I've consulted with major research firms who use government data. Universally we find that the data haven't been verified and a little work shows massive inconsistencies therein. When recovery.org was showing jobs in zip codes that don't exist, etc., I wasn't surprise - it's par for the course.
I'll reserve judgment, but making data available is one thing; collecting usable data is something entirely different.
* 1. Make sure that the font's period (.) sign and the comman (,) sign is BIG, to aid in the debugging process.
Good advice.
Many many many moons ago I wrote a script to markup foreign language HTML pages with English glosses (rollover translations). It just compared strings to db entries, and inserted some DHTML around the original string.
Running it in French returned a bunch of unfound strings for words and phrases I knew were in the db. I finally had it throw an error with the original string and the five closest matches from the db. It read:
Original: c'est ; closest match: c'est.
I was so frustrated I took a screenshot and overlaid the two strings. The apostrophe was one pixel lower on the db's string than the original. It turns out that they were different ascii codes, but the font made them look the same.
In the social sciences, we see a different angle on this.
There is "research" under the umbrella "critical methods" which rather than control for bias, the authors state their bias and make no efforts to gather any evidence to refute their own a priori conclusions. Part of their justification is that opponents have the same opportunity: They can gather evidence to support their viewpoint, and then it will all get worked out in the literature.
The problem is that for many pressing issues, political correctness will not allow the other side to voice their opinion. So one end becomes more and more "founded" in "research" while the other - which may be more correct (from a positivist POV) - gets less and less credible.
I agree with you on this target variable because we can assess child literacy in a more objective manner.
However, there are plenty of psychological constructs for which self-assessment is the most accurate method of measurement. The results of those measures should at some point be compared against other non-self-reported data, but they are very useful.
I was at Mt. Rushmore this summer and they have a video of FDR visiting the site. He postulated that the sculpture would be around for 10,000 years and wondered what people then would think of us.
"Let us hope that at least they will give us the benefit of the doubt, that they will believe we have honestly striven every day and generation to preserve for our descendants a decent land to live in and a decent form of government to operate under."
That's not too much for to ask of people - to offer the benefit of doubt to others. But I think it's important to separate the people (the American citizenry) from the politicians (including the one who spoke those words). I'll give the Americans of the 1930's credit for doing what they thought was right, but no politician deserves that benefit.
If you had your way I imagine no one in their right mind would then go into politics.
You should google "General George Washington Resigning His Commission by John Trumbull". Or, better, you should go see that painting on display over Washington's saber at the Smithsonian.
The motivation for political *service* shouldn't be instrumental, but intrinsic.
After the election here in NY 23rd, some have wondered about the viability of third parties. The best comment I heard - which rang frustratingly true to this non-partisan - was this: "No third party can survive in the U.S. system because as soon as their platform becomes popular enough to win a seat, one of the major parties will assimilate that platform."
As I thought about this, I came to the conclusion that the two-party system isn't the disease, but a symptom.
The decimal comma is an SI standard as much as the decimal point and its usage is preferred (according to Wikipedia) in Germany, France, Belgium, Denmark, Italy, Serbia, Croatia, Bosnia, Estonia, Finland, Hungary, Poland, Czech Republic, Slovakia, French Canada, Romania, Sweden and much of the rest of Europe.
I was in the Louvre looking at the old French crown jewels when I heard someone read the display: "Fifty-four THOUSAND carats!?!?! WOW!"
While I've never seen anything like celebrating an opponent's death, in my social science experience, I've witnessed rampant conclusion-driven methodology.
"Do you think that because we included XYZ in our sampling that it's clouding the results?"
"Don't tell me what the data say; I know what's really happening and the data are wrong!"
etc.
The way science is funded is not amenable to honest science. If the track you're leading dries up, switching tracks isn't really an option because all the other tracks have people leading them already.
Re:Gee, it's almost like they have a monopoly or s
on
Less Than Free
·
· Score: 5, Interesting
When I first read this I thought about IBM back in the day. They could put a small company out of business simply by announcing, "Yeah, we're working on that too." And they had to fight off some well-founded lawsuits. Eventually, IBM became known for quiet and consistent R&D (Giant MR comes to mind) because they had to watch what they said.
Will that day come for Google? I think not (or it's a long way off). IBM's issues with the courts came around the same time Ma Bell was dismantled, which couldn't happen now.
