Amos,
I think, the problem is the word 'continuous' in my too short post. 'Continuous outcome variables' (as opposed to binary) would have been much clearer. Examples: pulse, blood indicators, etc.

It seems to me you have understood that one outcome variable is measured continuously over time. I have meant that a variable that can take many values on a continuum is measured (once, or even more often).

If then, the effect size decreases with the sample size that is an indicator of a couple of possible artifacts inflating the real effect size. And that would also hold in the world of electrons, BTW.

Wolfgang