The estimation engine
Before measuring anyone, it is worth checking that the instrument recovers what you put into it. Here the data are generated, the true parameters are known, and you can check whether the engine finds them again.
What this page is for
You set up a simulated person by moving the sliders: peak height, time to peak, share that never returns. The program generates a set of measurements with realistic noise, then forgets the true values and tries to reconstruct them from the sampled points alone.
The right-hand column compares what you set with what the engine estimated, together with the interval within which the value lies with 95% probability.
Raise the noise, or cut the number of sampled points, and watch the interval widen. That is how an honest instrument behaves: when it knows less, it says so.
The repeatability test
The button at the bottom simulates thirty people measured twice each, and computes how stable each parameter is within the same person. It is the first gate described in the protocol, applied here to invented data to show how it will be applied to real ones.
What this page is not
It is not a result. The data come from a formula chosen for convenience, not derived from people. It exists to show the reasoning and to open the method to discussion: the shape of the curve, the parameters to extract, and the way uncertainty is declared are all open to revision with those who know the phenomenon.