The lang blog guy hypothesizes that the differences might not be found in replication. Great. The study is pretty straight forward, go replicate it. It seems something that a few grad students in a seminar could knock out in a semester. Go for it.
I had an lawyer acquaintance with whom I discussed software licensing. He saw no problem with it, but I have since learned that many lawyers only think in terms of legal and illegal, not right or wrong, sensible or insensible, etc.
I mentioned that you cannot license a book, and he told me I was dead wrong; he had himself helped a manufacturer write the license for their *printed* maintenance manuals.
I sent him a link to the Penguin Books case that led to the doctrine of First Sale. He wrote back, "A) This appears to apply to the very limited context of book resellers. B) I'm sure there's been case law in the last 80 years to override this decision."
I personally still believe it's not legal, but IANAL.
I was on an evaluation team that was charged with determining how well a government program had addressed a "shortage" of a specific skill set. On the committee was an economist from a big university. He opened the meeting with the comment: There is no shortage; the government is just not willing to pay market value.
IAAL and it is legal malpractice to not double-check the prosecution's sentencing algorithm and recommendations to the judge...
That's great! So, where do I bring this up? What happens to the lawyers who make these mistakes?
I'm not being snarky; I sincerely wish to know.
When I moved to the East Coast I found it odd that I needed a lawyer to buy a house. I had bought and sold out West on a handshake and a contract. I was told that out here, where property has been bought and sold for centuries, the lawyers would check deeds, get the property surveyed, etc. OK, I got that.
But what happens if in ten years, somebody's great grandson comes by with a deed on the northern half of my land? Do I get my lawyer fees back?
Similarly, a family member of mine just settled on her divorce. When it came time to sign the papers, her ex acted shocked at the agreement. His laywer said, "You can't blame him. He just didn't understand the terms." So, then, can we blame the laywer who was supposed to explain it to him?
Coming back to the topic here: So the defense attorney screwed up. 1) What are the paths of recourse for those who suffered from the mistake? 2) What are the consequences to the lawyer who screwed up?
Because, in my dealing with lawyers, they almost never get called out on their mistakes.
But this isn't experience through "classes," it's experience in reading and discussion groups with colleagues in my department. I would consider myself also to be an "insider" to academic discourse, but I'm definitely not in the circle of discipline studies. Therefore the lens through which I view those fields contrasts them with traditional social science models.
The fact that you have tacitly acknowledged that these theories are simply alternative models and not reality itself is more than I have ever gotten from three books, two guest speakers, and a dozen faculty members. So, thank you for livening my spirits on the issue.
"They" are not shifting the meaning and "they" are not tossing out blanket accusations or immunities. Some extremists may be willing to go that far, but it's definitely the minority view.
Then I must be surrounded by extremists.
Kendall, whom I cited, is rather opaque in her attempts to remove traditional meaning from race-related terms. For example, when she cites the definitions of "prejudice" from the Random House Dictionary, she includes only the first and third definitions (without proper ellipses in the citation, I might add), leaving out the second definition that would question her hypothesis. I take that as an attempt to co-opt a term.
Also, when I speak with people who hold to these philosophies, they seem to assume their interlocutors hold to the same definitions. (See your comment's sibling for a similar experience.) Again, evidence that they are attempting to shift the language. That has not been my experience in social science outside of the discipline studies (Women's Studies, Queer Studies, Black Studies, Disability Studies, etc.) arena.
The author was clearly talking about increasing diversity in games and how the standard space-marine character pushes against that
Why do we need diversity anyway? Does it matter if you're playing a white space marine who shoots aliens or a black space marine who shoots aliens? Next up: chess is racist, because while you can play either white or black, there is no Native American side.
There is an area of academic pursuit that is actively trying to shift the meaning of "racism" and racist to encompass any white (written with a lowercase "w") member of society. Understanding White Privilege by Fances Kendall is a good read on the matter. Basically, our society is racist because members of different races exercise varying degrees of privilege. Because members of privileged groups cannot divorce themselves of the privileges that they receive from our racist society, all members of the privileged race are racists. Conversely, no Black (written with an uppercase "B") can be racist.
Such reasoning is extended to declare any member of the privileged sex "sexist," the privileged sexual orientation "homophobic," etc. This "privilege theory" was the basis for the now-retracted freshmen curriculum at the University of Delaware.
