The SRK Science Thread 2.0


#1042

Scientists Hatch Bold Plan to Save Planet From Supervolcano
https://www.nbcnews.com/mach/science/scientists-hatch-bold-plan-save-planet-supervolcano-ncna799166?cid=sm_pd_nn_ky_mc-a_170720&kwp_0=584631

NASA researchers say siphoning heat from the Yellowstone Caldera could lower the risk of a deadly eruption while generating electricity.


#1043

**Odors Alter Subjective Time Experience
**

http://english.cas.cn/newsroom/research_news/201705/t20170515_177067.shtml

Oh my science thread, how I’ll miss thee…

Never stop being curious guys!

https://encrypted-tbn0.gstatic.com/images?q=tbn:ANd9GcQQi6h8OZd0blEJzQK6Q458X--eekVu0QaLigVpJM5DsidzvYsj


#1044

My wife got her phd in pharmaceutics - thats like how drugs are metabolized by the body. since then i saw that there are some serious problems with science, confirmed by talking to other phds and doing some additional reading.

http://www.bbc.com/news/science-environment-39054778

A shit ton of studies cannot be replicated. This was especially bad when big pharma companies tried to replicate academic research results, and i think less than half could be reproduced.

Now I’m not saying vaccines cause autism. if anything, the discredited shit study that showed vaccines are bad is an example of exactly what the problem is. too much garbage going on in research. There are three main issues:

  1. There is no award for replicating a result. The goal in research is to get your shit published, so you can increase peen size (or shrink vag size if youre a woman scientist) and get more money. So if youre first to a result, great, and if youre second, nobody gives a fuck. this discourages people from pursuing same avenues of research

  2. You are punished if your hypothesis is wrong. A very critical part of the scientific method is testing your hypothesis. Yet if you are a phd student and your hypothesis is wrong, you can either spend another few years of your life wasting away to graduate, or just linear regression the fuck out of the data until hey look theres a trend with low p values if you squint hard enough! And if youre a post doc or researcher, nobodys going to publish a paper on how x,y,z did NOT work, which is a TRAGEDY, because NEGATIVE RESULTS ARE JUST AS VALUABLE AS AN EXPERIMENT CONFIRMING A HYPOTHESIS

  3. NIH grants. yes big government is bad. when talking to a lot of phd students, getting grants from the government relies on your hypothesis being correct (point 2) so you can build up on previous research and keep getting money. so if the original study was erroneous but government gave you money, you want to build upon the story, even if your data is bogus and can’t really be replicated, unless you squint your eyes and hand wae the math.

  4. there is no reward for just tinkering. in the book anti fragile by nassim taleb, he goes over these issues, and also discusses how there is a myth that academic research leads to industry revolution, where its really the reverse. its small, low level tinkering around that lead to many great inventions and discoveries. today’s research grants aren’t incentivizing that, its incentivizing research stories that build upon previous stories, even if the foundtion is shoddy (point 3)

  5. a lot of research publications are hidden behind paywalls like pubmed. DESTROY!!

So bringing it all together, today i am skeptical of initial studies unless its published in a high impact factor journal like cell, nature or science. older articles are legit if they’ve been cited multiple times.

How to solve this in future? this will require a lot of work. I think some way to incentivize replicating results, publishing negative results (hey my experiment didnt work can i have a phd anyway? can i have research money anyway), and small random tinkering as described in the book antifragile, will all go a long way. and destroy the paywall. something like an open source model to science where people can publish their results, other people can then vouch for replicating those results, and people whose results have been verified multiple times would have higher reputation, which would in turn increase reputation of people whose experiments they’ve verified.

thanks for listening!


#1045

I do tend to recall that articles published in the highest impact journals are more likely to be recalled than in the slightly lower ones, simply because Nature and Science gets to pick the coolest, most groundbreaking headlines that often turn out to be overblown or straight-up disprovable before people try to replicate the results. Yes, the leading scientific journals are tabloid as all fuck. The irony is mindblowing. Until you consider that several of them have profit margins of over 30% (!!!). Turns out having people whose work are funded by public money, turn over the rights to their work to you for free, and then charging them to get to access to it again later (which they have to do if they want to reference it), is a pretty easy way of making money.

