Repeated measures models

Q

Hi,

We ran some rat VBM and got some nice, corrected results but I am not sure we did the right thing. Angela ran a full factorial analysis adding time (rats were scanned twice a few weeks apart), sex and injury (did they have an injury or a sham) as factors.

That all sounds fine however I dont think it takes into account the fact that rats were scanned twice. Whilst there is a time factor that does not mean its the same rat over time. I am really struggling to get this right and I want to because there is alot of data that can be analyzed in the same way.

Do you have any ideas on how to look at the effect of injury, the effect of time and the effect of sex - but at the same time account for using the same rat twice???

A

If you look at models people make, they include a subject variable, which in a design matrix can be a single variable/column for each rat – examples below. That column would have two 1’s for the before and after scans, then all other rows would be 0. The scientific way to discover this is via google images search:

https://www.google.com/search?q=spm+repeated+measures+design+matrix&tbm=isch

😊

In CSPM, it will do this if the measures are repeated, although we don’t use that much.

Or