Lest I be flamed for this, let me be clear that I completely reject these notions. But they are central to many people's understanding of "racist." I've found that one's definition of the term to be central to many disagreements.
Both good points, but I wonder who those 35.5 kids are and what they're studying.
I know in most European countries students are tracked into a profession before age 15. Many components of U.S. education (literature, government, etc.) are only taught to people going into those fields. In the U.S., we teach general ed all the way through 12th grade.
Also the American education has - for good or ill - taken on the task of educating *everyone*. This means the deviant, the disabled, and the gifted are in the same system. Other countries can move a child to a special school. In the U.S., schools have to go to court to have a student reassigned to a more suitable institution.
So salary and class size may be relevant factors given the social complexities of teaching in America. But if you could enlighten us on how these issues play out in Japan and the other countries you cite, I'd appreciate it.
More importantly than the pay, I wasn't ALLOWED to be a good teacher. I was asked to teach stuff that was horrifically boring, in a boring way. Because success was determined based on how well kids filled in bubbles on a test. How do you demonstrate the ability to do science with a bubble-sheet? You don't. You demonstrate that you can MEMORIZE science facts....Standardized tests are blatantly anti-education. They measure the ability and motivation of a kid to memorize answers from other days, and fill in those answers on one day out of 180.
Ah, the misplaced hatred of standardized tests. Never mind that such a label is also applied to psychological profiles that are beneficial in classification and therapy decisions, or that those "other countries" who are supposedly so far ahead of the U.S. use standardized tests with higher stakes than Americans could imagine. (When was the last time someone committed suicide for failing their state tests?)
The effectiveness of an assessment is largely independent of its format. I've seen rote-recall essay and practical (lab) assessment tasks, and I've seen critical thinking restricted-response items. But good items take work to develop - work that most states are not willing to invest. The typical method is for the state to contract out the development of their tests to a textbook publisher - who will often sell the tests as a loss leader for textbooks. My state (NY) releases the technical reports for the publishers, but then doesn't do anything about low reliabilities (alpha of.50 on the CR items on the 3rd Grade Math in 2006), inaccurate placements (only 90% of 8th graders were accurately classified pass/fail on the English/Language Arts test in 2006), or other bizarre psychometric stats (only 24% of the variance in the student scores being explained by the dominant factor).
Rather than blame an inanimate objects (standardized tests), why not blame the policy makers who use them inappropriately and in violation of the 1999 Standards for Educational and Psychological Testing put out by AERA, APA, and NCME?
Oh, and the issue of testing 1 day out of 180 - Assessment people have known about that for almost a century. It's called Classical Test Theory and error due to occasion sampling. There are techniques to establish and mitigate its effect on test scores, but, again, states don't really care about the quality of the assessments.
Slashdot Mirror
I've consulted with major research firms who use government data. Universally we find that the data haven't been verified and a little work shows massive inconsistencies therein. When recovery.org was showing jobs in zip codes that don't exist, etc., I wasn't surprise - it's par for the course.
I'll reserve judgment, but making data available is one thing; collecting usable data is something entirely different.
Good advice.
Many many many moons ago I wrote a script to markup foreign language HTML pages with English glosses (rollover translations). It just compared strings to db entries, and inserted some DHTML around the original string.
Running it in French returned a bunch of unfound strings for words and phrases I knew were in the db. I finally had it throw an error with the original string and the five closest matches from the db. It read:
I was so frustrated I took a screenshot and overlaid the two strings. The apostrophe was one pixel lower on the db's string than the original. It turns out that they were different ascii codes, but the font made them look the same.
That's when I started using hex editors.
Am I the only one who looked at the that lab and rails and thought of Gordon Freeman "On a Rail"?
I thought so.
In the social sciences, we see a different angle on this.
There is "research" under the umbrella "critical methods" which rather than control for bias, the authors state their bias and make no efforts to gather any evidence to refute their own a priori conclusions. Part of their justification is that opponents have the same opportunity: They can gather evidence to support their viewpoint, and then it will all get worked out in the literature.
The problem is that for many pressing issues, political correctness will not allow the other side to voice their opinion. So one end becomes more and more "founded" in "research" while the other - which may be more correct (from a positivist POV) - gets less and less credible.
I agree with you on this target variable because we can assess child literacy in a more objective manner.
However, there are plenty of psychological constructs for which self-assessment is the most accurate method of measurement. The results of those measures should at some point be compared against other non-self-reported data, but they are very useful.