Another big problem is that scientists aren’t incentivised to actually solve these problems. Publishing in Open Access-journals is punished both because it gives them lower chances of getting grants and because it’s often less expensive in the short term: OA-journals tend to require more money upfront to help pay for the publishing, while the big journals tend to be swimming in cash and don’t need to charge as much for the publishing process. There’s also little political will behind this, since impact factor is something that can be measured (in fact, that was the entire point of it when it was introduced in the 80s: making it easier for politicians to compare scientists by having a number attached to them) and can thus be used in political debates, and as I addressed last page, this issue really only directly hits the people working in science even though the societal implications are huge.

It’s a tricky subject. :<


#1046

but I would say that its a good thing that articles get recalled. because nature and science are the big publications, people will be wowed by the results and actually try to replicate and invalidate. the peer review process is good and integral, but it is prone to errors and politics. like that stem cell scam from japan a few years ago. how did that pass peer review? whereas for lesser journals, a phd student trying to graduate might get rejected at other publications but go to a lower tier one because its more likely to slide under the radar, versus coming up with a new thesis and starting from scratch. i had no idea about these issues until i met my wife and then i started talking to my other friends who had phds. One guy who was a post doc actually quit his lab to become a real estate photographer because he didn’t want to publish his research because it was too sketchy. not sure the solution but its good to raise awareness of the issue. my hope isn’t that people go: “HA VACCINES CAUSE AUTISM” but realize that science is the best, and as such the problems with academia and research can be corrected.

this is not a big of a problem with physics or math where you can “prove” a result, or computer science where you can similarly either prove something, or release software that people can run and tweak themselves very easily


#1047

The problem is obviously not that these studies are getting replicated, that’s a good thing. The problem is that it’s as common as it is specifically in the high profile journals compared to lower profile journals. There’s such a high pressure to publish in these journals that scientists are, knowingly or unknowingly, publishing results that should not be good enough for the highest standards of scientific publication. Likewise, there are problems coming from the publications as well, who should require better scientific rigor from the articles they publish. It’s problematic and sets a bad precedent. Unfortunately, there are very strong incentives for both parties to just keep on doing as they are, and it hurts science as a whole.

I think we’re mostly in agreement on this subject though, the detail of how trustworthy high profile journals are aside.


#1048

Still think chess was invented by a god or a goddess… Maybe some other higher being…


#1049

Reproducibility crisis is science is well known among… Scientists. The general public that reads “cool science” articles will only be impacted. When it’s too late.


#1050

I think you’re going to like this @fishjie


#1051

https://www.yahoo.com/news/vets-fish-skin-heal-bears-015222079.html

wrappings made out of fish skin to treat burns.

wtf! the images are the wildest shit ever. no pun intended.


#1052

Yeah, it’s not a new concept. Looks weird, but it drastically increase survivability rates of server burn victims and reduce recovery times. Also pig stomach lining is used as skin graffs


#1053

More proof people need to shut the fuck up about GMOs

The Fruit and Veggies we have now don’t exist at all in nature.


#1054

ah yes i remember reading about that in the book parasite rex. cause we evolved to fight parasites when there are none the immune system overreacts in the forms of allergies and other annoying shit. or like how sickle cell anemia is resistant to malaria. also parasites are the best explanation for sex. if there weren’t parasites, evolving sex wouldn’t make sense, cause asexual would be easier. but asexual means everything same same and gets pwned by parasites


#1055

oh shit are you a scientist?!?!?!??!!

well i was mainly saying if its high profile, at some point someones going to bother reproducing the findings. but a lot of phd students (not my wifes university lab when she was getting her phd or anything…) or postdocs can publish in lower profile journals and fly under the radar… like peer review might get it rejected at many publications but sooner or later it gets in, even though the statistics might have been played around with until some curve or line fit the hypothesis… speaking in generalities here, after i started learning about this stuff and talking to my other phd/post doc friends i learned both how cruel and soul crushing a phd is, and also how many of them didn’t really think much of their own thesis. but they just wanted to graduate

edit: hilarious website about grad school, sort of related to science since STEM grad students are slaving away doing bench work all day long


#1056

and so it begins


#1057

#1058

Heavy Metal is a Earth moving experience

https://youtu.be/vxfJbW6KDp4

Also what got cut out of the video


#1059

Alzheimer’s disease reversed in mice, offering hope for humans, new research shows


#1060

Hello Science thread, long time.

Been out of the loop for a while, but here’s a link to a MinutePhysics series I enjoyed, about relativity (3 chapters at the time of posting):


#1061

This is pretty damn cool; the applications this can have are endless!