You can find some statistical power with any number of observations. Doesn't mean results can be generalized, which is what the article wants to do. 20 extremely low. At that level it is more reasonable to ask why…
N=20
Well, you're right, i did overstate. I think the researchers and journalists did too. Perhaps i would not be so annoyed with this, had the journalists specifically, just worded it in less bombastic terms.
MRI N= 17 (23 control) The study is as precise and generalizable as 17 anecdotes...
It's not a very good study, i made a Twitter thread om why: https://twitter.com/TheAntsSleep/status/1420313386252972045?...
You can find some statistical power with any number of observations. Doesn't mean results can be generalized, which is what the article wants to do. 20 extremely low. At that level it is more reasonable to ask why…
N=20
Well, you're right, i did overstate. I think the researchers and journalists did too. Perhaps i would not be so annoyed with this, had the journalists specifically, just worded it in less bombastic terms.
MRI N= 17 (23 control) The study is as precise and generalizable as 17 anecdotes...
It's not a very good study, i made a Twitter thread om why: https://twitter.com/TheAntsSleep/status/1420313386252972045?...