Hacker News
new
|
past
|
comments
|
ask
|
show
|
jobs
|
submit
login
solveit
on Nov 13, 2020
|
parent
|
context
|
favorite
| on:
Effect of exercise training for five years on all ...
Statistical significance is not binary. And, personally, I have more confidence in a p=0.07 study than a p=0.048 study
CGamesPlay
on Nov 13, 2020
[–]
> And, personally, I have more confidence in a p=0.07 study than a p=0.048 study
Why on earth would that be?
solveit
on Nov 13, 2020
|
parent
[–]
There's a depressingly good chance that the p=0.048 was p-hacked while we can trust the p=0.07 to be what it says on the tin.
Guidelines
|
FAQ
|
Lists
|
API
|
Security
|
Legal
|
Apply to YC
|
Contact
Search:
