If you can read this, either the style sheet didn't load or you have an older browser that doesn't support style sheets. Try clearing your browser cache and refreshing the page.

(The New Yorker)   You can love science AND question it. As a matter of fact, you SHOULD love science and question it   (newyorker.com) divider line 93
    More: Obvious, science journalist, Neanderthal DNA, string theory, molecular biology, John Horgan  
•       •       •

1639 clicks; posted to Geek » on 09 Nov 2013 at 4:37 PM (44 weeks ago)   |  Favorite    |   share:  Share on Twitter share via Email Share on Facebook   more»



93 Comments   (+0 »)
   
View Voting Results: Smartest and Funniest

First | « | 1 | 2 | » | Last | Show all
 
2013-11-09 03:51:15 PM
That's a good idea.  We should come up with some kind of method that formalizes that.
 
2013-11-09 04:04:55 PM
As opposed to gods, who take it all personal when you do that.
 
2013-11-09 04:26:54 PM
Yeah, science!
 
2013-11-09 04:34:05 PM
Yes, the brainstrust in the Yahoo News comments loudly lambasting scienticians, calling for creationism in schools and mandatory burning of fossil fuels (to show them damn dirty hippies), and refusing to let their snowflake's precious bodily fluids be contaminated with autism juice, all because they heard a fourth hand account of a scientists work reported by a "journalist" with dubious credentials is the same as peer review and scientific consensus.
 
2013-11-09 04:54:17 PM
A key component of empirical thought is questioning your methodology and re-evaluating conclusions?

tnation.t-nation.com
 
2013-11-09 05:02:57 PM
If I conducted an experiment that disproved a major theory, I would be made famous.

I don't see why people think science is conspiracy lol.
 
2013-11-09 05:03:35 PM

verbal_jizm: Yes, the brainstrust in the Yahoo News comments loudly lambasting scienticians, calling for creationism in schools and mandatory burning of fossil fuels (to show them damn dirty hippies), and refusing to let their snowflake's precious bodily fluids be contaminated with autism juice, all because they heard a fourth hand account of a scientists work reported by a "journalist" with dubious credentials is the same as peer review and scientific consensus.


There's no harm in believing creationism, and people have the right to decide what they're injected with. Deal with it.

Reason will always prevail in the end, and science has always been about arguing with everyone, not agreeing with everyone. If you prefer to ignore or even ban certain opinions instead of arguing them, then you have no place in science. It can be a long hard road sometimes arguing with people constantly about the same topics, but if you are truly right you will be shown as such eventually. Maybe 50 years after you're dead, but still.
 
2013-11-09 05:07:31 PM
J. Frank Parnell:
There's no harm in believing creationism, and people have the right to decide what they're injected with. Deal with it.

There is harm believing in creationism because to do so you have to reject science entirely. It isn't a competing theory backed by evidence lol, In fact all of the arguments in favor of creationism are "science is wrong".
 
2013-11-09 05:09:22 PM
Because funding for the sciences is so scarce, especially theoretical sciences like subatomic theory or paleontology--fields that can't easily produce a weapon or recreational drug--researchers have had to turn themselves into carnival barkers. They talk up every minor advance in the media, make it sound like they're about to invent light-speed travel tomorrow, and hope to FSM they can scrounge up enough money to keep the heat on.
 
2013-11-09 05:11:16 PM
It's science. If you're not questioning it then it's not really science now is it?
 
2013-11-09 05:18:34 PM

devek: There is harm believing in creationism because to do so you have to reject science entirely.


No you don't.

God created evolution. Boom, done, moving on.
 
2013-11-09 05:20:28 PM

doglover: devek: There is harm believing in creationism because to do so you have to reject science entirely.

No you don't.

God created evolution. Boom, done, moving on.


When people hear or talk about creationism, they usually mean *young earth* creationism.
 
2013-11-09 05:22:38 PM

Mentat: That's a good idea.  We should come up with some kind of method that formalizes that.


devek: If I conducted an experiment that disproved a major theory, I would be made famous.

I don't see why people think science is conspiracy lol.


The problem isn't a conspiracy or that there is no system for checking work that someone has done. It is that the work isn't being checked enough. Sure, the big once in a life time things will attract replication like a dog attracts fleas. But the small things? A lot gets published without ever being replicated. That way a lot of things that are demonstrably false don't get found out about because there is no one demonstrating anything. This one "scientist" merely made up some numbers and the peer review saw that the maths worked out without and the conclusion wasn't to outlandish and cleared it for publication.

Peer review and articles were once imagined to be a replacement for actual demonstrations. To reach the larger scientific community, as only a few people would fit in a laboratory at a time. The article would be read and people would actually check if things were true and write rebuttals. Over time, however, the system changed subtly. "Peer reviewed" became synonymous with "this is true" instead of "someone checked my maths for major errors and maybe someone else should actually replicate my study." Just keep an eye out and you'll find it everywhere. Even on Fark people will ask for "peer reviewed" articles as the final arbitrator of truth. Not reports of replications. Peer reviewed. As if the two people who were nowhere near the experiment (and often don't even work in the direct field) and only read the information provided could actually make an informed conclusion about the validity of the content.

Things need to change, especially now that we have the internet and electronic communication. Every publisher could add a counter to an article to show how often it has been successfully replicated or refuted. They could even add references to the articles. That information would be a lot more useful to see how reliable an article is than the vague "peer reviewed" status that has been elevated in importance to a position that is entirely undesirable.
 
2013-11-09 05:23:06 PM

Mentat: That's a good idea.  We should come up with some kind of method that formalizes that.


Exactly.

Too bad you scientards are too busy accepting everything science says like the lemmings you are.

/jk, jk
 
2013-11-09 05:25:12 PM
J. Frank Parnell:

There's no harm in believing creationism, and people have the right to decide what they're injected with. Deal with it.

Well, until someone gets a disease and passes it on to you because you had a compromised immune system, so you couldn't get the vaccine, and you end up in the hospital. And nothing wrong with creationism until you run into an antibiotic-resistant bacteria, other than that it's perfectly harmless.
 
2013-11-09 05:33:46 PM

Felgraf: doglover: devek: There is harm believing in creationism because to do so you have to reject science entirely.

No you don't.

God created evolution. Boom, done, moving on.

When people hear or talk about creationism, they usually mean *young earth* creationism.


Really? No method for checking? That's what you're going with? Peer review seems to work pretty well, but what would I know.
 
2013-11-09 05:34:13 PM

To The Escape Zeppelin!: It's science. If you're not questioning it then it's not really science now is it?


This is exactly why I hate the whatshisname picture with the "science is true even if you don't believe in it" macro. Science isn't true. Very little that has been discovered is actually true. It is truer than what we used to know and it is true enough to do a lot of cool stuff. But "science" isn't "truth." Do you think that we'll ever get nuclear fusion to work if we used the various "constants" that were "scientifically discovered" and "verified" 150 years ago? We'd get really close to a large hole in the ground, if we got anything to happen at all. Things get replaced and updated all the time. Even the universal constants have been edited many times. And you know what? None of those values are true, those values are merely the best we can do at this point.

People keep seeing that farking macro and go "Well, a smart man said it and he should know what he is talking about." Or people lash on to it in order to get people to get people interested in science. But at the same time those same people are instilling the idea that the science, as done by predecessors, is infallible.

I started reading the Foundation series last week (I know, the book has been recommended thousands of times and I should have read it sooner) and it perfectly illustrates this issue. Science is dying in the empire because everyone is reaching back to previous scientists. With the question about archaeology and the location of the Birth Planet of the human race being the most direct example given. Science becomes reading what other people have done and drawing conclusions from that. Doing the work yourself is considered useless because surely those scientists of olden days did the work already? There are tons of books and references to be found, and those are held to be rather infallible to a degree. It no longer is about seeking what is right and what is wrong, but providing a lot of information and drawing a conclusion about what other people think is right or wrong. With no effort being made by people to make discoveries of their own.
 
2013-11-09 05:34:45 PM

DerAppie: Over time, however, the system changed subtly. "Peer reviewed" became synonymous with "this is true" instead of "someone checked my maths for major errors and maybe someone else should actually replicate my study." Just keep an eye out and you'll find it everywhere. Even on Fark people will ask for "peer reviewed" articles as the final arbitrator of truth. Not reports of replications. Peer reviewed. As if the two people who were nowhere near the experiment (and often don't even work in the direct field) and only read the information provided could actually make an informed conclusion about the validity of the content.


It's intellectual laziness. Just tell me what to think, there are TV shows to watch.

With the wealth of information online, even full university courses, anyone can have a decently educated opinion on even the most sciency of topics. We all don't need to look to some 'expert' or some agency to tell us what to think.
 
