Is science a meme?

Is science a meme?

>inb4 someone links me to /sci/

This board has no topic you faggot

Attached: reproducibilitycrisis.jpg (659x729, 121.43K)

Other urls found in this thread:

smithsonianmag.com/science-nature/scientists-replicated-100-psychology-studies-and-fewer-half-got-same-results-180956426/
twitter.com/NSFWRedditVideo

The temptation and pressure is always there to throw out the data that doesn't make a coherent narrative. Nobody wants to publish a paper that says "we tried a bunch of stuff but don't see any clear trends because there were too many complicating factors."

Isn't that the fascinating thing about science?, Trying things that always lead you to another question.

it's because 90% of studies in modern times rely on extremely niche subjects to begin with and then they use statistics to bend the data in their favor. When I was in college, I took a class that taught you how to pick your statistical analysis and how to make it sound as significant as possible. In reality, most of these results barely imply anything or just use such a small sample size that the bias/lurking variables are incredibly difficult to avoid.

Science is controlled by capitalism now. If scientists can't get their article in a journal, because their hypothesis was wrong, they'll lose their job, so they have to fudge the data. And nobody is redoing published experiments to see if they get different data because it's not as exciting, despite being necessary. What you've got is a bunch of contradictory theses that haven't been checked by anyone else.

Lol BS explain to me how the NSF is capitalist.
Whenever I've had research funded by a company you sign all sorts of NDAs and very little of the work is published.

gravity's not real. that shits in your head!

Well an unfortunate number of people are in the field not because they have a passion for science but because they want to be scientists.

Also experiment costs can add up quickly, there's not enough time to run additional experiments, have to have data to publish before the end of program reviews etc.

>did our experiment work?
>no, another failure!
>you fleshbag, you forgot to incense the pitri dish!
>Ah, there it goes; praise the Omnissiah!
Are we /chadmech/ yet?

Attached: Happy Ones and Zeroes.jpg (500x418, 52.27K)

Yep, science is a meme. It's just the new religion now, mostly spouted by idiots who have no idea what they're reading, believing whatever they hear from their pastors("experts") without understanding a single thing about it.

As for scientists themselves, they are mostly underfunded and, but this is more a problem with non "hard" sciences. Those are the ones that are really easy to bullshit. You'd be wise never to believe anything about social sciences or psychology.

If you've ever been in a lab this bullshit will not surprise you, people are awful and statisticians are worse.

it funds its operation by selling the journal, aka capitalism
and journals sell better when they contain mind-blowing results
so journals are more likely to refuse you if your conclusion isn't something cool and epic

Everyone knows psychology is a joke, but let's not throw out the baby with the bathwater. Chemistry, physics, neurology, etc. are all still reliable sciences.

I thought neurology was fairly new and therefore not all that reliable.

human brain might have reached its capacity of figuring shit out so many scientists probably just make shit up these days

>Chemistry, physics, neurology, etc. are all still reliable sciences.
Citation needed

Neurology definitely has a long way to go, but from what I've read they usually do a good job of taking that into account in their studies. They don't claim to have figured out depression or whatever, mostly just "we did X and saw increased activity in section Y." As such, neurology experiments aren't exactly groundbreaking, but also not false.

The reliability of science is directly proportional to it's hardness (i.e. use of math). Physics is the most reliable science due to it's intense mathematical rigor, and chemistry comes in second. I agree with you on neurology though.

science isn't the be all end all of the universe. science isn't a meme but it's not 100 percent right.

The answer is simple: science tends to be on the cutting edge of discovering new things. The more novel your research, the harder it will be to replicate, because your findings are more likely to be the first. If scientists just went around writing papers about things we already understood really well, guess what? The replication rate would be 99%. Is that better?

Psychology only replicates slightly worse than the hard sciences, but it has a stronger statistical power.

I see. However, I think there's a particularly strong incentive to fudge data in the field because of it's political/social implications.

Neurology is still objective, though. Measuring activity in the brain doesn't have room for bias. It's when people try to apply an interpretation as to why those increases / decreases of activity happen that you get bias.

There is a huge issue with reproducibility with medical research.

>Psychology only replicates slightly worse than the hard sciences

smithsonianmag.com/science-nature/scientists-replicated-100-psychology-studies-and-fewer-half-got-same-results-180956426/

And when don't they try to interpret the data? All science is subject to a degree of bias insofar as the data has to be interpreted.

Medicine and engineering have a 55% replication rate. So?

publish or perish
you have to write papers or else you're considered ineffective so you write safe, incremental improvements or carefully play with the numbers

>Is science a meme?
Science used to be cool and "reliable" but now it's some lab results with a bunch of manipulation. Especially if private companies pay for it.
This.

Attached: smart science man1.png (479x359, 191K)

Most people who fund research don't want to hear "yeah, it didn't work at all"

>Physics is the most reliable science due to it's intense mathematical rigor
that still leaves room to completely fuck up the scientific part. you can make equations for bullshit but its still bullshit

even computer science should have rigor but people can't reproduce AI results