I was at Mt. Rushmore this summer and they have a video of FDR visiting the site. He postulated that the sculpture would be around for 10,000 years and wondered what people then would think of us.
"Let us hope that at least they will give us the benefit of the doubt, that they will believe we have honestly striven every day and generation to preserve for our descendants a decent land to live in and a decent form of government to operate under."
That's not too much for to ask of people - to offer the benefit of doubt to others. But I think it's important to separate the people (the American citizenry) from the politicians (including the one who spoke those words). I'll give the Americans of the 1930's credit for doing what they thought was right, but no politician deserves that benefit.
You should google "General George Washington Resigning His Commission by John Trumbull". Or, better, you should go see that painting on display over Washington's saber at the Smithsonian.
The motivation for political *service* shouldn't be instrumental, but intrinsic.
After the election here in NY 23rd, some have wondered about the viability of third parties. The best comment I heard - which rang frustratingly true to this non-partisan - was this: "No third party can survive in the U.S. system because as soon as their platform becomes popular enough to win a seat, one of the major parties will assimilate that platform."
As I thought about this, I came to the conclusion that the two-party system isn't the disease, but a symptom.
I was in the Louvre looking at the old French crown jewels when I heard someone read the display: "Fifty-four THOUSAND carats!?!?! WOW!"
While I've never seen anything like celebrating an opponent's death, in my social science experience, I've witnessed rampant conclusion-driven methodology.
"Do you think that because we included XYZ in our sampling that it's clouding the results?"
"Don't tell me what the data say; I know what's really happening and the data are wrong!"
etc.
The way science is funded is not amenable to honest science. If the track you're leading dries up, switching tracks isn't really an option because all the other tracks have people leading them already.
When I first read this I thought about IBM back in the day. They could put a small company out of business simply by announcing, "Yeah, we're working on that too." And they had to fight off some well-founded lawsuits. Eventually, IBM became known for quiet and consistent R&D (Giant MR comes to mind) because they had to watch what they said.
Will that day come for Google? I think not (or it's a long way off). IBM's issues with the courts came around the same time Ma Bell was dismantled, which couldn't happen now.
My high school trig teacher tried to get out of a ticket once by working out the problem with the cop, but using sine instead of cosine.
Serious?!?!
I guess you didn't actually read Twilight , did you?
Which is why replication so important.
The lang blog guy hypothesizes that the differences might not be found in replication. Great. The study is pretty straight forward, go replicate it. It seems something that a few grad students in a seminar could knock out in a semester. Go for it.
I had an lawyer acquaintance with whom I discussed software licensing. He saw no problem with it, but I have since learned that many lawyers only think in terms of legal and illegal, not right or wrong, sensible or insensible, etc.
I mentioned that you cannot license a book, and he told me I was dead wrong; he had himself helped a manufacturer write the license for their *printed* maintenance manuals.
I sent him a link to the Penguin Books case that led to the doctrine of First Sale. He wrote back, "A) This appears to apply to the very limited context of book resellers. B) I'm sure there's been case law in the last 80 years to override this decision."
I personally still believe it's not legal, but IANAL.
I was on an evaluation team that was charged with determining how well a government program had addressed a "shortage" of a specific skill set. On the committee was an economist from a big university. He opened the meeting with the comment: There is no shortage; the government is just not willing to pay market value.
So the company that insures my title will get the money back from the lawyer whom I paid to cleared the title?
Or does the lawyer keep the money regardless of the outcome?
You need to add the stipulation "... in our FUBAR system."
There are ways to reform it, but short of a revolution, they're not happening.
That's great! So, where do I bring this up? What happens to the lawyers who make these mistakes?
I'm not being snarky; I sincerely wish to know.
When I moved to the East Coast I found it odd that I needed a lawyer to buy a house. I had bought and sold out West on a handshake and a contract. I was told that out here, where property has been bought and sold for centuries, the lawyers would check deeds, get the property surveyed, etc. OK, I got that.
But what happens if in ten years, somebody's great grandson comes by with a deed on the northern half of my land? Do I get my lawyer fees back?
Similarly, a family member of mine just settled on her divorce. When it came time to sign the papers, her ex acted shocked at the agreement. His laywer said, "You can't blame him. He just didn't understand the terms." So, then, can we blame the laywer who was supposed to explain it to him?
Coming back to the topic here: So the defense attorney screwed up. 1) What are the paths of recourse for those who suffered from the mistake? 2) What are the consequences to the lawyer who screwed up?
Because, in my dealing with lawyers, they almost never get called out on their mistakes.
I can agree with that.
But this isn't experience through "classes," it's experience in reading and discussion groups with colleagues in my department. I would consider myself also to be an "insider" to academic discourse, but I'm definitely not in the circle of discipline studies. Therefore the lens through which I view those fields contrasts them with traditional social science models.
The fact that you have tacitly acknowledged that these theories are simply alternative models and not reality itself is more than I have ever gotten from three books, two guest speakers, and a dozen faculty members. So, thank you for livening my spirits on the issue.
Then I must be surrounded by extremists.
Kendall, whom I cited, is rather opaque in her attempts to remove traditional meaning from race-related terms. For example, when she cites the definitions of "prejudice" from the Random House Dictionary, she includes only the first and third definitions (without proper ellipses in the citation, I might add), leaving out the second definition that would question her hypothesis. I take that as an attempt to co-opt a term.
Also, when I speak with people who hold to these philosophies, they seem to assume their interlocutors hold to the same definitions. (See your comment's sibling for a similar experience.) Again, evidence that they are attempting to shift the language. That has not been my experience in social science outside of the discipline studies (Women's Studies, Queer Studies, Black Studies, Disability Studies, etc.) arena.
Which color goes first in chess? I'm just sayin'.
There is an area of academic pursuit that is actively trying to shift the meaning of "racism" and racist to encompass any white (written with a lowercase "w") member of society. Understanding White Privilege by Fances Kendall is a good read on the matter. Basically, our society is racist because members of different races exercise varying degrees of privilege. Because members of privileged groups cannot divorce themselves of the privileges that they receive from our racist society, all members of the privileged race are racists. Conversely, no Black (written with an uppercase "B") can be racist.
Such reasoning is extended to declare any member of the privileged sex "sexist," the privileged sexual orientation "homophobic," etc. This "privilege theory" was the basis for the now-retracted freshmen curriculum at the University of Delaware.
Lest I be flamed for this, let me be clear that I completely reject these notions. But they are central to many people's understanding of "racist." I've found that one's definition of the term to be central to many disagreements.
Both good points, but I wonder who those 35.5 kids are and what they're studying.
I know in most European countries students are tracked into a profession before age 15. Many components of U.S. education (literature, government, etc.) are only taught to people going into those fields. In the U.S., we teach general ed all the way through 12th grade.
Also the American education has - for good or ill - taken on the task of educating *everyone*. This means the deviant, the disabled, and the gifted are in the same system. Other countries can move a child to a special school. In the U.S., schools have to go to court to have a student reassigned to a more suitable institution.
So salary and class size may be relevant factors given the social complexities of teaching in America. But if you could enlighten us on how these issues play out in Japan and the other countries you cite, I'd appreciate it.
Ah, the misplaced hatred of standardized tests. Never mind that such a label is also applied to psychological profiles that are beneficial in classification and therapy decisions, or that those "other countries" who are supposedly so far ahead of the U.S. use standardized tests with higher stakes than Americans could imagine. (When was the last time someone committed suicide for failing their state tests?)
The effectiveness of an assessment is largely independent of its format. I've seen rote-recall essay and practical (lab) assessment tasks, and I've seen critical thinking restricted-response items. But good items take work to develop - work that most states are not willing to invest. The typical method is for the state to contract out the development of their tests to a textbook publisher - who will often sell the tests as a loss leader for textbooks. My state (NY) releases the technical reports for the publishers, but then doesn't do anything about low reliabilities (alpha of .50 on the CR items on the 3rd Grade Math in 2006), inaccurate placements (only 90% of 8th graders were accurately classified pass/fail on the English/Language Arts test in 2006), or other bizarre psychometric stats (only 24% of the variance in the student scores being explained by the dominant factor).
Rather than blame an inanimate objects (standardized tests), why not blame the policy makers who use them inappropriately and in violation of the 1999 Standards for Educational and Psychological Testing put out by AERA, APA, and NCME?
Oh, and the issue of testing 1 day out of 180 - Assessment people have known about that for almost a century. It's called Classical Test Theory and error due to occasion sampling. There are techniques to establish and mitigate its effect on test scores, but, again, states don't really care about the quality of the assessments.