Just want to add one more thing... Most people assume that neurovascular coupling is a linear system, which is often a valid assumption. I think the question that you are concerned with is whether the psychological and/or neuronal processes are linear with ITI, rather than whether neurovascular coupling is linear with ITI. The best thing to do is to measure whatever psychological process you are studying and see if it is linear. For example, look at RT or accuracy and see if it changes with ITI. It would be optimal if you could stay in the range of ITIs where RT/accuracy remains constant. If this isn't possible, you should explicitly model these non-linear effects, though this requires additional assumptions about whether the "linear" component is the same across ITIs.
jack
|