scientific political bias

you can use statistical analysis of your results to get whatever result you want, take a look at data dredging

>>8295979
It honestly doesn't matter because the data

>>8295988

always shows through anyways

it just makes it INSUFFERABLE to go through

Imagine reading paper after paper and saying it's shit only to have some idiot play wordsmith and spin it another angle

It's basically where politics come in, and that's why bias is very important to take into consideration. The data is there, but you have to see through the lies and read between the lines

>>8296013
>it just makes it INSUFFERABLE to go through
This, for example:
https://en.wikipedia.org/wiki/Study_329

>>8295979
Because science relies on statistics to find truth and you can lie with statistics if you are motivated to do so.

>>8295979
>Surely if the logic in the paper follows then it doesn't matter who funded it

There's more ways than one to conduct a study and even more ways in which you can slant the presentation of the results.

File: 20100115-talebone.gif (43KB, 245x300px) Image search: [Google]

43KB, 245x300px

>>8295979
Their own minds.

>>8295979
The problem isn't science per se, but the publics (and the media's) understanding of how science is done. That is to say there's far too many people that think when a study is published the conclusions are affirmative and not just indicative. That's not to say there isn't some malfeasance (holy shit this >>8296028), but I think the former problem is bigger.

In studies on drug effects in humans you can be biased in the patients you select, whether they are allocated to placebo or active treatment, in how you judge whether the drug was succesful or not and so on - bias can be introduced in every stage of the process.

These days pharma sponsored papers are usually pretty good - they are very well funded and under a great deal of scrutiny from regulators. Most of the really egregious bias or downright fraud comes from shitty groups/institutions with incentives to publish anything positive or just get involved with the latest fad flavour of the month.

In general the industry is probably more transparent than it has been in the past, particularly with the disclosure of incentives/competing interests and efforts to minimise publication bias (the overrepresentation of positive studies in the literature) with registries and prespecified outcomes.

>>8296208
this is a very popular blog by a medicinal chemist mostly about the pharmaceutical industry, drug development and clinical trials
http://blogs.sciencemag.org/pipeline/

>>8296194
>I think the former problem is bigger.
Not even close, the main issue is that "everyone" wants a study that says

>"HOLY FUCKING SHIT LOOK AT THIS! ANTIGRAVITY AND A CURE FOR CANCER!"

or alternatively something very bleak like

>"Tap water found to be main reason for EVERY SINGLE DISEASE!"

Because that's how you get funding, how people make money, how you gain rockstar staus and so on.

A study that concludes

>"nothing of interest found, whole lot of money wasted"
Is not really publishable, and could even end your career.

tl:dr Clickbait advertisement mentality is too prevalent and severely degrades the quality of science.

>>8296213
>Not even close
But the two examples you give are direct conclusions of the problem I point out.

>Clickbait advertisement mentality is too prevalent and severely degrades the quality of science.

Again, that only works because the public don't understand how science is done.

>>8295979
Exactly, "Ï trust no study than mine I faked on my own."

>>8296226
It's not about the public it's within the fields that people chase the fancy headlines.