The (ongoing) War on Data

I know we've been through this before, but from the New York Times (via ataxingmatter):

One bill, introduced in the House by Jeff Duncan of South Carolina, would effectively end all surveys by the bureau, except for the decennial census, and even that would be limited to counting noses — a silly interpretation of the census’s mandate. Banning the surveys would make it impossible to compile reliable data on employment, productivity, health, housing, poverty, crime and the environment, to name a few of the affected fields.

This bill would be too wacky to worry about, but its lunacy makes the other know-nothing bill look moderate. That bill, introduced in the House by Ted Poe of Texas and in the Senate by Rand Paul of Kentucky, targets the American Community Survey. Started in 2005 to replace the long-form census, the survey is the indispensable source of information on factors that define American life, including family configurations, education levels, work and living arrangements, income and insurance coverage. Credible information is the basis for a responsive government, an efficient economy and, by extension, a functional society. It also gives American policy makers and businesses a competitive edge, because it encourages decisions based on hard data as opposed to guesses or other faulty rationales that dominate in the absence of credible data.

About three million people receive the survey every year, and, as with the census, answering it is required by law. Mr. Poe and Mr. Rand want to make it voluntary, which would make the results less reliable, and potentially worthless, because fewer people would answer and those who did would not be a representative sample.

Canada recently replaced its mandatory long-form census with a voluntary survey — and now lives with the sorry results. To try to get an adequate level of response, the voluntary survey was sent to one in three Canadians instead of one in five, which increased costs. The response rate plunged anyway, from 94 percent to 68 percent. In a staggering one-fourth of Canadian communities, not enough people responded to make the data usable.

There are bets then there are bets -- developments in the rabbit ears war

Fox launched its terrestrial superstation, Movies!, on Memorial Day with the following line-up.

8:00AM / Take a Hard Ride

10:10AM / House of Bamboo

12:20PM / Backlash

2:25PM / Michael Shayne: Private Detective

4:10PM / The Man Who Wouldn't Die

5:45PM / Jumpin' Jack Flash

8:00PM / High Anxiety

10:05PM / Silent Movie

The channel is a collaboration with Weigel Broadcasting and there's definitely a Neal Sabin touch with lots of dog whistles for movie lovers (Goldsmith, Fuller, Sturges, Brooks), widescreen format and an emphasis on the pretty-good but hard-to-catch (Emperor of the North is playing muted as I write this). The channel also introduces the welcome practice of not editing for length.

Looking through the schedule for the next few days, I see a number of interesting titles, more than I'll have time to watch. I also see indications that a great deal thought went into scheduling, making NBC's COZI secure in its position as worst run terrestrial superstation.

When I rescanned my channels to pick up Movies!, four other new channels also showed up, including what appears to be a new subchannel from the CBS affiliate. At present, it simply rebroadcasts the same programming as the main channel, but if they really have gone from broadcasting one channel to broadcasting  two, it would seem to indicate that CBS is at least considering jumping into the terrestrial market (which would mean that all four of the big four networks would have terrestrial-only channels).

In the past week I also noticed an LA company that sells and installs outdoor antennas has launched a fairly sizable ad campaign. Along similar lines, I recently started seeing store windows in East and Central LA carrying a wide selection of motorized and/or amplified indoor antennas.

As we've mentioned before, there are two wildly divergent pictures of the state of over-the-air television: healthy and growing according to the 2012 Ownership Survey and Trend Report; or small and shrinking according to Nielsen. It appears that pretty much everybody with first hand knowledge and skin in the game (networks, broadcasters, regional media players, retailers, manufacturers) are putting their money on the first scenario, while the only prominent backers of Nielsen appear to be reporters for the New York Times and Reuters.

Noah Smith recently observed that trivial bets do not necessarily reveal beliefs. That's true, however the multimillion dollar ones do indicate a certain level of sincerity.

One more of these and I'm back to writing about old TV shows...

It's late, I hadn't planned to do another on this and I've got a logjam forming upstream, so I'm going to rush through this one.

Take a look at the following from the the Hartford Courant:

Rates Of Suspension Vary

The difference in rates of suspension between suburban and urban districts is substantial, with West Hartford and Farmington having five or fewer incidents of suspensions in this age group in 2012, while Bridgeport had 293 and Hartford had 238.

In some cases, the rates of suspension are quite different even between somewhat similar districts and schools.

For instance, Hartford has only about 240 more children enrolled in kindergarten and first grade than New Haven, but had 238 instances of suspension compared to 89 incidents in New Haven.

Amistad Academy in New Haven and Achievement First Hartford Academy are both public charter schools run by Achievement First, with very similar enrollment numbers in the early grades. But while Amistad had 38 instances of suspension during the last school year among children age 6 and younger, Achievement First Hartford Academy had 114 in the same age group.*

An even more dramatic comparison: The incidence of suspension of kindergartners and first graders at Achievement First Hartford Academy last year was an estimated nine times the rate in Hartford public schools.
...
Marc Michaelson, regional superintendent for Achievement First, said the school, where students annually out-perform their Hartford peers by significant margins on state standardized tests, has "a very high bar for the conduct of our students and that's because we've made a promise to our scholars and our families that we are going to prepare them for college."

And from another Courant article:

For the 2012-13 year, the state is spending about $67.7 million on 17 charters with 6,451 students, including Achievement First's network of schools in New Haven, Bridgeport and Hartford, said Jim Polites, a spokesman for the state Department of Education. Last fall, the state budget was $52.8 million for nearly 6,100 charter school students.

State Education Commissioner Stefan Pryor is a former board member and co-founder of Amistad Academy, Achievement First's flagship school in New Haven.

A few points:

1. There is a great deal of money changing hands, often with little scrutiny...

2. And despite decidedly mixed results, the money seems to be increasing.




3. Selection effects matter.

4. These effects can be amplified by social dynamics.

5. Add to that the natural extremes of childhood and you get certain kids who are 80/20 for experienced teachers and 95/5 for the newbies.

6. Thus even a small change in class roster can cause a big change in class behavior and performance.



7. Under movement reforms, one of the best ways to get ahead financially and politically, is to find a way to dump certain students.

8. If you repeated suspend five-year-olds, their parents will take them elsewhere.



9. The hardest-to-reach kids should be the ones you care about the most.

10. If you don't feel that way, you shouldn't be working with kids.



* In case I haven't mentioned it lately, there is a huge range in quality and professional ethics among different charters.

More on the looting phase of the education reform movement

Having watched the rise and fall (?) of the education reform movement for years now, I often find my attention split the story and the metastory. The story is what's being done in education, how it's working and what should be done differently. The metastory is about the way the reform narrative grew and entrenched itself (particularly with parts of the left) and about the way movement reformers (again, particularly on the left) failed to deal with certain longstanding concerns.

Putting aside the question of political agendas (which really shouldn't be considered when assessing arguments -- bad motives don't preclude good reasoning), the concerns fell into three broad categories:

1. The underlying pedagogical and economic assumptions were flawed;

2. There were intractable data issues that were being overlooked;

3. The system was vulnerable to (and had huge incentives for) gaming at a level that could undo any of the benefits that might come from the proposed reforms.

It is the third concern that prompted me to speculate about affinity effects (a point disputed here). Though evidence supporting all three has been accumulating for years, discussions of the first two got very technical very quickly. Support for the third, though, was obvious on a common sense level.

When a group repeatedly fails to notice something as it becomes increasingly obvious, you can legitimately start looking for some external influence, and the problem with gaming has become increasingly obvious. We started with clear vulnerabilities, then saw the rise of charismatic figures who asked us to trust them with tremendous amounts of money and power based on extravagant but questionable claims. Almost immediately after this, reports starting flowing in of suspicious test results, extensive student dumping and other signs of aggressive data manipulation.

