Critical thinking means coming up with a coherent set of thoughts, supported by some amount of evidence, that is more likely to reveal the truth than pure guessing.
This is particularly true in fields like nutrition, where the funding available is low relative to the complexity of the problem being studied.
Most studies in the field of nutrition fall into the "inconclusive" category.
The more complex a system you are talking about, the more apparently contrasting evidence is bound to be found.
Now you're talking about this kind of thing. Fortunately, if someone does yet another study to determine whether coffee causes cancer, they summarize previous research briefly, in order to explain why their study is special.
A lot of times the problem is the media misrepresenting a study in order to be sensationalistic, and if you read the actual paper instead of the reports, it's more tame and accurate. (and sometimes scientists are complicit in it, by the title of their paper something surprising or by giving interviews to the media that are not entirely supported, or they are misquoted.)
I don't think it's that bad. Most papers follow a format kind of like this:
1) Assuming X (which relies on other studies)
2) We tested Y
3) And conclude Z
Usually the parts you are interested in are Y and Z. So the first thing to look for is the methodology. Is it an experiment you could reproduce and get the exact same results (assuming you had funding, of course)? In other words, did they describe it in enough detail to be reproducible?
The next thing to look for is error bars. Have the calculated the margin of error, and does the calculation make sense?
The next thing to look for is the effect size. A lot of studies baaaarely show any significant effect at all, and that's a good sign that the paper isn't very good.
If you get past those things, there's a good chance the study would be able to be reproduced if someone tried. Of course, that's the next step: reproducing the study if you want to be sure.
COBOL was created in collaboration between Americans and Europeans, and it nearly broke down over the number seperator, with one researcher emotionally declaring, "I will never use a period as a decimal point!" Eventually they came to a compromise but not before a tombstone was made for COBOL. https://www.computerhistory.or... Next let's tackle the controversy of order of operations! Left to right is of course the proper order.
I really wonder how these things end up online, given that most consumer routers don't accept incoming connections by default. Are people really going out of their way to put this stuff on the open internet, or is something else going on here?
There's a lot you don't have any choice about when it comes to raising your kids. "Shouldn't I be able to let them play in the front yard alone?" It might seem reasonable but it also might get CPS called on you.
I don't even think you are right. Systemic reviews tend to be written by the same crappy scientists who wrote the original crappy studies. In my experience they are poorly written as often as not. Ultimately if you don't understand statistics and can't calculate the margin of error on a study, then you are hosed and there is nothing you can do.
If he spends a lot of time in R&D, that explains why nothing impressive has come out of Microsoft research, despite spending tons of money and hiring good people. He spends time working on Microsoft AI? The most notable thing to come out of Microsoft AI was Tay.
Would argue that writing was more important than the printing press by an order of magnitude. The world went from a technological advancement every few thousand years to several a century.
You should go on systemic reviews published in high impact factor journals.
Systemic reviews aren't great, either. If it's a topic you care about, you should read the study yourself and evaluate it. Reproduce it if it's important enough, but with a knowledge of statistics you should be able to filter out most of the problematic ones.
You absolutely should be allowed to modify the Autosteer software on a Tesla. Whether or not you should be allowed on the road with those modifications is another question, just like if you make physical modifications to your car.
No one knows how to fix a psychopath, don't kid yourself. Psychiatrists are like doctors from the 1800s who had very little understanding of how it worked (but they knew some things that didn't work, like mesmerism). At least we don't do lobotomies anymore, yuck.
Slashdot Mirror
Do you call two-dimensional squares "flat, one-dimensional lines"?
I call them flat, two-dimensional squares.
You just like to argue.
Critical thinking means coming up with a coherent set of thoughts, supported by some amount of evidence, that is more likely to reveal the truth than pure guessing.
Well said.
This is particularly true in fields like nutrition, where the funding available is low relative to the complexity of the problem being studied.
Most studies in the field of nutrition fall into the "inconclusive" category.
The more complex a system you are talking about, the more apparently contrasting evidence is bound to be found.
