Generously suppose that a busy airport has 100 terrorists go through it a year, and 30 million passengers. 85% of terrorists is 85 true positives; 8% of non-terrorists is 0.08*(30,000,000 - 100) = 2,399,992 (!) false positives. That's a total of 2,400,092 positives, so the probability of a failed test correctly identifying a terrorist is 85/2,400,092 or about 0.0035%.
0.0035%. Generously. Maybe there is some use for this test, but screening every passenger like that tells you practically nothing. Just another excuse to feel self-righteous about irritating people.
That's true -- assuming you choose such bad criteria for your profile. Race, religion, etc., as you point out, are easily skirted. Why? Because they have fundamentally got nothing to do with terrorism. Blond, white Irish Catholics would have been a reasonable choice. Now it's dark haired, tan Muslims. In ten years, maybe it'll be Mexican Catholics upset about immigration laws. Maybe Eskimos who want their seals back.
But profiling does work. Here's my profile: Nervous facial cues or body language. Suspicious baggage. Suspicious behavior. Probably there are more criteria that would be effective.
The point is that terrorists are very likely to be nervous, and are by definition up to something suspicious. Yes, it's possible to hide these traits as well. But now a terrorist recruiter needs to find someone who is willing to die for a cause, willing to buy into that particular cause, willing to kill lots of civilians, and is in addition a very convincing actor, with the discipline to not stop acting for one second in the airport. The category of potential undetected terrorists is now relatively miniscule.
These magical, invisible terrorists also have to be in the right (wrong) place at the right time and get recruited. I feel pretty good about my odds of getting blown up under a scheme like that.
Malcolm Gladwell has a good article about this stuff.
Considering the amount posted on this in the last month or so, it's kind of astounding that there exists a slashdot reader who doesn't know this.
Anyway, google ->
This is not a complete list and Sony-BMG continues to refuse to make such a list available to consumers. Consumers can spot CDs with XCP by inspecting a CD closely, checking the left transparent spine on the front of the case for a label that says "CONTENT PROTECTED." The back of these CDs also mention XCP in fine print. You can find pictures of these and other telltale labeling at http://www.eff.org/IP/DRM/Sony-BMG/.
Yes, true. But sort of orthogonal to the thrust of the article, which it sounds like you would agree with. We have this data -- women's interest in computer science is dropping off rapidly.
The questions that raises are: What does that mean? Do we care? and If so, what do we do about it?
Your answers seem to be that it means what it means (that less women are in computer science), that we shouldn't care, and that we shouldn't do anything about it. You're probably right, in that increasing female enrollment in CS for its own sake is not something most people care about. But there are other valid interpretations of the data, and they imply other things.
The article claims not that declining numbers of women is bad on its own, but that it is bad because it signals a general decline in CS interest. It is important, they argue, because there are so many tech jobs being created that companies will have no choice but to outsource. That doesn't mean we should care yet, because those outsourced jobs are for people who chose a different career - they are talking about outsourcing the jobs of imaginary programmers. But somebody definitely does care about that. The proposed solution is an implied suggestion that CS classes be better taught; less technical details, more problem solving and engaging ideas. I'll buy that, as long as the students have the chance to learn those technical details somewhere when they need them.
I would take it a step further. It's not just colleges that are doing a bad job of engaging students; it's the whole school system. When I was in high school, I wanted to take a programming class. But there were only two: Basic and C. And Basic was a prerequisite for C.
Most high schools teach writing at a useful level. If they have the resources to teach programming, why not teach computer science at a useful level? The only high school appropriate program that I know of is the teachscheme thing, but I assume that if there were more interest there would be more appropriate textbooks and curriculums. If we want more, better computer science graduates, we need to start engaging them as soon as possible and teaching them the things that are really important.
Of course, that just means more competition for me. So hooray for declining enrollment.
Attractive nuisance? That is not even close to applying. An attractive nuisance is when something dangerous on your property entices small children to trespass, and they then are physically injured. Like a deep, fun looking hole in your backyard that some kid falls into.
If a child got at his wireless router and it somehow managed to injure them, maybe. But that has nothing to do with this lawsuit anyway.
I don't know about the master's statistic, but statistics on bachelor's degrees were pretty easy to find. The US census page on education has this table that says 23.8% of americans have one.
The census doesn't have data on degrees higher than that. The NCES probably does somewhere, if you really care enough to find it. But really, why bother? This is a not-very-relevant side topic of a side topic already.
This post seems totally off base.
1. The relevant part of the study was done on caucasian males. Even if IQ tests aren't standardized across race and gender, then, that issue is irrelevant.
2. Determining the exact gene may be very difficult. So what? That doesn't mean researchers can't do it. In this case, they did. Or more precisely, they found a rough but significant estimate for the effects of a specific gene on IQ.
3. The meaning of IQ in this sense, as even the inventor of IQ tests said, is really just "what IQ tests measure". So they are accurate by definition. It is definitely valid, though, to argue whether that sense of IQ is a useful one. Or the best one, which it almost definitely isn't.
Unfortunately, the linked article is one of the less informative ones. The Olympian http://www.theolympian.com/apps/pbcs.dll/article?A ID=/20051203/NEWS/51203003 has a lot more, including an estimate from Jirtle that 25% of the Caucasian population has the variant (low IQ) gene. Assuming the sample of 300 was half male and half female, that would mean about 37.5 males with the variant gene and an average IQ of 85, and about 112.5 had an average IQ of 105. That makes the IQ of the group average to 100; the population average now is higher than 100, so that's skewed a bit, but it doesn't really matter if we're talking about significance.
It's been a while since I knew statistics, but I think a Chi square test would be appropriate.
That point is addressed, but the explanation makes no sense. From TFA:
Rebates are used, Baker says, because unlike regular sales, people perceive them as a one-time opportunity to get a product at a lower price than it would normally be sold at. "You want to make believe that there is a special opportunity here and rebates are the best mechanism for that," he says. They are especially valuable to electronics retailers because they don't scale pricing up and down the way some other retailers do. "Their customers haven't been trained," he says.
So... supposedly the reason is that sales aren't percieved as special and rebates are. Really? I don't buy that. No justification for that assertion is provided, and later he admits that customers "absolutely hate rebates". Am I completely missing something here?
When are they adding the continuation-stored-in-the-server feature? Having to do a CPS transform essentially by hand to all CGI scripts is ridiculous. Oh yeah...Perl/PHP/etc. don't support that. Why not?
It depends on where you live. I recall reading that in the midwest, something like 60-70% of water usage is on lawn watering. In wetter climates I believe it's more like 40%. [pdf]http://www.energyrating.gov.au/library/pubs/w a-wateruse.pdf is a relevant study done in western Australia; page 32 has a nice pie chart that shows 51% of water is used on lawns, and 8% on toilets. In non-pdf land, Concord, California says their breakdown is 40% grass / 7% toilets (http://www.ci.concord.ca.us/living/recycle/env-wa ter-use.htm
These are both from reasonably rainy areas. So yes, your gut feeling is very right.
"Your rights end when you walk through the door."
Not true.
From Tinker v Des Moines:
"It can hardly be argued that either students or teachers shed their constitutional rights to freedom of speech or expression at the schoolhouse gate."
etc.
A school, like any other government organization, needs a specific and compelling reason to violate your constitutional rights. The usual argument (as far as I know, IANAL) is that the speech would damage the educational environment if it were allowed.
The watermarks in films at the theater are pretty obvious -- they're big brown dots that flash on the screen for a second or so a couple times during the movie. They're right in the middle of the screen, usually. Even so, I never noticed until I started looking for them. Check it out some time.
Actually, Regal gives a $100 reward to employees that catch someone with a cam. I know cause I used to work for them, and someone at my theater caught someone last summer.
Does anyone besides me notice that the framing at multiplexes sucks?
Slashdot Mirror
http://www.cbsnews.com/stories/2007/04/13/national /main2683325.shtml/
Web 2.0.
To take this a little further:
Generously suppose that a busy airport has 100 terrorists go through it a year, and 30 million passengers. 85% of terrorists is 85 true positives; 8% of non-terrorists is 0.08*(30,000,000 - 100) = 2,399,992 (!) false positives. That's a total of 2,400,092 positives, so the probability of a failed test correctly identifying a terrorist is 85/2,400,092 or about 0.0035%.
0.0035%. Generously. Maybe there is some use for this test, but screening every passenger like that tells you practically nothing. Just another excuse to feel self-righteous about irritating people.
That's true -- assuming you choose such bad criteria for your profile. Race, religion, etc., as you point out, are easily skirted. Why? Because they have fundamentally got nothing to do with terrorism. Blond, white Irish Catholics would have been a reasonable choice. Now it's dark haired, tan Muslims. In ten years, maybe it'll be Mexican Catholics upset about immigration laws. Maybe Eskimos who want their seals back.
But profiling does work. Here's my profile: Nervous facial cues or body language. Suspicious baggage. Suspicious behavior. Probably there are more criteria that would be effective.
The point is that terrorists are very likely to be nervous, and are by definition up to something suspicious. Yes, it's possible to hide these traits as well. But now a terrorist recruiter needs to find someone who is willing to die for a cause, willing to buy into that particular cause, willing to kill lots of civilians, and is in addition a very convincing actor, with the discipline to not stop acting for one second in the airport. The category of potential undetected terrorists is now relatively miniscule.
These magical, invisible terrorists also have to be in the right (wrong) place at the right time and get recruited. I feel pretty good about my odds of getting blown up under a scheme like that.
Malcolm Gladwell has a good article about this stuff.
Sony seems to think so.
http://www.eff.org/deeplinks/archives/004145.php
Wait...Gray Davis is Bush's brother?? Wow, those Bushes are sneaky.
Considering the amount posted on this in the last month or so, it's kind of astounding that there exists a slashdot reader who doesn't know this.
Anyway, google ->
From EFFYes, true. But sort of orthogonal to the thrust of the article, which it sounds like you would agree with. We have this data -- women's interest in computer science is dropping off rapidly.
The questions that raises are: What does that mean? Do we care? and If so, what do we do about it?
Your answers seem to be that it means what it means (that less women are in computer science), that we shouldn't care, and that we shouldn't do anything about it. You're probably right, in that increasing female enrollment in CS for its own sake is not something most people care about. But there are other valid interpretations of the data, and they imply other things.
The article claims not that declining numbers of women is bad on its own, but that it is bad because it signals a general decline in CS interest. It is important, they argue, because there are so many tech jobs being created that companies will have no choice but to outsource. That doesn't mean we should care yet, because those outsourced jobs are for people who chose a different career - they are talking about outsourcing the jobs of imaginary programmers. But somebody definitely does care about that. The proposed solution is an implied suggestion that CS classes be better taught; less technical details, more problem solving and engaging ideas. I'll buy that, as long as the students have the chance to learn those technical details somewhere when they need them.
I would take it a step further. It's not just colleges that are doing a bad job of engaging students; it's the whole school system. When I was in high school, I wanted to take a programming class. But there were only two: Basic and C. And Basic was a prerequisite for C.
Most high schools teach writing at a useful level. If they have the resources to teach programming, why not teach computer science at a useful level? The only high school appropriate program that I know of is the teachscheme thing, but I assume that if there were more interest there would be more appropriate textbooks and curriculums. If we want more, better computer science graduates, we need to start engaging them as soon as possible and teaching them the things that are really important.
Of course, that just means more competition for me. So hooray for declining enrollment.
Attractive nuisance? That is not even close to applying. An attractive nuisance is when something dangerous on your property entices small children to trespass, and they then are physically injured. Like a deep, fun looking hole in your backyard that some kid falls into.
If a child got at his wireless router and it somehow managed to injure them, maybe. But that has nothing to do with this lawsuit anyway.
I don't know about the master's statistic, but statistics on bachelor's degrees were pretty easy to find. The US census page on education has this table that says 23.8% of americans have one.
The census doesn't have data on degrees higher than that. The NCES probably does somewhere, if you really care enough to find it. But really, why bother? This is a not-very-relevant side topic of a side topic already.
Insightful this is.
This post seems totally off base.
1. The relevant part of the study was done on caucasian males. Even if IQ tests aren't standardized across race and gender, then, that issue is irrelevant.
2. Determining the exact gene may be very difficult. So what? That doesn't mean researchers can't do it. In this case, they did. Or more precisely, they found a rough but significant estimate for the effects of a specific gene on IQ.
3. The meaning of IQ in this sense, as even the inventor of IQ tests said, is really just "what IQ tests measure". So they are accurate by definition. It is definitely valid, though, to argue whether that sense of IQ is a useful one. Or the best one, which it almost definitely isn't.
Unfortunately, the linked article is one of the less informative ones. The Olympian http://www.theolympian.com/apps/pbcs.dll/article?A ID=/20051203/NEWS/51203003 has a lot more, including an estimate from Jirtle that 25% of the Caucasian population has the variant (low IQ) gene. Assuming the sample of 300 was half male and half female, that would mean about 37.5 males with the variant gene and an average IQ of 85, and about 112.5 had an average IQ of 105. That makes the IQ of the group average to 100; the population average now is higher than 100, so that's skewed a bit, but it doesn't really matter if we're talking about significance.
It's been a while since I knew statistics, but I think a Chi square test would be appropriate.
That point is addressed, but the explanation makes no sense. From TFA:
So... supposedly the reason is that sales aren't percieved as special and rebates are. Really? I don't buy that. No justification for that assertion is provided, and later he admits that customers "absolutely hate rebates". Am I completely missing something here?
I don't actually expect Apache to support this, ever. But it's a neat idea. Here's a paper by people who know more and write better than me:n ke.pdf
: ase.informatik.uni-essen.de/olbib/2001graunke.pdf
Slow PDF: http://ase.informatik.uni-essen.de/olbib/2001grau
Ugly HTML conversion: http://64.233.161.104/search?q=cache:Cx4ndEP2FHYJ
When are they adding the continuation-stored-in-the-server feature? Having to do a CPS transform essentially by hand to all CGI scripts is ridiculous. Oh yeah...Perl/PHP/etc. don't support that. Why not?
It depends on where you live. I recall reading that in the midwest, something like 60-70% of water usage is on lawn watering. In wetter climates I believe it's more like 40%. [pdf]http://www.energyrating.gov.au/library/pubs/w a-wateruse.pdf is a relevant study done in western Australia; page 32 has a nice pie chart that shows 51% of water is used on lawns, and 8% on toilets. In non-pdf land, Concord, California says their breakdown is 40% grass / 7% toilets (http://www.ci.concord.ca.us/living/recycle/env-wa ter-use.htm
These are both from reasonably rainy areas. So yes, your gut feeling is very right.
"Your rights end when you walk through the door." Not true. From Tinker v Des Moines: "It can hardly be argued that either students or teachers shed their constitutional rights to freedom of speech or expression at the schoolhouse gate." etc. A school, like any other government organization, needs a specific and compelling reason to violate your constitutional rights. The usual argument (as far as I know, IANAL) is that the speech would damage the educational environment if it were allowed.
The watermarks in films at the theater are pretty obvious -- they're big brown dots that flash on the screen for a second or so a couple times during the movie. They're right in the middle of the screen, usually. Even so, I never noticed until I started looking for them. Check it out some time.
Actually, Regal gives a $100 reward to employees that catch someone with a cam. I know cause I used to work for them, and someone at my theater caught someone last summer. Does anyone besides me notice that the framing at multiplexes sucks?