How were so many otherwise alert and skeptical movement reformers caught off guard by problems anyone could have seen coming? I believe part of the answer lies in the culture and narratives of the reform movement. Cultural affinity meant that reformers (who generally had good intentions themselves) tended to project similar values on those who displayed the correct cultural signifiers while hero/villain narratives made it difficult to accept the idea that some of those cast in heroic role might be corrupt while some of those cast in the villainous parts might be right.

Though overt cheating has been getting most of the press lately, larger, more systemic data manipulation is potentially the bigger concern. Mike the Mad Biologist has been on this beat and sends us to this depressing data point.[emphasis added]

One of the first things a visitor sees when stepping into Kipp Academy is a graph that shows how Kipp is outperforming Metro schools in every subject.

However, Kipp Academy is also one of the leaders in another stat that is not something to crow about.

When it comes to the net loss of students this year, charter schools are the top eight losers of students.

In fact, the only schools that have net losses of 10 to 33 percent are charter schools.

"We look at that attrition. We keep an eye on it, and we actually think about how we can bring that back in line with where we've been historically," said Kipp Principal Randy Dowell.

Dowell said Kipp's 18 percent attrition is unacceptable.

MNPS feels it's unacceptable as well, because not only are they getting kids from charter schools, but they are also getting troubled kids and then getting them right before testing time.

"That's also a frustration for the zoned-school principals. They are getting clearly challenging kids back in their schools just prior to accountability testing," said MNPS Chief Operating Officer Fred Carr.

Nineteen of the last 20 children to leave Kipp Academy had multiple out-of-school suspensions. Eleven of the 19 are classified as special needs, and all of them took their TCAPs at Metro zoned schools, so their scores won't count against Kipp.

"We won't know how they perform until we receive results and we see. We would be happy to take their results, frankly. The goal is getting kids ready for college. The goal is not having shiny results for me or for anyone on the team," Dowell said.

There's a surreal quality to Dowell's responses here. The kind of attrition we're talking about in this case is almost entirely at the administrator's discretion. What's more, it generally takes a great deal of time and paperwork to make it happen.

As mentioned before, these schools already have selection bias and social norming working in their favor. If, on top of that, a principal like Dowell selectively gets rid of almost one out of five students based on behavior and performance, it is difficult for a school not to look good.

In a sane society, this would have been the lead story the day it broke

David Biello writing for Scientific American (via Thoma):

That's the message from the U.S. Geological Survey's evaluation of how the U.S. is managing its aquifers. Or mismanaging. For example: water levels in the aquifer that underlies the nation's bread basket have dropped in some places by as much as 160 feet.

The rest of the world isn't doing any better. A conference of water scientists just issued the so-called Bonn Declaration, which declares that this lack of foresight will cause the majority of people alive in 2050 to face "severe" freshwater shortages.

Ethics and Ayn Rand

James Kwak:

This idea is obviously echoed in Ayn Rand’s novels, which celebrate the individual genius standing out against the backdrop of collectivist mediocrity. It has also trickled into the contemporary conservative worship of the ultra-rich. The phrase today is “job creators” (whatever that means), but it has the same moralistic overtones as in Nietzsche and Hayek—a class of people who are better than the rest of us, on whom we depend for our salvation and prosperity, and whom we should not presume to question or constrain through, say, safety regulation or higher taxes (“penalizing success,” in the jargon).

This really has been one of the more insidious ideas of the twenty-first century-- that wealth and success are evidence of moral character.  One problem with this philosophy is that it replaces a creed of "it isn't whether you win or lose, it is how you play the game" with an unfortunate alignment between moral virtue and business success.  Under these conditions, how do you restrain the quest for wealth to within a code of ethics?

Instead we have a purely outcomes based ethic, that talks about the good that comes from the wealthy without asking hard questions -- like would things be better if wealth were more equal?  In particular, the logical connection between rich people becoming more rich and more employment opportunities being created are less than completely clear.  I don't want to say that the effect size is actually zero, but rather that it seems odd to neglect other possible sources of variation.

The Project Gutenberg Project

Joseph and I have been going back and forth on the best way to get the most out of the tidal wave of open data of which we are only seeing the beginning.

