"By your broken logic, a cop should be able to jam a camera up your ass since you might be carrying illegal narcotics up there."
Current cases studies on such:
"Last week, news wires, blogs and pundits lit up with the horrifying story of David Eckert, a New Mexico man who last January was subjected to a series of invasive and degrading drug search procedures after a traffic stop. The procedures, which included x-rays, digital anal penetration, enemas and a colonoscopy, were all performed without Eckert's consent..."
" Coursera A/B tests most every aspect of the student learning experience and makes decisions based on what results in the best student learning outcome. "
Citation needed? I'm really curious about this, because (for example) Udacity makes the same claims and on closer inspection it seems to be basically BS. So if there are concrete examples of how Coursera has done this, I'd like to read about it.
"But only if your sample is not biased because you determine the average based upon your sample. If the sample is badly distributed or biased in some way then your estimate of the population mean will be similarly biased and you will not be able to make meaningful inferences regarding the population mean. Problems are worse with small samples."
You are almost terrifyingly confused; what you wrote above is nonsense. What I wrote earlier, "the likelihood that your one sample average is usefully close to the population average", is a probability question, and it's not dependent on any one sample. It is in fact the probability that your sample is biased or not (as you put it). Saying that this probability changes because of one oddball sample is incorrect.
Look, go to your linked applet. Set the sample sizes to 25. Keep the default population shapes (no skew, i.e., normal) and click "simulate". The Type I error rate will almost certainly be 0.05 (rounded off). Now change the population shapes to "severely skewed" (the most severe available), keep the sample size at 25, and run again. The Type I error rate will again be almost exactly 0.05.
The population shape makes no difference. For a sample of at least 25 or anything higher, the shape of the possible sample means (not shown in the applet) is normally distributed anyway, regardless of the population shape. That's the subject of the CLT anytime it's being discussed. Only for itty-bitty sample sizes (less than 25; really the subject of the applet) does the population shape make any difference.
"You're usually applying your test to a single sample and that may well be very non-normal."
No, usually you're applying your test to the average of a single sample, and since possible averages are always normally distributed (for a fair sample size), you can indeed use that to assess the likelihood that your one sample average is usefully close to the population average.
Argggh, you guys are all missing the point that the Central Limit Theorem is about the sampling distribution of the sample mean, i.e., the sample space for possible averages that you get as a result of your sampling process. (Or a proportion, equivalent to an average on booleans 0 or 1.) So you can always use this knowledge, for a fair sample size, to assess how likely it is that your sample mean is usefully close to the population mean.
What are some things that definitely have an approximately normal distribution for a fair sample size? The average of anything. Yes to biological length or height. Yes to mechanical error. Yes to the average of some diameters, surface areas, or volumes of berries or anything else. All sample averages, or sums or differences of averages, or proportions (or more fundamentally any statistic based on addition), are in fact normally distributed for a fair sample size. No doubt about it.
"If they aren't biased [outliers], then their effects cancel."
Oh, god no. The book I teach out of says that if outliers exist, it's required to do the regression both with and without the outliers and compare. Frequently there will be a big difference. (Weiss, Introductory Statistics, Sec. 14.2)
As others have said, the problem in many cases is not computational power, but expense or difficulty or even ethics of getting a large data set. What about a medical trial -- do want to necessitate giving some experimental medicine to 10,000 people before assessing whether it's a good idea or not?
"Because you don't need to see two copies of the same title."
So do I still get the benefit of, "If it's at the store, you could just buy it there"? Wouldn't doing so remove the ability for anyone else to see that title?
Why does everyone assume that performance follows a bell curve? Why not right-skewed, reverse-J shaped, or multimodal?
For what it's worth, most of the college statistics tests that I give have a bimodal distribution. Mostly A's, lots of F's, (you either get it or you don't) very little in the middle ground. I think it's best to be honest about that and not delude ourselves with manipulated data.
"Bookstores could downsize their physical presence, keep most of their inventory in inexpensive rural warehouses like Amazon, and offer free overnight or 2-day shipping to the store, no membership required."
So I get to travel to the store twice and not even view the product in the meantime? I guess?
"Some of this going to be generational... some people who are very attached to it are still alive."
Weak argument in a lot of ways. You seem to make the assumption that DST is some kind of majority-rules thing, when it's not. DST is one of those fairly boneheaded ideas that come up once in a while in a representational Congress (kind of like setting pi to 3), which is dumb but so low-priority that there's no major political pushback against it. And it's led by a fairly small group of marketers who think there's more shopping time from it. I bet that right now or any time in the past if there was a general vote on the issue it would be abolished.
At any rate, the lack of economic benefit and demonstrated increase in heart attacks around DST makes it unwarranted if not downright cruel. But, you know, America.
In principle, I agree with you. But in America, which is seriously more likely to happen: (a) widespread socialist safety net, or (b) imprisoning tens of millions of people that can't support themselves? The "technology will free us from work" prayer has utterly failed in our economic context for over 100 years now.
"It's a distinctly human behavior: empire building."
This metaphor is no better. An empire is a bunch of kingdoms or geographic entities forged together into one institution. If Amazon were buying up/taking over existing brick-and-mortar stores, then empire-building would be a pretty good metaphor. But it's not -- it's consuming their custom and destroying those businesses entirely. So in that regard the predation metaphor is really better.
Possibly because emergency drivers are breaking the normal rules of driving like going through red lights and such. Will the robo-cars be equipped with a subroutine to know when it's safe to pass on a solid dividing line and blow through a red light? Perhaps not.
That's not true. MTA is not a private organization, it's a New York "public benefit corporation". The governing board has its members appointed by regional governments (5 by NYS governor, 4 by NYC mayor, 3 by nearby counties, etc.)
The files are PAPER FILE FOLDERS. With notes ON PAPER. The very first thing in the article is a PICTURE of said PAPER FILES. On PAPER.
"But it wasn’t until a month later, on Sept. 10, that Hudson was informed by Bosch that five files including her handwritten and typed notes from interviews with numerous confidential sources and other documents had been taken during the raid."
Slashdot Mirror
"In short, you have to piss a lot of people off and be committed to accomplishing your grim task."
Great turn of phrase. Sounds a lot like teaching remedial math.
"This is why 'Stop and Frisk' was outlawed in New York (after way too long of it being done as the court case dragged on)."
As delightful as it would be for that to be the case, Stop and Frisk was NOT outlawed in New York.
In August a federal judge demanded that monitoring of the practice be put in place. Then in October that was overturned on appeal.
http://www.nytimes.com/2013/11/01/nyregion/court-blocks-stop-and-frisk-changes-for-new-york-police.html
"By your broken logic, a cop should be able to jam a camera up your ass since you might be carrying illegal narcotics up there."
Current cases studies on such:
"Last week, news wires, blogs and pundits lit up with the horrifying story of David Eckert, a New Mexico man who last January was subjected to a series of invasive and degrading drug search procedures after a traffic stop. The procedures, which included x-rays, digital anal penetration, enemas and a colonoscopy, were all performed without Eckert's consent..."
http://www.huffingtonpost.com/2013/11/11/anal-probes-and-the-drug-_n_4254600.html
" Coursera A/B tests most every aspect of the student learning experience and makes decisions based on what results in the best student learning outcome. "
Citation needed? I'm really curious about this, because (for example) Udacity makes the same claims and on closer inspection it seems to be basically BS. So if there are concrete examples of how Coursera has done this, I'd like to read about it.
Slashdot is clearly not the place for you. Go here instead, it's more for people like you: http://www.timecube.com/
"But only if your sample is not biased because you determine the average based upon your sample. If the sample is badly distributed or biased in some way then your estimate of the population mean will be similarly biased and you will not be able to make meaningful inferences regarding the population mean. Problems are worse with small samples."
You are almost terrifyingly confused; what you wrote above is nonsense. What I wrote earlier, "the likelihood that your one sample average is usefully close to the population average", is a probability question, and it's not dependent on any one sample. It is in fact the probability that your sample is biased or not (as you put it). Saying that this probability changes because of one oddball sample is incorrect.
Look, go to your linked applet. Set the sample sizes to 25. Keep the default population shapes (no skew, i.e., normal) and click "simulate". The Type I error rate will almost certainly be 0.05 (rounded off). Now change the population shapes to "severely skewed" (the most severe available), keep the sample size at 25, and run again. The Type I error rate will again be almost exactly 0.05.
The population shape makes no difference. For a sample of at least 25 or anything higher, the shape of the possible sample means (not shown in the applet) is normally distributed anyway, regardless of the population shape. That's the subject of the CLT anytime it's being discussed. Only for itty-bitty sample sizes (less than 25; really the subject of the applet) does the population shape make any difference.
"In the meantime, with a P of 0.05, you'd label it as a tentative conclusion, a likely theory."
Great, so we agree: Getting a P of less than 0.05 with a sample of 200 gets you published and other machinery acting on that signal.
"You're usually applying your test to a single sample and that may well be very non-normal."
No, usually you're applying your test to the average of a single sample, and since possible averages are always normally distributed (for a fair sample size), you can indeed use that to assess the likelihood that your one sample average is usefully close to the population average.
Argggh, you guys are all missing the point that the Central Limit Theorem is about the sampling distribution of the sample mean, i.e., the sample space for possible averages that you get as a result of your sampling process. (Or a proportion, equivalent to an average on booleans 0 or 1.) So you can always use this knowledge, for a fair sample size, to assess how likely it is that your sample mean is usefully close to the population mean.
What are some things that definitely have an approximately normal distribution for a fair sample size? The average of anything. Yes to biological length or height. Yes to mechanical error. Yes to the average of some diameters, surface areas, or volumes of berries or anything else. All sample averages, or sums or differences of averages, or proportions (or more fundamentally any statistic based on addition), are in fact normally distributed for a fair sample size. No doubt about it.
"If they aren't biased [outliers], then their effects cancel."
Oh, god no. The book I teach out of says that if outliers exist, it's required to do the regression both with and without the outliers and compare. Frequently there will be a big difference. (Weiss, Introductory Statistics, Sec. 14.2)
As others have said, the problem in many cases is not computational power, but expense or difficulty or even ethics of getting a large data set. What about a medical trial -- do want to necessitate giving some experimental medicine to 10,000 people before assessing whether it's a good idea or not?
"Because you don't need to see two copies of the same title."
So do I still get the benefit of, "If it's at the store, you could just buy it there"? Wouldn't doing so remove the ability for anyone else to see that title?
Why does everyone assume that performance follows a bell curve? Why not right-skewed, reverse-J shaped, or multimodal?
For what it's worth, most of the college statistics tests that I give have a bimodal distribution. Mostly A's, lots of F's, (you either get it or you don't) very little in the middle ground. I think it's best to be honest about that and not delude ourselves with manipulated data.
" I imagine the crackhead thinks the idea of life without crack is absurd, an unrealistic pipe dream."
Or an unrealistic pipeless dream, as the case may be.
"Bookstores could downsize their physical presence, keep most of their inventory in inexpensive rural warehouses like Amazon, and offer free overnight or 2-day shipping to the store, no membership required."
So I get to travel to the store twice and not even view the product in the meantime? I guess?
I enjoyed 1984 a lot more back when it was fictional.
http://www.youtube.com/watch?v=uj2dmQruJXs
"except that nobody ever loses their job as a bus driver. public unions ftw!"
Liar.
Google: "bus driver loses job".
About 1,840,000 results (0.32 seconds)
"Some of this going to be generational... some people who are very attached to it are still alive."
Weak argument in a lot of ways. You seem to make the assumption that DST is some kind of majority-rules thing, when it's not. DST is one of those fairly boneheaded ideas that come up once in a while in a representational Congress (kind of like setting pi to 3), which is dumb but so low-priority that there's no major political pushback against it. And it's led by a fairly small group of marketers who think there's more shopping time from it. I bet that right now or any time in the past if there was a general vote on the issue it would be abolished.
At any rate, the lack of economic benefit and demonstrated increase in heart attacks around DST makes it unwarranted if not downright cruel. But, you know, America.
http://www.sciencedaily.com/releases/2012/03/120307162555.htm
"I trust the police implicitly and require no evidence in support of their claims! Corruption in law enforcement is not rampant!"
In principle, I agree with you. But in America, which is seriously more likely to happen: (a) widespread socialist safety net, or (b) imprisoning tens of millions of people that can't support themselves? The "technology will free us from work" prayer has utterly failed in our economic context for over 100 years now.
"It's a distinctly human behavior: empire building."
This metaphor is no better. An empire is a bunch of kingdoms or geographic entities forged together into one institution. If Amazon were buying up/taking over existing brick-and-mortar stores, then empire-building would be a pretty good metaphor. But it's not -- it's consuming their custom and destroying those businesses entirely. So in that regard the predation metaphor is really better.
Possibly because emergency drivers are breaking the normal rules of driving like going through red lights and such. Will the robo-cars be equipped with a subroutine to know when it's safe to pass on a solid dividing line and blow through a red light? Perhaps not.
That's not true. MTA is not a private organization, it's a New York "public benefit corporation". The governing board has its members appointed by regional governments (5 by NYS governor, 4 by NYC mayor, 3 by nearby counties, etc.)
http://en.wikipedia.org/wiki/Metropolitan_Transportation_Authority_%28New_York%29#Governance
Regarding FOIA requests, it even has a web page informing one where and how to send such a request:
http://www.mta.info/nyct/rules/freedom.htm
The files are PAPER FILE FOLDERS. With notes ON PAPER. The very first thing in the article is a PICTURE of said PAPER FILES. On PAPER.
"But it wasn’t until a month later, on Sept. 10, that Hudson was informed by Bosch that five files including her handwritten and typed notes from interviews with numerous confidential sources and other documents had been taken during the raid."
http://dailycaller.com/2013/10/25/exclusive-feds-confiscate-investigative-reporters-confidential-files-during-raid/2/