2013-11-09 05:35:15 PM

machodonkeywrestler: Felgraf: doglover: devek: There is harm believing in creationism because to do so you have to reject science entirely.

No you don't.

God created evolution. Boom, done, moving on.

When people hear or talk about creationism, they usually mean *young earth* creationism.

Really? No method for checking? That's what you're going with? Peer review seems to work pretty well, but what would I know.


Fark. Meant to quote DerApple.
 
2013-11-09 05:36:19 PM

devek: If I conducted an experiment that disproved a major theory, I would be made famous.

I don't see why people think science is conspiracy lol.


Hey - I had to work hard to get these orthodoxic robes! Learning the secret handshake upon reaching Level 7 was no walk in the park either.

/Kiss the ring or be labelled heretic.
 
2013-11-09 05:36:36 PM
Jesus loved science.
 
2013-11-09 05:37:12 PM
i.imgur.com
 
2013-11-09 05:37:39 PM

J. Frank Parnell: verbal_jizm: Yes, the brainstrust in the Yahoo News comments loudly lambasting scienticians, calling for creationism in schools and mandatory burning of fossil fuels (to show them damn dirty hippies), and refusing to let their snowflake's precious bodily fluids be contaminated with autism juice, all because they heard a fourth hand account of a scientists work reported by a "journalist" with dubious credentials is the same as peer review and scientific consensus.

There's no harm in believing creationism, and people have the right to decide what they're injected with. Deal with it.

Reason will always prevail in the end, and science has always been about arguing with everyone, not agreeing with everyone. If you prefer to ignore or even ban certain opinions instead of arguing them, then you have no place in science. It can be a long hard road sometimes arguing with people constantly about the same topics, but if you are truly right you will be shown as such eventually. Maybe 50 years after you're dead, but still.


Interesting. I wasn't aware a mentioned anything about banning ideas or taking away their right to believe in fantasies (though you and I will disagree with their right to control what they are injected with when they live in a society).

The only point I made was that there is a difference between public opinion popping up in comment sections, forums, and blogs, and the standardized methods we have for explaining reality in incremental steps of precision and accuracy (i.e. science). The general public can question science all they like but usually the only thing it really provides for science in general is a measure of the ability of the public to understand the theories and models that currently form the framework of our understanding. Only when someone goes through the formal methods of putting forward and testing hypotheses (especially at a rigorous level that passes peer review) does the questioning have any bearing on the actual scientific framework.
 
2013-11-09 05:37:57 PM

DerAppie: Mentat: That's a good idea.  We should come up with some kind of method that formalizes that.

devek: If I conducted an experiment that disproved a major theory, I would be made famous.

I don't see why people think science is conspiracy lol.

The problem isn't a conspiracy or that there is no system for checking work that someone has done. It is that the work isn't being checked enough. Sure, the big once in a life time things will attract replication like a dog attracts fleas. But the small things? A lot gets published without ever being replicated. That way a lot of things that are demonstrably false don't get found out about because there is no one demonstrating anything. This one "scientist" merely made up some numbers and the peer review saw that the maths worked out without and the conclusion wasn't to outlandish and cleared it for publication.

Peer review and articles were once imagined to be a replacement for actual demonstrations. To reach the larger scientific community, as only a few people would fit in a laboratory at a time. The article would be read and people would actually check if things were true and write rebuttals. Over time, however, the system changed subtly. "Peer reviewed" became synonymous with "this is true" instead of "someone checked my maths for major errors and maybe someone else should actually replicate my study." Just keep an eye out and you'll find it everywhere. Even on Fark people will ask for "peer reviewed" articles as the final arbitrator of truth. Not reports of replications. Peer reviewed. As if the two people who were nowhere near the experiment (and often don't even work in the direct field) and only read the information provided could actually make an informed conclusion about the validity of the content.

Things need to change, especially now that we have the internet and electronic communication. Every publisher could add a counter to an article to show how often it has been successfully replicated or refuted. They could even add references to the articles. That information would be a lot more useful to see how reliable an article is than the vague "peer reviewed" status that has been elevated in importance to a position that is entirely undesirable.


To extrapolate, most scientist use parts of most papers to design their assays. While the experiment may not be reproduced in toto, enough parts are usually reproduced and most scientists will not take a novel approach as fact.
 
