|
|
|
|
|
|
by sciguy77
3943 days ago
|
|
|
First off, there's no need for all-caps. This isn't 4chan. > The fact that assumptions are considered as some unavoidable, forgivable, intricate part of science is part of what fuels anti-science and politics. No one here, as far as I can tell, is saying, 'oh well, science is full of assumptions therefore science is invalid.' The problem is not with science in general being valid or invalid, but rather with the sorts of experiments being conducted right now. Studies that are not replicable most of the time are bad science, which is different than science itself being 'bad.' I'm not saying such faulty experiments and studies are unavoidable and forgivable. Quite the opposite, they're flawed and need to be scrutinized more, not less. Finally, this is not an incurable problem. Bayesian math is one potential solution. There are others. |
|
|
No, they are saying "science is based on assumption, but in this case 'many scientists' were making the wrong assumptions." The title should have said "scientists find errors in p-values as premises." Assumptions are avoided, not depended upon. Bad assumptions would certainly lead to irreplicable experiments also.
The distinction you make between science and its experiments is not a common distinction. If science = experiments, then you totally agree that this kind of science is bad science. Which was precisely my point.