group, as compared with control group. Our hypothesis is the treatment will normalize Amygdala-prefrontal connectivity (more similar to control). Thus we have two groups (Patient and control) and two conditions (pre-treatment, post-treatment), and we are particularly interested in "spreading" (ordinal) interactions (i.e. pathways that patients improved over treatment, but controls remain the same), as opposed to "cross-over" (disordinal) interaction (for example, pathways that patients improved over treatment, but controls decreased over treatment). I wonder if CONN is able to do this type of flexible ANOVA test? and How? Many thanks
Joyce
Threaded View
| Title | Author | Date |
|---|---|---|
| juicefoods | Sep 26, 2016 | |
| Alfonso Nieto-Castanon | Sep 28, 2016 | |
| juicefoods | Sep 28, 2016 | |