2013-11-09 05:38:49 PM

J. Frank Parnell: verbal_jizm: Yes, the brainstrust in the Yahoo News comments loudly lambasting scienticians, calling for creationism in schools and mandatory burning of fossil fuels (to show them damn dirty hippies), and refusing to let their snowflake's precious bodily fluids be contaminated with autism juice, all because they heard a fourth hand account of a scientists work reported by a "journalist" with dubious credentials is the same as peer review and scientific consensus.

There's no harm in believing creationism, and people have the right to decide what they're injected with. Deal with it.

Reason will always prevail in the end, and science has always been about arguing with everyone, not agreeing with everyone. If you prefer to ignore or even ban certain opinions instead of arguing them, then you have no place in science. It can be a long hard road sometimes arguing with people constantly about the same topics, but if you are truly right you will be shown as such eventually. Maybe 50 years after you're dead, but still.


How can you be so obtuse? Is it deliberate?
 
2013-11-09 05:50:14 PM

malaktaus: How can you be so obtuse? Is it deliberate?


Solitary! A month!
 
2013-11-09 05:51:37 PM
I hate science but don't question it.
 
2013-11-09 05:52:27 PM

J. Frank Parnell: DerAppie: Over time, however, the system changed subtly. "Peer reviewed" became synonymous with "this is true" instead of "someone checked my maths for major errors and maybe someone else should actually replicate my study." Just keep an eye out and you'll find it everywhere. Even on Fark people will ask for "peer reviewed" articles as the final arbitrator of truth. Not reports of replications. Peer reviewed. As if the two people who were nowhere near the experiment (and often don't even work in the direct field) and only read the information provided could actually make an informed conclusion about the validity of the content.

It's intellectual laziness. Just tell me what to think, there are TV shows to watch.

With the wealth of information online, even full university courses, anyone can have a decently educated opinion on even the most sciency of topics. We all don't need to look to some 'expert' or some agency to tell us what to think.



Keep in mind that while one can have a decently educated opinion without experts, it is also extremely unlikely that one would be able to meet the same level of knowledge and experience as an expert. What you're missing is the years and years of learning about and working with the subject matter, not just the simple  availability of information.

Also keep in mind that most people don't have the time nor energy to learn everything to a level equivalent to that of an expert. Expert opinion is a very useful heuristic in the face of the large amount and complexity of information out there, as well as the frequent necessity of making decisions in a timely manner.
 
2013-11-09 05:59:08 PM
Unless it is about climate change, in which case even sounding something similar to questioning gets you labeled immediately as a denier on par with a flat-earther.

The good news is that the science there is SETTLED. It is so settled, in fact, that Al Gore created a website whose sole purpose is to spam public forums with their canned talking points so that it becomes even settleder.
 
2013-11-09 06:02:15 PM
Once a physical process becomes a well established mechanism to explain natural phenomena--the marrying of the understanding of the natural world with physics and/or chemistry--that theory is very difficult to overturn.

Evolution: the discovery that the phenotype of a population of organisms could change via the natural selection of "winning" physical traits from generation to generation was a good start. The discovery of genes, by which these traits could be handed from generation to generation, was even better. Now that we're down to understanding DNA at the molecular level and how things like mutations and interactions with microRNA are expressed at the organism level, evolution is one theory that's never going to get overturned.

Plate tectonics: the hypothesis of continental drift was a good start. The discovery of subduction zones and mid-oceanic ridges gave it some big oomph, though the drivers behind that weren't well known. The advancement of geophysics and how we now understand how mantle flows and  continental crust interacts with it, as driven by densities, has made plate tectonics another theory that'll never get overturned.

The greenhouse effect and anthropogenic global warming: discovering over a century ago that some gases absorb IR light differently than others made for an interesting hypothesis that the earth could warm if CO2 accumulated in the atmosphere. The use of paleoclimatology to help track historical changes in comparison to changes in the atmosphere and solar input has been a big help in understanding what's happening, but, by itself, didn't provide a physical model beyond some real basic stuff that could be inferred from it. However, models built on physics of heat exchange--those physical processes that try to understand the interaction of the atmosphere, ocean, and land, along with solar input--have gone a huge way in helping scientists understand what's happening, especially when tested against paleoclimatological and historical evidence. The current state of climatology is very unlikely to be overturned now that the physical premises behind it are fairly well known.

Like I said, once you nail those underlying physical mechanisms, it gets very difficult to overturn something.

/Kuhn's postmodernistic stuff about the advancement of science, that theories are very tightly held onto by scientists until some sort of revolution where one theory overturns another, is very useful in understanding the initial battles between theories when physical/chemical processes aren't very well know, but definitely has its limits
 
2013-11-09 06:04:04 PM

machodonkeywrestler: To extrapolate, most scientist use parts of most papers to design their assays. While the experiment may not be reproduced in toto, enough parts are usually reproduced and most scientists will not take a novel approach as fact.


machodonkeywrestler: Fark. Meant to quote DerApple.


I'll answer both of you at once, if you don't mind. And if you do.

I never said that there was no method for checking. Just that the method doesn't get used enough. Peer review isn't a check to see if the conclusion of the paper is valid. Just a check to see if the maths work out and if the conclusion isn't too outlandish. Which is why, not too long ago, a Dutch social psychologist got mixed up in a bit of a fraud investigation. It turns out that 55 of the 130 articles and 10 out of 24 book chapters he got published were fraudulent. The level of fraud varied from totally made up research projects to cherry picked data. And the kicker? All that stuff was peer reviewed. And yet the falsehoods weren't caught. He got referenced by other researchers, his theories were taught and life went on. Except that he got caught and it suddenly became virtually impossible to undo the damage done to the knowledge base which a lot of people were taught and trusted.

That a lot of people used bits of his research is no guarantee that the entire paper was correct. It doesn't even sufficiently guarantee that the fraud would be spotted if the theory based (partly) on the fraudulent work didn't pan out. There is a paper saying things are true and the next researcher added variables to it. Suddenly there is noise in where the blame is. And there are very view who'd say "well, I quoted peer reviewed articles and that the analysis showed that my theory is bull. Lets put the blame on the already published articles instead of on the variables added and the methodological differences of my study." A case in point, a lot of his post docs based their research on his work. The important work being "based on." As in not working on the same theory and thus still able to find valid results. The fact that some did and some did not find significant results doesn't prove anything any of the many short blurbs they quoted for their theoretical basis.

Peer review is bull crap when trying to decide if the paper and its conclusions are valid. Only actual replication can show that.

/Whe whe social psychology isn't science
//Still a valid example to show that peer review doesn't work anywhere near expected levels
 
2013-11-09 06:07:46 PM
Bah, maybe I should re-read do some editing before I hit Add Comment after a long post.
 
2013-11-09 06:08:29 PM
The comments there made some of my brain cells commit vicarious suicide. There were some seriously stupid arguments, their eloquence is irrelevant, they were still stupid arguments.

"Scientists and their acolytes are a priest class" was one of those comments. It's a pretty stupid comment. No one's worshipping science, they're only turning to it for answers because using the scientific method is the best way to try to figure out how things actually work. If it weren't for the scientific method, we'd probably still be bleeding people when they were sick, we'd think things like the woman is responsible for gender of child born to her, that any manner of illnesses from physical to mental were varying problems from curses to demonic possession.

Apparently they do they not understand that science seeks to understand all things with investigation, while religion typically gives you the answers and expects to accept them without investigation. Yes, sometimes scientists are wrong and other theories are developed which are better repeatable, making for a better theory. That doesn't invalidate the scientific process, the errors actually help us gain knowledge because we can learn from faulty theories. We can learn that the old theory is an example of how things don't work, and continue working towards learning how things actually do work.

I would bet such "thinkers" that believe in a "scientific priest class" are afraid of knowledge, inquiry and examination. Perhaps they're aware somewhere in their minds, though they refuse to confront it, that their non-scientific methodology is not repeatable and can't be validated.

These seem to be the types that believe in Intelligent Design (closet creationism) because they have a lack of understanding of high school biology and think that there's irreducible complexity standing in the way of evolution. If you cut out a complex part of a human, say their eye, that doesn't stop them from being human. The eye itself  is actually an astounding example of evolution. It started out as simple photosensitive cells which evolved into a complex multilayered organ. Taking away the cornea doesn't stop if from being a complex organ any more than taking out a cell nucleus changes a cell into something other than a cell lacking a nucleus. Just because a person doesn't understand something doesn't mean it's so complex it can't be understood, it just means you haven't figured out how it works yet.
 
2013-11-09 06:09:58 PM

Mentat: That's a good idea.  We should come up with some kind of method that formalizes that.



Seriously. When anyone utters a statement similar to subby's headline they have just effectively told you "I haven't a farking idea what 'science' actually is. I ignorantly think it's like religion, but with lab coats and bunsen burners."
 
