Hi Harry,
I assume you ran the 2-stage-model , i.e. long_mris_slopes script (else you would not have the long.thickness-avg, rate ,... files.
As described here: http://surfer.nmr.mgh.harvard.edu/fswiki/LongitudinalTwoStageModel and http://surfer.nmr.mgh.harvard.edu/fswiki/FsTutorial/LongitudinalTutorial
this model first reduces the analysis to a single variable per subject (e.g. the rate or percent change across time) using a linear fit in each subject. So in your case with 3 time points a line gets fit into the data (all three time points) for each subject and the rate is mm/year (if years is your time unit). You would then compare this rate across groups in the second stage.
If you rather want to do pairwise analysis, you can - simply create qdec.long tables with only two time points and run the same stuff (make sure you name your output differently each time). - You can run a simple analysis as described here http://surfer.nmr.mgh.harvard.edu/fswiki/RepeatedMeasuresAnova - Or you can create more complex models (e.g. piecewise linear, with a node at the second time point, if you are interested in detecting a change in slopes at the second time point) using mixed effects models: http://surfer.nmr.mgh.harvard.edu/fswiki/LinearMixedEffectsModels
An overviewon methods can be found here (including advantages / disadvantages): http://surfer.nmr.mgh.harvard.edu/fswiki/LongitudinalStatistics
Best, Martin
On 04/01/2013 11:40 PM, Harry Hallock [hhal5293@uni.sydney.edu.au] wrote:
Hi,
I'm currently doing a long analysis of a data set that contains two groups (training and control) and 3 timepoints (baseline -BL, follow up1- FU1 follow up2-FU2)
I've put it through qdec. But for the analysis qdec only gives me one measure for long.thickness-avg, rate, pc1 and spc. How do i know which timepoints are being analyzed? i.e. is it simply FU2-BL? Or is it FU2-FU1 or FU1-BL? Is there a way i can test for each of these changes over time?
Cheers, Harry
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer