5 ms·
The sample size is so small, that the results are nearly meaningless.
by rwj 7y ago
The sample size is so small, that the results are nearly meaningless.
- jlarocco 7y agoComments like this aren't helpful. Literally nobody is claiming the results are solid evidence for anything other than that one small study had an interesting and unexpected result. And the results aren't "nearly meaningless" - they're an indicator that maybe a larger study should be done. I doubt anybody gets funded for big studies shooting into the dark, so smaller "what if..." studies like this are valuable.
- ALittleLight 7y agoIf 9 people took a substance and died that would be a compelling enough sample size for me not to take thing. Likewise, 9 people reporting a positive result seems meaningful enough to encourage further study (provided effect size is large enough)..
- nikolay 7y agoWith the placebo being so strong, sample size of 9 is too small for any conclusion.
- James_Henry 7y agoDo you know how strong the placebo works on Horvath's epigenetic clock? I don't think that research has been done, but it will be soon I am sure and it will be incredibly interesting to find out. Even without knowing that though, I'm surprised, and so is Horvath himself, that they found a full-on reversal of the clock. I don't think anyone was expecting that. I'm betting we won't see that kind of reversal in a future placebo group, though if we do I'm going to be getting myself some good placebo!
- nikolay 7y agoThe placebo itself is magical enough so it could be extended to pretty much anything and that's why quality studies are done certain way. If can understand the placebo, then and only then we can then explain what it could do and it can't do.
- James_Henry 7y agoThis is false. We can explain a lot without a placebo comparison. Also, the placebo isn't magic, we just haven't studied exactly what it is. It clearly does have limitations: look at all the studies where the placebo only does so much!
- nikolay 7y agoWe have studied it and there are some hypotheses, but that's it.
- ALittleLight 7y agoThe article mentions the effects persisted even a year after treatment. That's far beyond the placebo effect for anything I'm aware of.
- nikolay 7y agoThe placebo effect can definitely extend beyond the end of the trial.
- James_Henry 7y agoSure, but we know how it often acts in trials.
- SamReidHughes 7y agoThere is more to the sample size than the number 9. Every time unit before the experiment where the effects were not seen is a form of sample. For example, suppose nine old people took an unusual drug and then grew 1-3 inches over the next two years. The fact that it happened right after they took the drug, and not some other point in their life, would be one heck of a coincidence, the probability of which depends on the absolute rarity of that sort of event.