alt Hacker NewsThe point isn’t that they found subgroups, the point is the method they used to find them — namely, analyzing individual brain scans rather than averaging them out first.
The point isn’t that they found subgroups, the point is the method they used to find them — namely, analyzing individual brain scans rather than averaging them out first.
Same sort of thing - how would you expect to find subgroups by assigning the aggregate to individuals. It would be bad design. It's like they're surprised that they used a good design.