/sci/ - Science & Math

Trump's people are literally telling scientists what conclusions to reach, and demanding that the studies be altered to support their political agenda.
EPA researchers had put together a study showing that a particular regulation protecting streams and rivers would end up saving more money (by preserving resources associated with wetlands) than it would cost to implement. Scott Pruitt didn't like this, so he explicitly ordered those scientists to come up with a new study that ignored those benefits.
>https://www.nytimes.com/2017/08/11/us/politics/scott-pruitt-epa.html

Unsurprisingly, Rick Perry did something similar to this report.
A complete draft of it that was leaked found that the power grid was in good condition and that renewables weren't causing problems, and that coal plant shutdowns were driven by the natural gas boom (which is absolutely true). But the """"official"""" version that Perry's people edited tried to say the opposite, only they missed a few spots where the original conclusions were still mentioned.
>https://thinkprogress.org/trump-officials-rewrite-grid-study-ac7d3d188b6a/

Basically, you can't trust any official reports coming out of agencies controlled by Orange Man. The politicians write the conclusions first and try to force the data to fit them.

>>8635278
>>8635279
>>8635281
not to get too Jedi on you, but we have to be better than them. we can't let ourselves be ruled by fear of unseen cabals or rage against the vanity of the ignorant.
keep your mind level. focus your rage. uphold the principles of empiricism and objectivity.

>>8596811
medicine has its statistical limitations (mostly arising from the fact that ethical considerations preclude a lot of human experimentation, meaning that you frequently have small sample sizes with no controls). good researchers, however, are aware of this and are honest about what does and doesn't have robust statistical support.

all fields have hand-wavers who treat all observed trends as real and don't consider reasons that spurious correlations might arise.