that approach is accurate, but not scalable.

the effort to publish a fraudulent study is less (sometimes much less) than the effort to replicate a study.