GraphVar_2.01 GLM panel

Submitted by Yuki on

Hello!

I read GraphVar_SPSS_GLM_validation and GraphVar_Manual but I don,t know GLM panel.

Could you explain me the difference between them(between covariate,nuisance covariate,within covariate)?

All the best,

Yuki

 

Johann Kruschwitz

Thu, 06/21/2018 - 13:23

Hi,

please have a look at 1.NEWs_GraphVar_GLM_Turbo.pdf
Here the fields are explained.
Best,
Johann

 

Yuki

Sat, 06/23/2018 - 02:15

In reply to by Johann Kruschwitz

Hi,Johann

 

Thank you for your quick reply.

I read 1.NEWs_GraphVar_GLM_Turbo.pdf but I dont understand nuisance covariates.

 

I apologize for my lack of ability,

But please let me ask you a few question.

 

This may be a very question but I want to know How does covariates(i.e age, fantasy_score, sex, etc) affect Calculate Variables(i.e dependent variable).

Also, how is it different from the analysis in which nuisance covariates box is emptied and putting variables to be estimated in between covariates?

 

I'm not so good at English.

Sorry if it's wrong.

 

Best,Yuki

Johann Kruschwitz

Thu, 07/05/2018 - 12:35

In reply to by Yuki

Hi Yuki,
nuisance covariates in the GLM model are regressed out from the dependent variables prior to all analyses (the  dependent variable represents the output or outcome whose variation is being studied; e.g., a specific graph measure, or connectivity between regions). This is similar as removing the global signal or CSF signal in rs-fMRI analyses... here these signals are removed from the actual signal (i.e., their inflcuence is estimated in a regression analysis) and the residuals (i.e., the part of the signal that cannot be explained by the global signal or CSF) is used for further analysis. Subsequently, your actual analysis of interest is performed with the residuals. For example, an analysis with between covariates (continuous) would parallel a multiple linear regression, in which the unique contribution of entered variables [e.g. age, education, fantasy score] to the dependent variable is determined and associated with an effect (.e., a beta value).
I hope this helps in understanding the difference between nuisance variabe and between covariates in the GLM.
BEst,

Johann

Forums