3 ms·
Does it need fixing? No, why should it? If science could be done better, yes, we should fix it. The author identifies a very real problem - publication bias an
by yummyfajitas 15y ago
Does it need fixing? No, why should it?
If science could be done better, yes, we should fix it. The author identifies a very real problem - publication bias and data mining, and proposes a solution, namely pre-registration of experiments.
Only one of his mechanisms for enacting the solution (#1, "clinicaltrials.gov option") is centralized.