New Comment
2 comments, sorted by Click to highlight new comments since:

What are good rules to identify bad popsci?

The tweet asks to identify bad articles based on the headline. Given that the headline is usually not written by the person who wrote the article, that seems like a bad idea. 


RCTs still have to be justified via “equipoise” that we don’t know whether the results will be beneficial or not. Trials will be ended early if that stops being the case, or will be designed initially to minimize the problem by eventually giving all patients the treatment. The still ongoing COVID vaccine trials unblinded participants once other vaccines became available, for one example. The polio vaccine case they make clear that these solutions don’t work due to the seasonal outbreaks making it all or nothing. But it seems like the obvious solution for them is, if they could manufacture N doses, take 2N volunteers. Presumably volunteers outnumbered doses.

 What I couldn’t find from Wikipedia just now is why confidence was so high ahead of time for this trial.  It wasn’t just Salk: he sold 9 million doses before the trial completed. But there had a already been at least two failed vaccines, one failing from lack of safety and one of efficacy. Why not a third?