help > Between Subjects analysis
Showing 1-3 of 3 posts
Feb 13, 2020 12:02 AM | nadiak
Between Subjects analysis
Hi, I am new to imaging and conn and just want to make sure i am
doing things right.
I have 30 subjects, each with just 1 session (15 seizure free, 15 not) and want to do a between subject analysis and compare rs connectivity between the two groups. I know in the set up there is an option to create two conditions for between subjects, and then in second level analysis i can specify contrasts.
My first question: Does it matter which order I input my data for when I later analyze them? Ie, based on how i specify i would need to input my non seizure free patients first, and then seizure free. And thus, the conditions would look something like this:
sz free: 00000011111
non sz free: 1111110000
I am unsure of how conn would otherwise be able to identify the different groups.
Second question: to perform the between subjects analysis would I select my contrasts as [1-1] to analyze differences between groups?
Hopefully that made some sense. Thanks so much!
I have 30 subjects, each with just 1 session (15 seizure free, 15 not) and want to do a between subject analysis and compare rs connectivity between the two groups. I know in the set up there is an option to create two conditions for between subjects, and then in second level analysis i can specify contrasts.
My first question: Does it matter which order I input my data for when I later analyze them? Ie, based on how i specify i would need to input my non seizure free patients first, and then seizure free. And thus, the conditions would look something like this:
sz free: 00000011111
non sz free: 1111110000
I am unsure of how conn would otherwise be able to identify the different groups.
Second question: to perform the between subjects analysis would I select my contrasts as [1-1] to analyze differences between groups?
Hopefully that made some sense. Thanks so much!
Feb 13, 2020 01:02 PM | Alfonso Nieto-Castanon - Boston University
RE: Between Subjects analysis
Hi Nadiak
Yes, everything looks fine except perhaps that you are mentioning "sz free" and "non sz free" as two 'conditions', when in CONN these should be defined as second-level covariates instead. In particular, in CONN, between-subjects factors (e.g. these two different groups of subjects in your study) are entered in the Setup.Covariates (2nd-level) tab, while within-subjects factors (e.g. if you had two scans per subject, one before and another one after an intervention) are entered in Setup.Condtitions
Hope this helps
Alfonso
Originally posted by nadiak:
Yes, everything looks fine except perhaps that you are mentioning "sz free" and "non sz free" as two 'conditions', when in CONN these should be defined as second-level covariates instead. In particular, in CONN, between-subjects factors (e.g. these two different groups of subjects in your study) are entered in the Setup.Covariates (2nd-level) tab, while within-subjects factors (e.g. if you had two scans per subject, one before and another one after an intervention) are entered in Setup.Condtitions
Hope this helps
Alfonso
Originally posted by nadiak:
Hi, I am new to imaging and conn and just want
to make sure i am doing things right.
I have 30 subjects, each with just 1 session (15 seizure free, 15 not) and want to do a between subject analysis and compare rs connectivity between the two groups. I know in the set up there is an option to create two conditions for between subjects, and then in second level analysis i can specify contrasts.
My first question: Does it matter which order I input my data for when I later analyze them? Ie, based on how i specify i would need to input my non seizure free patients first, and then seizure free. And thus, the conditions would look something like this:
sz free: 00000011111
non sz free: 1111110000
I am unsure of how conn would otherwise be able to identify the different groups.
Second question: to perform the between subjects analysis would I select my contrasts as [1-1] to analyze differences between groups?
Hopefully that made some sense. Thanks so much!
I have 30 subjects, each with just 1 session (15 seizure free, 15 not) and want to do a between subject analysis and compare rs connectivity between the two groups. I know in the set up there is an option to create two conditions for between subjects, and then in second level analysis i can specify contrasts.
My first question: Does it matter which order I input my data for when I later analyze them? Ie, based on how i specify i would need to input my non seizure free patients first, and then seizure free. And thus, the conditions would look something like this:
sz free: 00000011111
non sz free: 1111110000
I am unsure of how conn would otherwise be able to identify the different groups.
Second question: to perform the between subjects analysis would I select my contrasts as [1-1] to analyze differences between groups?
Hopefully that made some sense. Thanks so much!