Joseph tends to be more skeptical on this subject. He almost has to be. He's approaching this both as an epidemiologist (where privacy and ethical issues create huge headaches) and an academician (where open data can create tremendous perverse incentives to rush out mediocre work in order to beat out other researchers looking at the same data). The promise of open data is very much field specific.

I tend to be more optimistic about the subject. I'm more the data miner of the blog and to find myself living in an age when anyone with a refurbished desktop computer, a copy of R and Python and a decent internet connection can do real, interesting research is tremendously exciting.

There is at least one area, though, where I am possibly more skeptical than Joseph and that's in the chances of these huge data initiatives self-organizing along anything near an optimal configuration...

I started to write something here about market forces in research and incentives and non-rival goods but then the phone rang and by the time I got off I realized that would be a lot of work (at least it would if I did enough research to make sure I wasn't making an idiot of myself). Chances are, that discussion would just be a long winded way of saying if we want to effectively coordinate all these researchers so that information flows where it needs to flow and data is fully explored and we can keep track of what's going on, we need to think this thing through.

Which brings me to the Project Gutenberg Project. Project Gutenberg has, of course, a huge and growing database. It's set up to be researcher-friendly and the system readily lends itself to automated approaches. The possibilities for text-mining are endless and a tremendous number of interesting research questions can be addressed with nothing more than a reasonably up-to-date computer and some free software (I previously posted a couple of examples here).

This would seem to be an ideal test case for setting up procedures and sites for dealing with large, open databases . Here are a few possibilities:

A place to submit and comment on proposed hypotheses;

A place to report preliminary findings;

A place to report negative findings;

A place to report confirmation of previous findings;

A database connecting approaches, hypotheses and data points;

Multiple ranking systems;

A way of identifying under-explored parts of the data.

Obviously this is a first pass and I'm just throwing out some ideas. Some might be impractical. Others, as Joseph would point out, will not be applicable to other data sets. And I have a nagging feeling that I've left something obvious out.

But that, of course, is the nature of a blog post.

Maybe they're hinting at something

I've been experimenting with different blog formats over the past few years with a couple of pop culture blogs and a fairly serious one (You do the math) on education. They've definitely been on the back burner recently but they are active and they continue to improve slowly in traffic and page rank so if I decide to put one in gear (which I may do with the You Do the Math fairly soon), I won't have to start from zero.

These sites run ads. To date I haven't seen any money from them (my account is around forty dollars and you apparently have to have one hundred for them to cut you a check), but like I said they've been in a groundwork stage and I have plans to go after at least one of them more aggressively in the future.

Which is why this email from Google is a tad troubling:

Hello,

As part of our internal audit review, we identified a credit under your AdSense account that may be due and owing to you.

Since the date of your last transaction, there has been no activity from your account. As such, these funds are in danger of being escheated to the state. Unless we are informed of your ownership or beneficial interest in these funds, they will be placed in the custody of the State of California. If these funds are transferred to the State, you will be required to submit a claim in accordance with the statutory provisions of the State's unclaimed property laws in order to recover it.

Should you wish to discontinue using AdSense, you can close your AdSense account. If you close your account, you'll no longer have access to the Google code and you'll receive no further emails from us. Your remaining earnings will be paid out according to the payment schedule outlined in the AdSense Terms and Conditions. Provided there are no holds, you have accrued more than 10 USD and we have a valid form of payment on file, you should receive your payment in the cycle following your account closure. For detailed instructions on how to cancel your account, click here.



Sincerely,

The Google AdSense Team

Here's the funny part... There doesn't seem to be any way for me inform them of my ownership other than to click the link and cancel. There is no link in the middle paragraph. The email was sent from a no-reply address. When I go to the Adsense contact page I am told that my account is not eligible for direct email consultation and I am limited to "specialized contact forms," none of which seem to come even close to fitting my situation.

Are they trying to get me to leave or are they really this bad at business communications?