Now you're talking about this kind of thing. Fortunately, if someone does yet another study to determine whether coffee causes cancer, they summarize previous research briefly, in order to explain why their study is special.
A lot of times the problem is the media misrepresenting a study in order to be sensationalistic, and if you read the actual paper instead of the reports, it's more tame and accurate. (and sometimes scientists are complicit in it, by the title of their paper something surprising or by giving interviews to the media that are not entirely supported, or they are misquoted.)
Yeah. Selling core memory was not a sustainable business model, but it sure made a lot of people a lot of money for a while.
I don't think it's that bad. Most papers follow a format kind of like this:
1) Assuming X (which relies on other studies)
2) We tested Y
3) And conclude Z
Usually the parts you are interested in are Y and Z. So the first thing to look for is the methodology. Is it an experiment you could reproduce and get the exact same results (assuming you had funding, of course)? In other words, did they describe it in enough detail to be reproducible?
The next thing to look for is error bars. Have the calculated the margin of error, and does the calculation make sense?
The next thing to look for is the effect size. A lot of studies baaaarely show any significant effect at all, and that's a good sign that the paper isn't very good.
If you get past those things, there's a good chance the study would be able to be reproduced if someone tried. Of course, that's the next step: reproducing the study if you want to be sure.
I don't understand people who can't respond to email. What kind of emails are they getting that are so hard?
COBOL was created in collaboration between Americans and Europeans, and it nearly broke down over the number seperator, with one researcher emotionally declaring, "I will never use a period as a decimal point!" Eventually they came to a compromise but not before a tombstone was made for COBOL. https://www.computerhistory.or... Next let's tackle the controversy of order of operations! Left to right is of course the proper order.
I really wonder how these things end up online, given that most consumer routers don't accept incoming connections by default. Are people really going out of their way to put this stuff on the open internet, or is something else going on here?
There's a lot you don't have any choice about when it comes to raising your kids. "Shouldn't I be able to let them play in the front yard alone?" It might seem reasonable but it also might get CPS called on you.
Yeah that's true
I don't even think you are right. Systemic reviews tend to be written by the same crappy scientists who wrote the original crappy studies. In my experience they are poorly written as often as not. Ultimately if you don't understand statistics and can't calculate the margin of error on a study, then you are hosed and there is nothing you can do.
If I were a teacher, I would go crazy if I had to grade video homework every day.
If he spends a lot of time in R&D, that explains why nothing impressive has come out of Microsoft research, despite spending tons of money and hiring good people. He spends time working on Microsoft AI? The most notable thing to come out of Microsoft AI was Tay.
Would argue that writing was more important than the printing press by an order of magnitude. The world went from a technological advancement every few thousand years to several a century.
Yeah, it's overstating it to say that India wants American companies. They want American CEOs to come stand trial, at least one of them....
Just use common sense
wtf is common sense?
You should go on systemic reviews published in high impact factor journals.
Systemic reviews aren't great, either. If it's a topic you care about, you should read the study yourself and evaluate it. Reproduce it if it's important enough, but with a knowledge of statistics you should be able to filter out most of the problematic ones.
A lot of sane people are walking away from facebook
Yeah, right to Instagram. Mark Zuckerburg is so sad about that.
It is hard to imagine, impossible actually, this conversation happening 50 years ago.
Anti-vaxxers have been around almost as long as vaccines
I haven't died yet!
That particular AC was referring to a certain poster that is not liked here for some reason. So kind of a woosh.
Point me to where the constitution says anything about what car's I can drive
second amendment bro, now stand in front of my headlights, I'm going to try something.
You absolutely should be allowed to modify the Autosteer software on a Tesla. Whether or not you should be allowed on the road with those modifications is another question, just like if you make physical modifications to your car.
No one knows how to fix a psychopath, don't kid yourself. Psychiatrists are like doctors from the 1800s who had very little understanding of how it worked (but they knew some things that didn't work, like mesmerism). At least we don't do lobotomies anymore, yuck.