| Parachutes are a device for saving lives. We know they work. We have evidence of efficacy. But this evidence does not come from randomized controlled trials. If it was a pill, the standards for evidence should not change. Randomized controlled trials make things easier to prove. But they are not the only way to collect evidence of efficacy. > And your final example of showing up in scientifically illiterate society wielding a miracle drug and declaring that they need no evidence of its efficacy, ignoring the fact that such evidence was all on you prior to arriving there? No evidence?? That's the exact opposite of what I'm trying to say. Your objections... okay, would a better thought experiment be an invention of a new antibiotic? There is no existing proof of anything. I go to the nearest town and give the pill to every single resident with gonorrhea, 50 of them, and a few days later only 48 of them have the disease any more. They otherwise took no drugs they hadn't already been taking for months. I didn't randomize at all, and I didn't set up a control group. But it's statistically impossible for that to happen by chance. Also, I and a team of respected researchers searched for confounding factors just as hard as someone performing a randomized controlled trial, and we could not find any. Did I provide evidence of efficacy? If not, why not? |