Weekend blogging -- Uncle Art's Funland

Art Nugent doesn't have the mathematical following that Dudney or Loyd (he seldom hit the depths of those two), but he was a pretty good puzzle maker and a damned fine cartoonist and his output was extraordinary. For more than forty years he put out newspaper features and comics pages filled with puzzles, games, riddles, activities and magic tricks.

From Just a Pile of Old Comics.

Spinach is supposed to make you stronger -- Infrastructure and the Kinsley/Krugman fight

Virtually everyone who reads this blog has heard about (and is probably sick of hearing about) Michael Kinsley's contrarian defense of austerians that ended with this analogy:

Austerians don’t get off on other people’s suffering. They, for the most part, honestly believe that theirs is the quickest way through the suffering. They may be right or they may be wrong. When Krugman says he’s only worried about “premature” fiscal discipline, it becomes largely a question of emphasis anyway. But the austerians deserve credit: They at least are talking about the spinach, while the Krugmanites are only talking about dessert.

To get a feel for just how odd this analogy is, you need to remember that a large part of this 'spinach' is saying no to people who want to borrow money almost interest-free and spend it on infrastructure, education and research thus avoiding far greater costs in the future.

These are all urgent issues but the infrastructure crisis in particular demands immediate action. Civil engineers have been ringing this alarm bell for years:

Back in March, when the American Society of Civil Engineers issued an infrastructure report card for the entire country, its very best grade — a B-minus — went to solid-waste disposal. Thanks to our decent progress in recycling, the United States’ overall grade-point average in subjects ranging from aviation to water systems actually ticked up from the previous GPA.

To a pitiful 1.30, that is, on a 4.00 scale.

Those warnings became considerably less abstract yesterday:

SEATTLE — A large section of a bridge on Interstate 5 north of Seattle collapsed Thursday evening, sending vehicles and people plunging into the swirling, frigid waters of the Skagit River.
Three people were hospitalized in stable condition, officials said. No one was killed.

The bridge failed without warning between the towns of Burlington and Mount Vernon on the major route linking Seattle with the Canadian border, the Washington State Patrol said.

"Without warning" here is a bit of a relative term:

The 58-year-old bridge in Washington, a crucial link to the Canadian border traveled heavily by trucks, was inspected every two years, most recently in November, state Department of Transportation spokesman Bart Treece told the Los Angeles Times.

“It’s an old bridge. We have to look into the specifics. We do have a lot of old, aging structures, and a lot of them hold up really well,” he said.

The National Bridge Inventory lists the bridge as “functionally obsolete,” with “somewhat better than minimum adequacy to tolerate being left in place as is.” It received a sufficiency rating of 57.4 out of 100.

Putting aside for the moment the question of public safety, the economic impact of bad infrastructure can be huge (from the same story):

Washington’s main north-south thoroughfare, though, was likely to remain closed 60 miles north of Seattle for an indefinite period, state officials warned. The nearly 71,000 vehicles a day that travel the bridge between Mt. Vernon and Burlington were diverted through city streets to another nearby bridge.

Just to be clear:

dessert = making repairs now

spinach = deferring repairs now and making more costly ones later when interest rates will be higher

Kinsley's analogies are like America's bridges; they need a lot of work.

Notes on an unwritten paper -- Naive Bayesian Classifiers and Order of Composition

[Update: I've got some more thoughts on Gutenberg-based research in my latest post.]

I'm planning on writing some posts on the potential of and the potential concerns about open data (possibly even getting Joseph to join in) so I thought I'd dust off a somewhat relevant idea I had a few years back. If anyone wants to see if they can get something publishable out of this, feel free. In the meantime, I plan on getting some mileage out of it as an example.

A few years ago, I wrote some code for text mining. It was really basic, standard stuff -- using naive Bayesian classifiers and n-grams (normally techniques for assigning authorship) -- but it worked well and was fun to play around with. I used various books from Project Gutenberg as test data and selected authors with styles and backgrounds ranging from close (Dickens and Trollope) to out there (Thorstein Veblen) with a translation of Verne as someone neutral. The two Victorians also had the advantage of having written lots of books over many years.

