Re: [Freesurfer] Linear Mixed Effects model

[Diers, Kersten /DZNE]

Hi Kasper,

your design matrix X should have as many rows as there are measurements. Say if 
you have m subjects with n measurements each, it will be m*n rows. If different 
subjects have different numbers of measurements, that's also OK; in such cases 
the number of rows per subject in X will also differ.

Then, X should contain the following columns:

1. intercept 
this column should be all ones

2. time from baseline
if you have two time-points for each subject, this column should look as 
follows:
0 follow-up-time_subj1 0 follow-up-time_subj2 0 follow-up-time_subj3 etc.
here, follow-up-time is the difference in time between time-point 1 and 2, i.e. 
(time-point 2 minus time-point 1).

3. group indicator
0 for controls, 1 for patients

4. time x group interaction
multiply the values of columns 2 and 3

Further, it is important that the rows of X are ordered according to time for 
each subject. In your case with two measurements for each subject (if I 
understood correctly), this means that for the rows of X, time should be the 
faster changing index, and subject the slower changing index. This results in a 
sequence like subj1_time1 subj1_time2 subj2_time1 subj2_time2 subj3_time1 etc.
Needless to say, this must be the same in the ordering of your data variable Y.

Hope this helps,

Kersten



Am 08.08.2017 um 14:52 schrieb Kasper Jessen :

> Dear FreeSurfer,
> 
> I have a longitudinal study (patients and controls) with two time points 
> (i.e., baseline and six-weeks follow-up). I would like to use the LME model 
> as it can handle unequal timing and different number of time points across 
> subjects due to missing data. 
> 
> I would like to test if there is an effect of (a) time (b) group and (c) 
> group x time interaction.
> 
> However, i am trouble with my design matrix X:
> 
> (1) How would i write X? 
> 
> Thanks in advance
> Best regards
> Kasper
> ___
> Freesurfer mailing list
> Freesurfer@nmr.mgh.harvard.edu
> https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
> 
> 
> The information in this e-mail is intended only for the person to whom it is
> addressed. If you believe this e-mail was sent to you in error and the e-mail
> contains patient information, please contact the Partners Compliance HelpLine 
> at
> http://www.partners.org/complianceline . If the e-mail was sent to you in 
> error
> but does not contain patient information, please contact the sender and 
> properly
> dispose of the e-mail.


___
Freesurfer mailing list
Freesurfer@nmr.mgh.harvard.edu
https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer

[Freesurfer] Linear Mixed Effects model

[Kasper Jessen]

Dear FreeSurfer,
I have a longitudinal study (patients and controls) with two time points (i.e., 
baseline and six-weeks follow-up). I would like to use the LME model as it can 
handle unequal timing and different number of time points across subjects due 
to missing data.
I would like to test if there is an effect of (a) time (b) group and (c) group 
x time interaction.
However, i am trouble with my design matrix X:
(1) How would i write X?
Thanks in advance
Best regards
Kasper___
Freesurfer mailing list
Freesurfer@nmr.mgh.harvard.edu
https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer


The information in this e-mail is intended only for the person to whom it is
addressed. If you believe this e-mail was sent to you in error and the e-mail
contains patient information, please contact the Partners Compliance HelpLine at
http://www.partners.org/complianceline . If the e-mail was sent to you in error
but does not contain patient information, please contact the sender and properly
dispose of the e-mail.