Daily samples of a measure- how much variance is there for each participant?

Hi guys,

I have a set of data where someone has collected a measure of performance on a motor task every day from the same participants (N=9) for 28 days. What I would like to see is how much variance there is on a daily basis for each participant. I am a bit confused as to what sort of statistical approach I should take. Ideally, I want to see what sort of 'band' of variance there is for each participant- does this make sense? So, the idea is that we can predict for each participant what sort of range of performance they will fall in to (so simple regression maybe?).

At first, I broke down the measures in to weeks- so had 4 averages and then performed a bivariate correlation (r=.96, p<.001). But this doesnt seem to be the best approach as I am discarding a lot of data with the averaging approach. Can anyone offer advice as to how they would tackle this data set?