The idea was to approach this less as a classification problem and more of a question of distance between points in a literary space. Here the "likelihood score" was more a measure of similarity. As you would expect, Great Expectations was more similar to Nicholas Nickleby than to Barchester Towers, more similar to Barchester Towers than to a translated Master of the World and more similar to Master of the World than to Theory of the Leisure Class. It also worked as expected when you compared works of the same author written at different points in his career: Great Expectations (1860 to 1861) was more similar to Our Mutual Friend (1864 to 1865) than to Nicholas Nickleby (1838 to 1839).

Obviously this was a tiny trial run, but it did suggest that there's something out there, as did a recent literature search which turned up at least one related paper from 2011 ("Predicting the Date of Authorship of Historical Texts" by A. Tausz) which used NBCs to determine absolute rather than relative dates. Still even with Tausz' paper (which is very interesting, by the way) there still should be room for research into intra-author questions and, more importantly, into lots of other questions using data from project Gutenberg.

And on top of that you can apparently find interesting stuff to read at the site as well.

Scandal, metareporting and the dumber-reader theory

Everybody has heard of the greater fool theory of investing where you buy a stock not because you think its assets are undervalued or because there's a good chance that the company will make money but because you believe there is someone out there who will pay significantly more than you paid.

I've noticed a somewhat analogous trend in journalism today, particularly involving the coverage 'scandals.' I apologize for the quotes but they're there for a reason I'll get to in a bit. In the traditional model of reporting, the journalist implicitly claims that the information being reported is accurate, representative and significant enough to justify the readers' time.

Over the past few years, though, journalists seem to have gotten more likely to downplay these traditional elements (what we might call the fundamental value of the story) and focus on what the impact of the story will be if people other than the reader believe it (the dumber-reader theory). In generic form, the stories go something like this: "A made accusations against B. There is no reason to believe these accusations but if they gain traction, they could hurt B."

Perhaps the most dramatic example of the past few years was swiftboating where most of the attention was paid to how Kerry's handling of the charges would affect his campaign while relatively little was given to the charges' validity (a question that in previous times would probably have been considered a necessary condition for the story to advance).

Don't get me wrong. Coverage has always included questions about the impact of scandals, but it seems like the process before had more of a tree structure: ask question A and then, based on the answer, ask either question B or C. I'm not saying that this rose to the level of hard and fast rule, just that it was the norm. First you asked if an accusation was true. If the answer was yes you asked how serious was the offense; if the answer was no you asked if the accuser had been deliberately misleading. And so on...

I can see how moving away from that structure is a good thing for journalists. For the rest of us, however, it does not look like a good thing at all.

Pets

Frances Woolley:

Indeed, when a person selects a pet, life expectancy is one of the last things considered (see, for example, this pet selection guide, or this one or this one). Instead, "experts" recommend choosing a pet who will be a good match for his or her owner in terms of activity level, sociability, and so on. Good health matters - sensible owners avoid breeds prone to health problems. But not life expectancy per se.

I think there is a good point here -- life expectancy is not the only good that people are interested in.  Sure, I do not want to die young.  But if terrible quality of life was the only way to extend one's life span that would seem sub-optimal too. 

Really what we want to maximize is high quality life.  In cases where high quality and life contradict each other then one has to choose (and it is never an easy decision).  So it is not surprising that people adopt pets that they are compatible with.  But just ask a dog owner what they will do to extend the life of a sick Labrador Retriever and you might be surprised . . .

So which preference is dominant?  The breed decision or the attempt to prolong the life of one's furry friend? 

Un-self-awareness at the New Republic -- more Rhee-views

Michele Rhee has popped up in a couple of notable posts this week.

