partial coefficients

AFNI Message Board
1

Dear Colleagues,

Hi. I hope you are all coping with everything well.

I have a question about this text from the 3dDeconvolve help file: “It is VERY important to realize that statistics (F, t, R^2) computed in 3dDeconvolve are MARGINAL (or partial) statistics. For example, the t-statistic for a single beta coefficient measures the significance of that beta value against the regression model where ONLY that one column of the matrix is removed” (https://afni.nimh.nih.gov/pub/dist/doc/program_help/3dDeconvolve.html).

Got it. But what about the beta coefficients themselves? Are they also partial? I assume so because the t-statistics are testing their significance, but I thought I would check.

Sincerely,

Dante

2

Hi Dante,

Everything in the time series regression model is partial because they all share the same plate of food (i.e., data). In other words, it is not like one regressor walks into the kitchen first and takes as much as it wants; rather, they all split everything at the same time in an impartial fashion even though each one gets a partial proportion.

1 Like
3

Mmm, data.