1. You need to be careful here demeaning. When you set up a linear model, you have two parameters: the intercept and the slope. Without demeaning, the intercept is defined to be where the line intercepts the y axis at x=0 (in this case age=0, birth). When you demean age, you do the test at x=mean age; this will have a different intercept and so a different result.
2. It should not matter because you are looking at the BMI slope which will be independent of offset
On 6/17/2021 8:33 PM, WON JONG CHWA wrote:
External Email - Use Caution
Dear FreeSurfer experts,
I am currently running longitudinal analyses using the paired analysis tutorial for two time points. There are two tests I am trying to run:
- Calculating significant changes in volume between the two time
points, controlling for age and gender. 2. Correlating BMI with change in volume between the two time points, controlling for age and gender.
For (1), would it be necessary to demean age for a one-sample longitudinal analysis? And for (2), would I need to demean both BMI and age?
Thank you for your help.
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer