The way it should be is that the metrics for performance are the aggregate quality and impact of the work, not the number of publications or the impact factors of the journals they go into. Why doesn't this work? Because administrators generally don't understand the science that they are "administering."
That isn't why it doesn't work. It doesn't work because there's no particularly good objective metric for "quality" or "impact." For "impact" you have number of citations and where the work was published. If you want to get fancy, you can make a metric that takes into account what those citations were. Quality is pretty much impossible to judge objectively, particularly if you want to compare across fields. It doesn't matter how competent or knowledgeable your assessors are--that's not the limiting factor. In fact, what are you are calling for is already being done: when grant applications are judged, it's done by competent people who are knowledgeable in the field. They judge based upon the quality of the application, its novelty, its practicality, presence of pilot data, and, of course, if you have managed to publish this stuff before and what the publications looked like. The system isn't perfect, but I can't see how it can easily be changed to make it radically better.
Saying "Publish or perish must go" is great, we all like the sound of that. But then what do you replace it with? Any metric that you come up with will be gamed if the people being measured know how you're measuring them. It's easy to point the finger at journals, funding committees, and hiring committees and say that the publish or perish mentality is their fault. But it's also the fault of researchers who choose to play the game. Researchers choose to break down papers into many smaller ones in order to increase publication count. Researchers choose to waste everyone's time by gambling and submitting to progressively lower tier journals until the paper sticks, rather than being honest with themselves and pitching the manuscript correctly from the start. Researchers choose to publish the shit stuff they barely believe anyway, wherever it'll get in, rather than consign it to the scrap heap and start over.
Who read about the University of Edinburgh physicist: He just won the Nobel prize, and has published a total of 10 papers in his entire career. As he said: today he wouldn't even get a job.
Obviously it's easy for someone in his position to take this stance (it would be suicide for most early career scientists), but it's still laudable. I've seen instances of this go the other way: Nobel appears and the person turns it into a licence to publish craptacular papers in top tier journals. When this happens it's bad on every level: harms the field, harms the first author, harms the journal.
If you don't know the difference between "breaks" and "brakes", will you really understand the answers to your questions?
Why not? Typos, missed auto-corrects, or brain-farts aren't a reflection of one's intelligence. I'm sure the the senator knows how a car stops, despite his spelling mistake.
What's to stop someone from forcefully taking down an Amazon drone, then placing it into a Faraday cage while they disassemble it and get the free hardware?
The fact that it's vapourware and will never see active service?
This is really a VR headset for watching movies, playing games, etc. It's not designed for the same applications Google Glass is designed for. If anything, this is a competitor to devices such as Oculus Rift, not Google Glass.
How safe is the car when you follow all driving laws like speed limits especially through turns?
I'm from England and currently live in NY state. I don't even think we have speed limits on turns in the UK (or if we do they're rare). Here in NY, loads of turns have speed limits on them and they're almost all unrealistically low or pointless. I have a large wagon and I can easily take these turns safely at twice the posted limit (e.g. doing 30 MPH through a 15 MPH turn). Many of the turns with posted speed limits are trivial, gradual, turns on roads that aren't particularly busy. In general US roads have an over-abundance of restrictions (double yellows everywhere, over-use of stop signs, etc). The authorities seem to think that more restrictions will lead to safer driving and lower fatality counts. My suspicion is that it just leads to "road sign inflation" with drivers either ignoring signs (because many are meaningless) or, possibly worse, no longer thinking for themselves what is the correct speed, etc, for a particular stretch of road.
Don't forget that intelligence is hard to define and test. The IQ test probes some correlates of intelligence, but it can be gamed and you can train for it (which is another reason to be cautious about the Flynn effect--conventional education effectively "trains" people for IQ tests and nowadays more people spend more time in education. The Flynn effect is tailing off in many 1st world countries, which is consistent with this explanation.). e.g. Digit span (forward and backward) is tested in an IQ test. Without training, most people have a hard time reaching ten digits. However, with training you can recall 100 or more digits. You haven't become smarter, you've just trained once particular thing. Ditto with other aspects of the test. This is why those "brain training" games are pseudo-scientific bollocks. They make you better at the game, they don't make you smarter. It's possible that regularly "using" your brain will stave off dementia, and experience in life counts for a lot, but nobody has shown that you become "smarter" through training.
Not at all. Lamarck's ideas involved concepts such as inheritance through use. e.g. animals that stretch to get higher leaves of trees will stretch their necks and these longer-necked animals pass on that trait to their children. This isn't what happens and it isn't what epigenetics does. Epigenetics add another layer to how natural selection works and might even accelerate natural selection, but it doesn't directly lead to changes in the genetic code and its effects can washed out over a few generations.
I can't tell if you're trying be funny, but Lamarkian evolution isn't the same thing as epigenetics and the existence of epigenetics is not a problem for the theory of evolution by natural selection.
I don't feel making the patient shop around for the best deal in medical care is, at this time, practical. Next time you use your medical insurance ask the doctor exactly how much each test and procedure costs and how much the insurance company will pay.
Absolutely: I think one reason costs are so high is because patients do not shop about at all. There is no shopping about because a) patients mostly don't know on what basis to compare services. b) They generally go wherever is closest. c) If there's something seriously wrong with you, you don't want to go looking for the cheapest treatment. d) In many cases, the prices only appear after you've been treated and get the bill. You usually don't know in advance who will be cheapest.
Where people can and do shop about, however, is insurance.
If I only have a small pool of money to pay tenured professors, why wouldn't I want to select the ones that have proven themselves?
That's a reasonable stance, and one the funding agencies and journals more or less follow, but it it's somewhat blinkered. Some researchers produce good work early in their career but then become stagnant or a little nuts in the second half of their career. However, if they have built up a sufficiently big name they are able to attract more funding and get papers into better journals than their contemporary output really deserves. The money would have been better spent on budding new minds.
Anecdote: There's a Nobel laureate in my field who routinely gets spectacularly shit papers into the very top journals. These papers take up space that should have gone to deserving work, they don't help the lead author (student or post-doc) because everyone knows the paper is over-sold shit, and they are bad for our field because people outside judge us by them. Furthermore, this person and progeny from the lab tend to stick together and block outsider's work from getting into these journals. The reviews that come back are obviously obstructionist bile, but the editors don't have the balls to do deal with it. I had a paper blocked this way: the second reviewer even wrote to the journal to call out the first, obstructionist, reviewer but it did no good. A shittier version of our work, authored by people more established than us (but not those who blocked us), eventually appeared in this top journal. I know others who've had worse experiences. It's terribly demoralising.
Another important reason is that, however unlikely it is to have your special numbers come up, it is not impossible. When they do come up and just that time you didn't play, you'll kick your own arse for the rest of your life. This risk is a strong motivation to keep playing, which can lead to gambling addiction. So to lower your risk of getting addicted to gambling, don't ever bet on the same numbers.
The grandmother of one of my sister's friends would "play" the lottery that way. Choosing numbers but not buying a ticket. I don't recall if she stuck with one set of numbers or chose new ones each week. Anyway: one week she "won." That was a bit of bugger.
Then have the consumer pay the doctor directly. If you want to buy insurance, that's fine - but it's the consumer who submits bills to the insurance company, and the insurance company pays the consumer. Make the consumer get involved in the financial side of their own healthcare, because right now the consumer really doesn't know how much anyone's paying, nor now much is being charged.
I agree: stripping things down would likely help. One good example is the Swiss system. Wages are high and taxes on par with the US (http://upload.wikimedia.org/wikipedia/commons/3/3d/Tax-Revenues-As-GDP-Percentage-%2875-05%29.JPG). From these wages the individual must take out their own health insurance. Some workers, such as those in education, can get discounted insurance. Your premium never costs more than 8% of your income because the government will chip in to cover costs over 8% should you have a low-paying job. The compulsory insurance plans are not allowed to discriminate based on pre-existing conditions, etc. The insured always pays a small proportion of the costs (which is capped, so you won't lose your house and savings if you get cancer) in order to dissuade over-use of services for trivial reasons, etc. IIRC, the insured always pays the full cost of an ambulance trip. You can also take out extra, private, insurance. These premiums are risk-assessed.
of the record labels. Before records, musicians made money by playing in live concerts. That's what musicians should do today,
I agree, but what do you mean "should do", it's what they do do. Search for "$ARTIST_NAME tour dates" and you'll either find the option to buy tickets or see evidence of recent touring activity.
Fairly predictably, this post has led to comments regarding which language is better for teaching kids. There is merit to that discussion, of course, but the main goal of teaching programming to kids to get them interested and maintain their interest. With regards to this, it matters more what they learn to do with the language than what the language is. The smart, creative, motivated kids will figure this out for themselves; those are the kids that will self-teach. The rest will need to be shown the utility of what they're learning. I remember trying to learning BASIC as a kid: I progressed a little but I never really "got" it. It was only much later, at university, when I had to learn MATLAB to analyse data, that things really clicked. It clicked because I could see how learning this language made my work radically easier and more effective. Since then I've dabbled a bit with Perl, Python, and C++.
Kids understand the web, so perhaps a good approach would be to teach them HTML and CSS so they can make a web page. A mark-up language isn't a bad first step into programming and building a web page is something a lot of kids would care about. Then teach them about dynamically generated pages and they power these provide using any language of your choice. You could even do some simple Arduino programming and get them moving motors, etc, over the internet via their webpage. Just some thoughts...
This is a good idea, but it touches on one of the main problems with the whole US healthcare system: that the costs in general are wildly high and not anchored to reality. e.g. $15 for administering a Tylenol to a hospital patient, $25 for an alcohol cotton swab, etc. Clearly that's not what those items cost, so what are what are we really paying for? Yes, they could add an "uninsured persons supplement" to the bill, but it would get lost in the shit storm of crazy prices already on there.
My opinion is that America isn't ready for nationalised healthcare. Trying to implement it is like using a bandaid to treat cancer. The first step is to disentangle those costs and make them transparent. Clearly there's wastage and corruption somewhere in the healthcare pipeline. That needs to be regulated down to sane levels. Then, after that's been done, we can worry about some form of nationalisation/socialisation. The system will never be perfect, but the current US system has a lot of room for improvement.
Search me. Just get a decent picture taken. Besides, I don't think it's true that photos make people look less attractive than RL (as the GP stated). It can often be opposite too: I've seen lots of photos where someone looks way more attractive than in RL. It doesn't even take Photoshop to get it that way. The challenge is to get a photo taken that is slightly flattering whilst remaining accurate.
Slashdot Mirror
The way it should be is that the metrics for performance are the aggregate quality and impact of the work, not the number of publications or the impact factors of the journals they go into. Why doesn't this work? Because administrators generally don't understand the science that they are "administering."
That isn't why it doesn't work. It doesn't work because there's no particularly good objective metric for "quality" or "impact." For "impact" you have number of citations and where the work was published. If you want to get fancy, you can make a metric that takes into account what those citations were. Quality is pretty much impossible to judge objectively, particularly if you want to compare across fields. It doesn't matter how competent or knowledgeable your assessors are--that's not the limiting factor. In fact, what are you are calling for is already being done: when grant applications are judged, it's done by competent people who are knowledgeable in the field. They judge based upon the quality of the application, its novelty, its practicality, presence of pilot data, and, of course, if you have managed to publish this stuff before and what the publications looked like. The system isn't perfect, but I can't see how it can easily be changed to make it radically better.
Saying "Publish or perish must go" is great, we all like the sound of that. But then what do you replace it with? Any metric that you come up with will be gamed if the people being measured know how you're measuring them. It's easy to point the finger at journals, funding committees, and hiring committees and say that the publish or perish mentality is their fault. But it's also the fault of researchers who choose to play the game. Researchers choose to break down papers into many smaller ones in order to increase publication count. Researchers choose to waste everyone's time by gambling and submitting to progressively lower tier journals until the paper sticks, rather than being honest with themselves and pitching the manuscript correctly from the start. Researchers choose to publish the shit stuff they barely believe anyway, wherever it'll get in, rather than consign it to the scrap heap and start over.
Who read about the University of Edinburgh physicist: He just won the Nobel prize, and has published a total of 10 papers in his entire career. As he said: today he wouldn't even get a job.
You mean Peter Higgs?
Obviously it's easy for someone in his position to take this stance (it would be suicide for most early career scientists), but it's still laudable. I've seen instances of this go the other way: Nobel appears and the person turns it into a licence to publish craptacular papers in top tier journals. When this happens it's bad on every level: harms the field, harms the first author, harms the journal.
The technique was not invented by the matrix creators. Take a look at this music video from 1985. http://www.youtube.com/watch?v=9el2lg2olpE (Accept - Midnight Mover)
I feel dizzy now.
That's lamarck in action not darwin.
?
If you don't know the difference between "breaks" and "brakes", will you really understand the answers to your questions?
Why not? Typos, missed auto-corrects, or brain-farts aren't a reflection of one's intelligence. I'm sure the the senator knows how a car stops, despite his spelling mistake.
What's to stop someone from forcefully taking down an Amazon drone, then placing it into a Faraday cage while they disassemble it and get the free hardware?
The fact that it's vapourware and will never see active service?
This is really a VR headset for watching movies, playing games, etc. It's not designed for the same applications Google Glass is designed for. If anything, this is a competitor to devices such as Oculus Rift, not Google Glass.
Thanks, I didn't know that.
How safe is the car when you follow all driving laws like speed limits especially through turns?
I'm from England and currently live in NY state. I don't even think we have speed limits on turns in the UK (or if we do they're rare). Here in NY, loads of turns have speed limits on them and they're almost all unrealistically low or pointless. I have a large wagon and I can easily take these turns safely at twice the posted limit (e.g. doing 30 MPH through a 15 MPH turn). Many of the turns with posted speed limits are trivial, gradual, turns on roads that aren't particularly busy. In general US roads have an over-abundance of restrictions (double yellows everywhere, over-use of stop signs, etc). The authorities seem to think that more restrictions will lead to safer driving and lower fatality counts. My suspicion is that it just leads to "road sign inflation" with drivers either ignoring signs (because many are meaningless) or, possibly worse, no longer thinking for themselves what is the correct speed, etc, for a particular stretch of road.
I don't know about the 3-D aspect, but the level of detail these guys can get back is crazy: http://www.nature.com/news/stealth-camera-takes-pictures-virtually-in-the-dark-1.14260 Compare the t-shirt text in the first and last images. It's almost like those shitty scenes in CSI where information seems to come from nowhere.
So why are IQ scores getting higher (http://en.wikipedia.org/wiki/Flynn_effect)? The more we use our brain, the smarter our offspring get.
There are plenty of other, less far-fetched, explanations for the Flynn effect. This is "only" a correlation but it brings up some important issues: http://en.wikipedia.org/wiki/File:IQatWoN_GDP_IQ.png
Don't forget that intelligence is hard to define and test. The IQ test probes some correlates of intelligence, but it can be gamed and you can train for it (which is another reason to be cautious about the Flynn effect--conventional education effectively "trains" people for IQ tests and nowadays more people spend more time in education. The Flynn effect is tailing off in many 1st world countries, which is consistent with this explanation.). e.g. Digit span (forward and backward) is tested in an IQ test. Without training, most people have a hard time reaching ten digits. However, with training you can recall 100 or more digits. You haven't become smarter, you've just trained once particular thing. Ditto with other aspects of the test. This is why those "brain training" games are pseudo-scientific bollocks. They make you better at the game, they don't make you smarter. It's possible that regularly "using" your brain will stave off dementia, and experience in life counts for a lot, but nobody has shown that you become "smarter" through training.
BTW, turns out Lamarck got it right.
Not at all. Lamarck's ideas involved concepts such as inheritance through use. e.g. animals that stretch to get higher leaves of trees will stretch their necks and these longer-necked animals pass on that trait to their children. This isn't what happens and it isn't what epigenetics does. Epigenetics add another layer to how natural selection works and might even accelerate natural selection, but it doesn't directly lead to changes in the genetic code and its effects can washed out over a few generations.
I can't tell if you're trying be funny, but Lamarkian evolution isn't the same thing as epigenetics and the existence of epigenetics is not a problem for the theory of evolution by natural selection.
I don't feel making the patient shop around for the best deal in medical care is, at this time, practical. Next time you use your medical insurance ask the doctor exactly how much each test and procedure costs and how much the insurance company will pay.
Absolutely: I think one reason costs are so high is because patients do not shop about at all. There is no shopping about because a) patients mostly don't know on what basis to compare services. b) They generally go wherever is closest. c) If there's something seriously wrong with you, you don't want to go looking for the cheapest treatment. d) In many cases, the prices only appear after you've been treated and get the bill. You usually don't know in advance who will be cheapest.
Where people can and do shop about, however, is insurance.
If I only have a small pool of money to pay tenured professors, why wouldn't I want to select the ones that have proven themselves?
That's a reasonable stance, and one the funding agencies and journals more or less follow, but it it's somewhat blinkered. Some researchers produce good work early in their career but then become stagnant or a little nuts in the second half of their career. However, if they have built up a sufficiently big name they are able to attract more funding and get papers into better journals than their contemporary output really deserves. The money would have been better spent on budding new minds.
Anecdote: There's a Nobel laureate in my field who routinely gets spectacularly shit papers into the very top journals. These papers take up space that should have gone to deserving work, they don't help the lead author (student or post-doc) because everyone knows the paper is over-sold shit, and they are bad for our field because people outside judge us by them. Furthermore, this person and progeny from the lab tend to stick together and block outsider's work from getting into these journals. The reviews that come back are obviously obstructionist bile, but the editors don't have the balls to do deal with it. I had a paper blocked this way: the second reviewer even wrote to the journal to call out the first, obstructionist, reviewer but it did no good. A shittier version of our work, authored by people more established than us (but not those who blocked us), eventually appeared in this top journal. I know others who've had worse experiences. It's terribly demoralising.
No, no, we're not locking up millions in prison camps, that would be fascism
That would be any oppressive totalitarian regime. The ideology is just the window-dressing necessary for getting into power.
Are you serious? there's no such word as alot.
No, but "allot" does exist, which probably fans the confusion.
Another important reason is that, however unlikely it is to have your special numbers come up, it is not impossible. When they do come up and just that time you didn't play, you'll kick your own arse for the rest of your life. This risk is a strong motivation to keep playing, which can lead to gambling addiction. So to lower your risk of getting addicted to gambling, don't ever bet on the same numbers.
The grandmother of one of my sister's friends would "play" the lottery that way. Choosing numbers but not buying a ticket. I don't recall if she stuck with one set of numbers or chose new ones each week. Anyway: one week she "won." That was a bit of bugger.
Then have the consumer pay the doctor directly. If you want to buy insurance, that's fine - but it's the consumer who submits bills to the insurance company, and the insurance company pays the consumer. Make the consumer get involved in the financial side of their own healthcare, because right now the consumer really doesn't know how much anyone's paying, nor now much is being charged.
I agree: stripping things down would likely help. One good example is the Swiss system. Wages are high and taxes on par with the US (http://upload.wikimedia.org/wikipedia/commons/3/3d/Tax-Revenues-As-GDP-Percentage-%2875-05%29.JPG). From these wages the individual must take out their own health insurance. Some workers, such as those in education, can get discounted insurance. Your premium never costs more than 8% of your income because the government will chip in to cover costs over 8% should you have a low-paying job. The compulsory insurance plans are not allowed to discriminate based on pre-existing conditions, etc. The insured always pays a small proportion of the costs (which is capped, so you won't lose your house and savings if you get cancer) in order to dissuade over-use of services for trivial reasons, etc. IIRC, the insured always pays the full cost of an ambulance trip. You can also take out extra, private, insurance. These premiums are risk-assessed.
of the record labels. Before records, musicians made money by playing in live concerts. That's what musicians should do today,
I agree, but what do you mean "should do", it's what they do do. Search for "$ARTIST_NAME tour dates" and you'll either find the option to buy tickets or see evidence of recent touring activity.
Kids understand the web, so perhaps a good approach would be to teach them HTML and CSS so they can make a web page. A mark-up language isn't a bad first step into programming and building a web page is something a lot of kids would care about. Then teach them about dynamically generated pages and they power these provide using any language of your choice. You could even do some simple Arduino programming and get them moving motors, etc, over the internet via their webpage. Just some thoughts...
This is a good idea, but it touches on one of the main problems with the whole US healthcare system: that the costs in general are wildly high and not anchored to reality. e.g. $15 for administering a Tylenol to a hospital patient, $25 for an alcohol cotton swab, etc. Clearly that's not what those items cost, so what are what are we really paying for? Yes, they could add an "uninsured persons supplement" to the bill, but it would get lost in the shit storm of crazy prices already on there.
In case you think I'm pulling numbers out of my ass, here's how the OECD see things: http://madvilletimes.com/wp-content/uploads/2013/07/Health-Care-GDP-OECD-1980-2012.jpg
My opinion is that America isn't ready for nationalised healthcare. Trying to implement it is like using a bandaid to treat cancer. The first step is to disentangle those costs and make them transparent. Clearly there's wastage and corruption somewhere in the healthcare pipeline. That needs to be regulated down to sane levels. Then, after that's been done, we can worry about some form of nationalisation/socialisation. The system will never be perfect, but the current US system has a lot of room for improvement.
Search me. Just get a decent picture taken. Besides, I don't think it's true that photos make people look less attractive than RL (as the GP stated). It can often be opposite too: I've seen lots of photos where someone looks way more attractive than in RL. It doesn't even take Photoshop to get it that way. The challenge is to get a photo taken that is slightly flattering whilst remaining accurate.