During the completion of my MA thesis in cognitive neuroscience my advisor told that, for example, you do not need to get as many (human) subjects as possible, as 500 measurements of 100 parameters on a one subject scientifically is rather comparable to 500 measurements of a 1 parameter in 100 subjects. As he was trained in animal learning, it seems rather logical, as in animal studies to train a rat to press a lever to get a reward the rat needs to perform 100s of random movements to learn that by hitting the lever it gets it and the researcher needs to measure all of them.
So I thought it may be a good idea to build a platform for all sorts of biohackers for (pseudorandomized[1]) decentralized at-home trials, with in-depth standardized data collection protocols, with some ML for personalized predictions and comparisons. Is it really? Would anyone like to participate?
[1]We take into account the dopaminergic system involvement in any "placebo" response
Not following this logic, what happens if there aren't 100 parameters to measure per person for whatever is being studied?