They're likely correlated, but not the exact same sub pops for several reasons, one of which you stated. Mathematically speaking they have to be fairly significant shared sets because of their respective results: you can't get the same divergence from sample population trend without the overlap. (As an aside, this is not driven by the small sample size; the effect is n-independent.) The exact magnitude of the overlap is the only thing that can't be pinned down without the underlying data, but from a statistician's eye it is about 50% at lowest. And anything above 50% means the disentanglement of the correlation and it's covariance has to be done eventually if we think these variables are valid (hence Ariana, plus the bigger Phase 2/3 will help; this is where the sample size n will matter). I don't know if the company will share that kind of granularity. They may at 52 weeks. They may also talk a little bit more about the top responders (e.g. the artist and pianist) and to which subset(s) they belong. I'm watching for that, but not expecting it.
Market Data 
Markets 
AI
