help
help > RE: second-level analysis in Conn
Aug 10, 2015 06:08 PM | Alfonso Nieto-Castanon - Boston University
RE: second-level analysis in Conn
Hi Diana,
The 'All effects' contrast is an F-test using an 'eye(N)' contrast (where N is the number of effects selected in your 'subject effects' list). This looks at the significance of the entire second-level model (which typically includes main connectivity effects as well as any covariates entered) and it can be understood as an OR conjunction of each individual subject-effect covariate entered into your model. The particular interpretation depends on the specific subject-effects selected, for example if you are selecting 'Group1', 'Group2', and 'Group3', then the 'all effects' contrast will be [1 0 0; 0 1 0; 0 0 1] which tests an OR conjunction of the individual [1 0 0], [0 1 0], and [0 0 1] contrasts, or, in other words, those areas where the connectivity with your seed regions is different from zero within any of your three subject groups.
Hope this helps
Alfonso
Originally posted by Diana Parvinchi:
The 'All effects' contrast is an F-test using an 'eye(N)' contrast (where N is the number of effects selected in your 'subject effects' list). This looks at the significance of the entire second-level model (which typically includes main connectivity effects as well as any covariates entered) and it can be understood as an OR conjunction of each individual subject-effect covariate entered into your model. The particular interpretation depends on the specific subject-effects selected, for example if you are selecting 'Group1', 'Group2', and 'Group3', then the 'all effects' contrast will be [1 0 0; 0 1 0; 0 0 1] which tests an OR conjunction of the individual [1 0 0], [0 1 0], and [0 0 1] contrasts, or, in other words, those areas where the connectivity with your seed regions is different from zero within any of your three subject groups.
Hope this helps
Alfonso
Originally posted by Diana Parvinchi:
> (Hi Alfonso,
Thank you very much for your reply. That was very helpful. We have another questions for you. When we select �any effect� from the options provided under the �Between-subjects contrast� we find more significant effects than when �main effect� or simple contrasts are selected. What type of analysis does it ( �any effect�) run? Is it an f-test? If so, why does it produce more effects?
> )
Thank you very much for your reply. That was very helpful. We have another questions for you. When we select �any effect� from the options provided under the �Between-subjects contrast� we find more significant effects than when �main effect� or simple contrasts are selected. What type of analysis does it ( �any effect�) run? Is it an f-test? If so, why does it produce more effects?
> )