sdm-help-list > Covariables
Showing 1-5 of 5 posts
Nov 15, 2016 09:11 PM | Azalea Reyes Aguilar - UNAM
Covariables
Hello Joaquim,
We have some questions with our meta-analysis.
We did a comparison between two groups with one and two covariables.
Firstly, for the analysis with one covariable we got three outputs: 1) 01m00, 2) 10m00, and 3) 10m01. So, we want to know what means each one of them.
And secondly, for comparison between two groups with two covariables we got four outputs: 1) 001m000, 2) 010m000, 3) 100m000, and 4) 100m010. it's the same question, i.e. what means each one of them.
Thanks!
We have some questions with our meta-analysis.
We did a comparison between two groups with one and two covariables.
Firstly, for the analysis with one covariable we got three outputs: 1) 01m00, 2) 10m00, and 3) 10m01. So, we want to know what means each one of them.
And secondly, for comparison between two groups with two covariables we got four outputs: 1) 001m000, 2) 010m000, 3) 100m000, and 4) 100m010. it's the same question, i.e. what means each one of them.
Thanks!
Dec 22, 2016 10:12 AM | Joaquim Radua
RE: Covariables
Dear Azalea,
sorry for the confusing names of the outputs.
I understand that the first regressor is the binary variable indicating the group, and the second and third regressors are the covariables. Then the outputs mean:
10m00 and 100m000: difference between groups
01m00 and 010m000: effect of the (first) covariable
001m000: effect of the second covariable
The other outputs are specific contrasts which may be of no interest here.
Hope this helps,
Joaquim
sorry for the confusing names of the outputs.
I understand that the first regressor is the binary variable indicating the group, and the second and third regressors are the covariables. Then the outputs mean:
10m00 and 100m000: difference between groups
01m00 and 010m000: effect of the (first) covariable
001m000: effect of the second covariable
The other outputs are specific contrasts which may be of no interest here.
Hope this helps,
Joaquim
Jul 25, 2017 02:07 PM | Gongjun Ji
RE: Covariables
Dear Joaquim
"01m00 and 010m000: effect of the (first) covariable"
*01m00*.nii represent "the effect of the first covariable" rather than "between-group differece after regress out the effect of first covariate"?
If so, how to setting the software to regress out the effect of covariates (such as the "age" when comparing patients and controls)?
Thanks a lot for your time!
JI
"01m00 and 010m000: effect of the (first) covariable"
*01m00*.nii represent "the effect of the first covariable" rather than "between-group differece after regress out the effect of first covariate"?
If so, how to setting the software to regress out the effect of covariates (such as the "age" when comparing patients and controls)?
Thanks a lot for your time!
JI
Sep 4, 2017 01:09 PM | Joaquim Radua
RE: Covariables
Dear Gongjun Ji,
I am sorry if this is confusing. I think that maybe it's easier if you send me an email with what exactly you want to do and I will try to tell you how.
Best,
Joaquim
I am sorry if this is confusing. I think that maybe it's easier if you send me an email with what exactly you want to do and I will try to tell you how.
Best,
Joaquim