This would be (barring outright fraud) nearly harmless if we had preregistration as a prerequisite for studies 'counting' (in meta-analyses, in discussions here

This would be (barring outright fraud) nearly harmless if we had preregistration as a prerequisite for studies 'counting' (in meta-analyses, in discussions here, in academic rankings, etc) http://theness.com/neurologicablog/index.php/registering-studies-reduces-positive-outcomes/ http://theness.com/neurologicablog/index.php/registering-stu... I suppose they could still add ridiculous interpretations (as is often the case when people's livelihood depends on their work being seen to promote a given side), but at least you could trust the data.