2013-11-09 06:10:13 PM

DerAppie: Mentat: That's a good idea.  We should come up with some kind of method that formalizes that.

devek: If I conducted an experiment that disproved a major theory, I would be made famous.

I don't see why people think science is conspiracy lol.

The problem isn't a conspiracy or that there is no system for checking work that someone has done. It is that the work isn't being checked enough. Sure, the big once in a life time things will attract replication like a dog attracts fleas. But the small things? A lot gets published without ever being replicated. That way a lot of things that are demonstrably false don't get found out about because there is no one demonstrating anything. This one "scientist" merely made up some numbers and the peer review saw that the maths worked out without and the conclusion wasn't to outlandish and cleared it for publication.

Peer review and articles were once imagined to be a replacement for actual demonstrations. To reach the larger scientific community, as only a few people would fit in a laboratory at a time. The article would be read and people would actually check if things were true and write rebuttals. Over time, however, the system changed subtly. "Peer reviewed" became synonymous with "this is true" instead of "someone checked my maths for major errors and maybe someone else should actually replicate my study." Just keep an eye out and you'll find it everywhere. Even on Fark people will ask for "peer reviewed" articles as the final arbitrator of truth. Not reports of replications. Peer reviewed. As if the two people who were nowhere near the experiment (and often don't even work in the direct field) and only read the information provided could actually make an informed conclusion about the validity of the content.

Things need to change, especially now that we have the internet and electronic communication. Every publisher could add a counter to an article to show how often it has been successfully replicated or refuted. They cou ...



If this sort of thing interests you, you might want to expand your discussion to include the differences between concepts of replicability and  reproducibility versus  corroborationand triangulation. What you're talking about is the former. However, from my experience, what is more common is the latter, in which the findings of a paper are explored from a slightly different angle. Sometimes it's with different methods, a different system or subject, an expanded scope, or other difference. This approach also provides confirmation of a paper's results, as well as providing robustness under different conditions that replication alone cannot provide.

Also note that reproducibility is also made very difficult when working with natural experiments or when working with systems that exhibit variability over time - you can't replicate a particular volcanic eruption, for example. What should also be mentioned is the fact that our current publication and funding paradigm prioritizes  novel results.
 
2013-11-09 06:10:43 PM

Felgraf: doglover: devek: There is harm believing in creationism because to do so you have to reject science entirely.

No you don't.

God created evolution. Boom, done, moving on.

When people hear or talk about creationism, they usually mean *young earth* creationism.


Well then, that's "people"'s failure. When I say something, I means it.

Personally, I like young earth creationists because not even the Church would agree with them. The monk Robert Bacon was basically Newton and Galileo rolled into one. St. Augustine would send them a very sternly worded letter about logic. Even the Templars didn't give a god's damn about such nonsense as long as they could fight Muslims.
 
2013-11-09 06:11:55 PM
Well, pretty much by definition, a 95% confidence interval means that one study in 20 will find a statistically significant result purely by random chance.  That's not even taking errors in methodology into account - perfect methods would still give an incorrect finding in those 5% of cases.  Then there's the whole asymmetry between accepting or rejecting the null hypothesis - a study is more likely to incorrectly accept the null hypothesis than to incorrectly reject it.

That's why important studies tend to be reproduced many times before they are generally accepted.  I don't know anyone outside of the popular press who will believe something just because a single study reported a possible link.  Even the authors don't generally go so far as to draw broad conclusions from a single study.
 
2013-11-09 06:15:23 PM

Damnhippyfreak: Also note that reproducibility is also made very difficult when working with natural experiments or when working with systems that exhibit variability over time - you can't replicate a particular volcanic eruption, for example. What should also be mentioned is the fact that our current publication and funding paradigm prioritizes  novel results.


Not to mention the much more obvious fact that years if not decades go into many publications - reproducing such is often impractical, with little benefit to the those doing the reproducing given the investment in time and effort.
 
2013-11-09 06:16:57 PM

Sum Dum Gai: Well, pretty much by definition, a 95% confidence interval means that one study in 20 will find a statistically significant result purely by random chance.  That's not even taking errors in methodology into account - perfect methods would still give an incorrect finding in those 5% of cases.  Then there's the whole asymmetry between accepting or rejecting the null hypothesis - a study is more likely to incorrectly accept the null hypothesis than to incorrectly reject it.

That's why important studies tend to be reproduced many times before they are generally accepted.  I don't know anyone outside of the popular press who will believe something just because a single study reported a possible link.  Even the authors don't generally go so far as to draw broad conclusions from a single study.



