Hey All great scientists and minds alike
I have a few questions regarding using "age" and other co variates on a second level analysis
as the age difference between the groups I am studying are significantly different...
1) when using age as a covariate why does the results differ so much when using A) 1 -1 0 vs B) 1 -1 0 0 in CONN?
...for A) the 0 involves the whole list of ages whereas B) the ages are seperated according to their groups
for example I tried both inputs and the functional connectivity results were inverted and totally different? What is happening here?
(please see attached)
2) Also what would happen if you use 0 0 1 -1 for example where by the ages are allocated as 1s and the groups I wish to look at are 0s... e.g. 0 0 = group A and B. 1 = Group A Ages. -1 = Group B ages. What would CONN explore and compute in this setting?
3) Also is it possible and does it make sense to regress more than one covariate that are independent of each other in CONN? for example subject age and subjects time since injury like this: 1 -1 0 0 (with 0s being age and time since injury respectively)
4) How would you control complicated scores such as motor or sensory scores when we only have values for one group (patients). Would you then have to either present a maximal score for healthy (e.g. 100) or would you put NaN for healthys instead? (although these scores are not missing, we would like to present the healthys as having a maximum score since they are healthy)
thanks so much for your time and attention, I know that this is a lot but I think it will help us all also!
Best wishes
Vincent
I have a few questions regarding using "age" and other co variates on a second level analysis
as the age difference between the groups I am studying are significantly different...
1) when using age as a covariate why does the results differ so much when using A) 1 -1 0 vs B) 1 -1 0 0 in CONN?
...for A) the 0 involves the whole list of ages whereas B) the ages are seperated according to their groups
for example I tried both inputs and the functional connectivity results were inverted and totally different? What is happening here?
(please see attached)
2) Also what would happen if you use 0 0 1 -1 for example where by the ages are allocated as 1s and the groups I wish to look at are 0s... e.g. 0 0 = group A and B. 1 = Group A Ages. -1 = Group B ages. What would CONN explore and compute in this setting?
3) Also is it possible and does it make sense to regress more than one covariate that are independent of each other in CONN? for example subject age and subjects time since injury like this: 1 -1 0 0 (with 0s being age and time since injury respectively)
4) How would you control complicated scores such as motor or sensory scores when we only have values for one group (patients). Would you then have to either present a maximal score for healthy (e.g. 100) or would you put NaN for healthys instead? (although these scores are not missing, we would like to present the healthys as having a maximum score since they are healthy)
thanks so much for your time and attention, I know that this is a lot but I think it will help us all also!
Best wishes
Vincent