First Nicholas Lemann writing in TNR:

Rhee is not one for exquisite sensitivity. She closed schools, fired teachers, and (though she assures us that “I had never sought the limelight”) became famous. She was on the covers of Time (holding a broom) and Newsweek, and was one of the stars of Waiting for Superman. It is usually a fundamental rule of politics that a department head isn’t supposed to do anything to make her boss unpopular or to upstage him. Rhee did not follow this rule. She has a special scorn for “politics” and often praises Fenty for not considering it when making decisions, but this is both un-self-aware (Rhee’s policies were very good politics in white Washington) and impractical. We live in a democracy, so officials have to contend with public opinion and with groups organized to promote their own interests. Many American politicians over the last generation, including all of the last five presidents, have been able to push education policies in the same realm as Rhee’s in a way that kept their coalitions together. That is what Rhee and Fenty were unusually bad at doing, and Rhee’s insistence that “politics” is a terrible thing that only her opponents practice was surely a big part of the reason why.

Lemann represents the pivot phase of the press's relationship with Rhee, not quite ready to address painful topics, but moving away from the hagiography that until recently marked much of Rhee's coverage (particularly at the New Republic).

It is, of course, difficult to pivot gracefully so it's not surprising to see an awkward turn or two here. Most memorable is the description of Rhee's stated view of politics as "un-self-aware." Not only is the term itself good for a chuckle, but by this criterion Rhee's un-self-awareness would apply to the vast majority of politicians (think of all the times you've heard candidate express a similar scorn). It is the most standard of standard campaign lies, made all the more transparent by Rhee's relentless and ruthless political maneuvers (including reaching her current position by climbing over the still-warm corpse of Adrian Fenty's career).

The press is slowly coming to terms with how badly it was played by Rhee, but they are getting there. The question now is how will the fall of one reformer affect the movement? Andrew Gelman sees this as indicative of something bigger.

My impression is that there has been a shift. A few years ago, value-added assessment etc was considered the technocratic way to go, with opponents being a bunch of Luddite dead-enders. Now, though, the whole system is falling apart. We can learn a lot from tests, no doubt about that, but there’s a lot less sense that they should be used to directly evaluate teachers. We’ve moved to a more modern, quality-control perspective in which the goal is to learn and improve the system, not to reward or punish individual workers.

This shift may have not happened yet at the political level, but it’s my sense that this is the direction that things are going. The Rhee story is symbolic of the fallacies of measurement.

I'm not so sure. From the beginning there have been at least four major concerns:

1. Given the ugly nature of the data (confounded, nested, etc.) we would not get anything usable out of the two or three years of data window we would have to evaluate a teacher;

2. The test might give us an inaccurate or incomplete picture of what students were learning;

3. The system would be vulnerable to cheaters;

4. The tests would distort education priorities.

Rhee's crash drives home 3 but I don't know that it says that much about the rest which is troublesome because those are the ones that bother me more.

A smart post from Felix Salmon

Felix Salmon:

My point here is that technology has a tendency to create its own norms. The classic example is the automobile — a technology which kills more than 30,000 Americans every year. From the 1930s through the 1990s, societal norms about who roads belonged to, and what people should do on them, were turned on their head thanks to the new technology. The dangerous new activity allowed by the new technology became the privileged norm, to the point at which just about all other road-based activity — and roads have been around for thousands of years, remember, since long before the automobile — essentially ceased to exist. Eventually, we reached the point at which elected representatives were happy saying that if a bicyclist gets killed by a car, it’s the bicyclist’s fault for being on the road in the first place

.I think that this is a very interesting point at two levels.  One, is that it does point out that society can change around innovation just as much as innovation can change society.  I think that this will be broadly applicable to innovations like driverless cars that are legal nightmares now, but could easy become the standard with enough adoption.  It's never clear when a technology will win this sort of breakthrough success (the innovation grave-yard is full of such examples).  But it does point out that some classes of argument are less likely to succeed.

But the second point is also really salient -- it is often amazing how much we overlook the subsidization of activities are social norms.  We don't see the use of roads for cars and not bicycles as a subsidization of the car.  Heck, I am often annoyed by bicyclists who can't decide what set of rules they are following (when they switch back and forth between being a fellow vehicle and a pedestrian it makes me nervous as I have a life-long goal to never hit a cyclist).  But the roads could just as easily be claimed by walkers, horses, bicycles and so forth in a much easier form of mixed use.