3 ms·
> The reality is that the research literature is filled with over 90% jun Based on what? There is a problem of non-replicated papers - but where exactly are yo
by Angostura 10y ago
> The reality is that the research literature is filled with over 90% jun
Based on what? There is a problem of non-replicated papers - but where exactly are you getting your claim of 9 out of 10 - particularly in high quality journals?
- nonbel 10y agoPersonal experience. It is great that the replication issue is getting attention, but really that is a minor aspect of the problem. Producing reproducible reports of your research is supposed to be the easy part... The real difficulty is how results are interpreted. What I was taught to do (along with basically my entire worldwide cohort of research colleagues) is test a "null hypothesis" rather than my hypothesis. Then when the null hypothesis was deemed to be inconsistent with the data, I was supposed to conclude my hypothesis was supported, with at most some hand-waving about other explanations. A crucial part of scientific research is distinguishing between different explanations for the results, and this is almost never found (at least for fields like medicine, psych, etc). If we can get the public eye to look behind that curtain, oh man...