> and sharing as reference material I'm hugely in favor of this. Lots of replication-crisis stuff focuses on handling statistical issues like salami slicing. T

> and sharing as reference material I'm hugely in favor of this. Lots of replication-crisis stuff focuses on handling statistical issues like salami slicing. That's great, but it's a one-problem fix. Releasing more comprehensive information is a far, far bigger advance, which enables everything from catching data fraud to verifying statistical analysis. Reinhart and Rogoff, for instance, was afflicted with nontrivial formula errors in the analysis. A standard of releasing data and code with papers would have allowed this to be discovered in months rather than years.