One result

Research round. Explain one result you trust, in four parts: the claim, the comparison, the evidence, and the limit. The interviewer asks what single observation would make you drop the claim. Answer with that observation, not a general doubt. Then they ask what the authors did not show. Name one missing comparison. Then they ask whether you would ship a product on this result today. Answer yes or no, and the reason is the falsifier you already named.