3 ms·
Reproducibility and reciprocity | Simply Statistics
- tel 14y agoThis is incredibly important to any kind of open science/transparency movement. As an even more stark example, imagine the problems with scientific rigor and transparency to a risk-bearing institution like a pharmaceutical company interested in defending a drug. If they publish their data openly and with easy access it becomes easy for an outsider to "discover" side effects or reduced effect sizes. In some cases these secondary analyses might be damning. In others, like the ones highlighted in this article, they might also be methodologically flawed. In either case this situation disincentivizes drug companies, even if you assume complete good faith and forward thinking, from publicizing their research in a too easily accessible format. ClinicalTrials.gov is a regulatory measure to enforce transparency, but it's difficult to understand what incentives could possibly motivate a risk-bearing evidence producer to submit to this transparency meaningfully. Even if you're completely confident your drug is side-effect free, it's expensive to fight off critics armed with your own data.