Variance between two time series (action and feedback)

#1
I would like to know how find the variance between two (or more) time series.

I have a series of timings (in milliseconds) which represent the timings of a subject's button presses. I also have series of milliseconds timings which represent the timings of an auditory tone which could either been produced by a subject or the experimenter. In the case where it is the subject that has produced the tone the difference between the action timing and the auditory tone timing should be 0, but I have conditions where the subject has less control over the tone (and the experimenter has more control). Those cases should result in larger differences (in ms) between the two time series (as the experimenters button presses will be different to those of the subject).

What I need is a measure of variance (e.g., ranging from 0-1) between action timings (button presses) and tone timings as a way of assessing how much control the subject actually had over the tone. This not as straightforward as I initially thought as it seems that different types of situations need to be taken into account (action with/without tone; tone without action...) . What the previous event was also seems important (action with/without tone; tone without action...).

I use r so any examples/thoughts in r would be particularly welcome.

CdB
 
Last edited: