Good research is not collecting facts. It is learning how to relate claims, evidence, incentives, and unanswered questions.