Good point. The analogy I like to use when talking about this is that we don't stand on the shoulders of giants so much as stand on the shoulders of many, many midgets.
 
2013-11-09 06:21:52 PM
There's a problem with the headline. It makes regular folks think any random half assed criticism is on the same level as peer review.

Yes, science needs criticism to thrive, but it's educated criticism, not your back of napkin debunking of general relativity, that it needs.
 
2013-11-09 06:39:11 PM

J. Frank Parnell: DerAppie: Over time, however, the system changed subtly. "Peer reviewed" became synonymous with "this is true" instead of "someone checked my maths for major errors and maybe someone else should actually replicate my study." Just keep an eye out and you'll find it everywhere. Even on Fark people will ask for "peer reviewed" articles as the final arbitrator of truth. Not reports of replications. Peer reviewed. As if the two people who were nowhere near the experiment (and often don't even work in the direct field) and only read the information provided could actually make an informed conclusion about the validity of the content.

It's intellectual laziness. Just tell me what to think, there are TV shows to watch.

With the wealth of information online, even full university courses, anyone can have a decently educated opinion on even the most sciency of topics. We all don't need to look to some 'expert' or some agency to tell us what to think.


How does an average person discriminate between good information and bad information they find on the internet?  Are average people capable of synthesizing information from many different sources and reaching valid conclusions when they lack the education and training often necessary to do so?

And let's not pretend checking a few more sources online other than Nature (or some other reputable science source) means you're not intellectually lazy.  Did you personally run dozens of carefully calibrated experiments or rigorously collect data in some fashion?  No?  Then you're trusting that the source you're gathering this wealth of information did so, right?  But then we're right back to where we started only we're apparently not (only) trusting experts.  What criteria do these "non expert" sources need to meet for you to consider them?  Why is your personal set of criteria an improvement on the process that experts have established?
 
2013-11-09 06:39:13 PM

SevenizGud: Unless it is about climate change, in which case even sounding something similar to questioning gets you labeled immediately as a denier on par with a flat-earther.


Nah, it's just that the questioning is based upon bullsh*t. That'll happen when you don't understand science.
 
2013-11-09 06:39:48 PM

Baryogenesis: How does an average person discriminate between good information and bad information they find on the internet?


They don't. That's why people pass around bogus stories all the time. Heck, it's the entire reason that Snopes exists.
 
2013-11-09 06:43:57 PM

Damnhippyfreak: If this sort of thing interests you, you might want to expand your discussion to include the differences between concepts of replicability and reproducibility versus corroboration and triangulation. What you're talking about is the former. However, from my experience, what is more common is the latter, in which the findings of a paper are explored from a slightly different angle. Sometimes it's with different methods, a different system or subject, an expanded scope, or other difference. This approach also provides confirmation of a paper's results, as well as providing robustness under different conditions that replication alone cannot provide.


That slightly different angle is why a lot of fraud (and errors) isn't found until it is too late. As I said in a later post, adding variables or slightly changing the method introduces noise. It might happen that just enough has been changed to create a corroborating account without actually having checked if the original theory is true. Which is why I am of the opinion that research need to be reproduced where possible. To get back to the social scientist: 55 fraudulent papers over a number of years would have been found a lot earlier if a lot of people didn't merely try and corroborate his results.

Which brings us to this:

Also note that reproducibility is also made very difficult when working with natural experiments or when working with systems that exhibit variability over time - you can't replicate a particular volcanic eruption, for example.

It is true that sometimes reproduction simply can't be done, as with the volcanic eruption. All you can do in those cases is pour over the data, verify all claims and hope that the data is correct. But that still makes it a mere peer reviewed paper. At best it can be used as a guide for analysing another eruption or even for preparing counter points based on the data. Other eruptions would be needed to corroborate or disproof the findings. Merely reviewing the paper along with the provided data isn't exactly proof of anything.

As a rule of thumb it can be said that the harder the science, the easier it is to reproduce a result. For the most part the trouble would be getting the required equipment. In the softer sciences reproducing might be an issue because of sample differences, but working in similar situations by the same method should still provide a clearer view of the validity of a claim than when there is a slightly different sample and something in the method has been changed.

What should also be mentioned is the fact that our current publication and funding paradigm prioritizes novel results.

Which is also why peer review is a crappy measurement of validity. The more novel an approach and the results, the less the reviewers will know about it. And the aim for novel results causes a neglect in the checking of previous works. Which is, once again, where fraud and errors creep in.

Anyway, I think this is drifting away from the main point of my statement: A peer reviewed article is not the arbiter of truth. The common misconception that peer review somehow elevates the content to truth is detrimental to science as a whole and a disservice to the purpose peer reviews were even started for.

How that should be fixed is open for debate, but I'll need some very good counter arguments to convince me that those 2 lines of text aren't true.
 
2013-11-09 06:58:53 PM
I would be cool with a scientific priest class. The idea of buying a certified fragment of Madame Curie's bones to get out of remedial algebra is very appealing to me.
 
2013-11-09 07:02:23 PM

DerAppie: Which is why I am of the opinion that research need to be reproduced where possible.


And that's why PLoS One and others have introduced The Reproducibility Initiative.
 
2013-11-09 07:12:07 PM

verbal_jizm: The general public can question science all they like but usually the only thing it really provides for science in general is a measure of the ability of the public to understand the theories and models that currently form the framework of our understanding. Only when someone goes through the formal methods of putting forward and testing hypotheses (especially at a rigorous level that passes peer review) does the questioning have any bearing on the actual scientific framework.


But my point was more that it isn't just science and the stupid people in the public who disagree. Within science itself there are skeptics to anything, or at least there should be if things are working properly. The idea all scientists frequently agree on things is a media fantasy. And it's like they want to make it that way now, by trying to silence or bury any dissent in scientific communities, when that is essential to science.

Here's a good article about that which i posted before, but it's even more fitting here.
 
2013-11-09 07:12:26 PM
SevenizGud: Unless it is about climate change, in which case even sounding something similar to questioning gets you labeled immediately as a denier on par with a flat-earther.

Unless it involves Elon Musk, in which case, the founder of PayPal can't be wrong. EVER!

/Yeah, he more engineering than science but the abandonment of skepticism and adoption of heroism in his case a bit disturbing
 
2013-11-09 07:14:08 PM

devek: If I conducted an experiment that disproved a major theory, I would be made famous.


Or maybe you would just be persecuted as a heretic by your colleagues and even your field's most notable Nobel prize winner.

An Israeli scientist who suffered years of ridicule and even lost a research post for claiming to have found an entirely new class of solid material was awarded the Nobel Prize for chemistry on Wednesday for his discovery of quasicrystals.

Three decades after Daniel Shechtman looked with an electron microscope at a metal alloy and saw a pattern familiar in Islamic art but then unknown at a molecular level, those non-stick, rust-free, heat-resistant quasicrystals are finding their way into tools from LEDs to engines and frying pans.

Shechtman, 70, from Israel's Technion institute in Haifa, was working in the United States in 1982 when he observed atoms in a crystal he had made form a five-sided pattern that did not repeat itself, defying received wisdom that they must create repetitious patterns, like triangles, squares or hexagons.

"People just laughed at me," Shechtman recalled in an interview this year with Israeli newspaper Haaretz, noting how Linus Pauling, a colossus of science and double Nobel laureate, mounted a frightening "crusade" against him, saying: "There is no such thing as quasicrystals, only quasi-scientists."

After telling Shechtman to go back and read the textbook, the head of his research group asked him to leave for "bringing disgrace" on the team. "I felt rejected," Shechtman remembered.

"Daniel Shechtman had to fight a fierce battle against established science ... His battle eventually forced scientists to reconsider their conception of the very nature of matter.


This story is only two years old
 
2013-11-09 07:17:26 PM

SevenizGud: Unless it is about climate change, in which case even sounding something similar to questioning gets you labeled immediately as a denier on par with a flat-earther.

The good news is that the science there is SETTLED. It is so settled, in fact, that Al Gore created a website whose sole purpose is to spam public forums with their canned talking points so that it becomes even settleder.


Scientific consensus? How does it work? Also, the problem with climate deniers is that they have no facts. That's pretty important to science. Oh, and people who aren't climate scientists generally aren't the best people to discredit climate change. Climate change is probably one of the most peer-reviewed scientific studies ever. If not the most.

Mainly because there is so much data, from so many different sources. All of it says yes the climate is changing in a way like never before and that the change started with the industrial revolution. IE: Humans.
 
Displayed 50 of 93 comments

First | « | 1 | 2 | » | Last | Show all

View Voting Results: Smartest and Funniest


This thread is closed to new comments.

Continue Farking
Submit a Link »






Report