help
help > RE: regression between connectivity and behavioral data
Jan 29, 2021 05:01 PM | Alfonso Nieto-Castanon - Boston University
RE: regression between connectivity and behavioral data
Dear Rui,
Re:(1) almost correct, but instead of [0 1 1 1] specify there, either:
a) the specific term that you want to analyze; e.g. [0 0 0 1] will test sex*age interactions
or b) several of those terms, if you want to run an F- test across all those terms; e.g. [0 1 0 0; 0 0 1 0;0 0 0 1] will test for the presence of either main sex effects (a [0 1 0 0] contrast), main age effects (a [0 0 1 0] contrast), or sex-by-age interactions (a [0 0 0 1] contrast)
Re:(2) yes, that is exactly correct
Best
Alfonso
Originally posted by Rui Li:
Re:(1) almost correct, but instead of [0 1 1 1] specify there, either:
a) the specific term that you want to analyze; e.g. [0 0 0 1] will test sex*age interactions
or b) several of those terms, if you want to run an F- test across all those terms; e.g. [0 1 0 0; 0 0 1 0;0 0 0 1] will test for the presence of either main sex effects (a [0 1 0 0] contrast), main age effects (a [0 0 1 0] contrast), or sex-by-age interactions (a [0 0 0 1] contrast)
Re:(2) yes, that is exactly correct
Best
Alfonso
Originally posted by Rui Li:
Dear Alfonso，
I performed a voxel-wise connectivity analysis with seed at PCC for a lifespan data, and now
1) I want to examine the sex, age, and the sex*age effects on the connectivity by multiple regressions. In the 'Group-analysis settings', I selected the 'All subjects', 'sex', 'age', and 'sex*age (interaction term)', with between-subjects contrast [0 1 1 1]; and selected the only one condition 'rest'. Am I right for the settings? and in the resulting files, is the 'spmT_0001' the effect of all the 3 terms on PCC connectivity, and if I want to examine respectively the statistical significance of sex, age, and sex*age using their respective regression coefficients, how to perform in CONN? I just only found beta_000x.nii files.
2)then I want to examine the correlation between connectivity and memory score with age as a covariate. In the 'Group-analysis settings', I selected the 'All subjects', 'memory', 'age', with between-subjects contrast [0 1 0], is the setting right?
Could you please correct me and help me with the settings? Thank you in advance, and thanks the great toolbox!
Best,
Rui
I performed a voxel-wise connectivity analysis with seed at PCC for a lifespan data, and now
1) I want to examine the sex, age, and the sex*age effects on the connectivity by multiple regressions. In the 'Group-analysis settings', I selected the 'All subjects', 'sex', 'age', and 'sex*age (interaction term)', with between-subjects contrast [0 1 1 1]; and selected the only one condition 'rest'. Am I right for the settings? and in the resulting files, is the 'spmT_0001' the effect of all the 3 terms on PCC connectivity, and if I want to examine respectively the statistical significance of sex, age, and sex*age using their respective regression coefficients, how to perform in CONN? I just only found beta_000x.nii files.
2)then I want to examine the correlation between connectivity and memory score with age as a covariate. In the 'Group-analysis settings', I selected the 'All subjects', 'memory', 'age', with between-subjects contrast [0 1 0], is the setting right?
Could you please correct me and help me with the settings? Thank you in advance, and thanks the great toolbox!
Best,
Rui
Threaded View
| Title | Author | Date |
|---|---|---|
| Rui Li | Jan 27, 2021 | |
| Alfonso Nieto-Castanon | Jan 29, 2021 | |