Dear Sophie,

If the main outcome of interest is the *difference* you can only use subjects that have data at both time points.  For such subjects, analysing the differences in a regression model (not a 2x2 anova) will work, including any covariates you like to account for variation in the 104w - 52w difference, including, of course, a patient-control group difference.

If you have interest in cross-sectional effects, of gender, baseline age, using all available subjects, or exact time effects (if not all subjects are exactly 52/104 weeks) I think you may want to check out the SwE toolbox for longitudinal data... it doesn't have any problem with imbalanced data.  There's more here: http://nisox.org/Software/SwE

-Tom

On Tue, Nov 13, 2018 at 6:16 AM Sophie <[log in to unmask]> wrote:

I'm hoping to run a longitudinal analysis looking at 2 groups (controls and patients) at two time points (baseline and follow-up) for an fMRI working task in a developmental sample. Some individuals only have follow-up data at 52 weeks and others only have at 104 weeks. I am hoping to optimize sample size by using whichever follow-up time point is available. I assume I need to include a covariate for time to account for this.  I'd also like to add in a few other covariates such as sex and neuropsych score.

My current plan is to:
1) Create difference contrasts for each group (ie: group baseline > group follow-up).
2) Run 1 sample t-test on the difference maps with time, sex, and neuropsych as covariates.

I'm only interested in looking at the groupXtime interaction while adjusting for sex and neuropsych score. Is this analysis plan appropriate? I reviewed the listserv on 2x2 rm anovas and the suggested model in which I put both groups in the model and look only at the interaction and main effect of time doesn't seem viable given the time point issue. If my plan is incorrect, are there other options I can use?

Thanks in advance for the help!

########################################################################

To unsubscribe from the FSL list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=FSL&A=1

--

__________________________________________________________

Thomas Nichols, PhD

Professor of Neuroimaging Statistics

Nuffield Department of Population Health | University of Oxford
Big Data Institute | Li Ka Shing Centre for Health Information and Discovery
Old Road Campus | Headington | Oxford | OX3 7LF | United Kingdom
T: +44 1865 743590 | E: [log in to unmask]
W: http://nisox.org | http://www.bdi.ox.ac.uk

---

To unsubscribe from the FSL list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=FSL&A=1