Did you catch this story on Drudge? Over 4 days in Phoenix, Arizona 4 boys, between 3 and 5 years old, intentionally set their homes on fire. That's weird enough, but what is weirder is that, according to the story, "All incidents were unrelated." The suggestion? "[Phoenix Assistant Fire Chief Bob] Khan recommends that parents enroll their children in the Youth Firesetter Intervention Program, which teaches youths about fire safety and the consequences of playing with fire."

I am most struck by the futility of the suggestion: These fires weren't accidental apparently, so a safety lecture will hardly help. But more to the point, what has happened when telling kids not to play with fire is now a municipally sponsored, centralized class in which one enrolls as opposed to something parents teach their kids? I also wonder if there are statistics available that can shed light on the success of the class in preventing fires (even suggestive statistics from before the class was offered compared to now.)

What's Belgian for "ant"?
This ant colony is completely creepy. It is a collection of millions (!) of nests that usually kill one another, but they are cooperating. It is sort of like the EU, minus the declarations of maximum permissible slope of anthill walls.

Joanne Jacobs links to a Heritage Foundation summary of a study done which purports to show that vouchers improved the test scores of low income black students. My first thought was that the student's whose parents had the gumption to get them the vouchers would of course have an advantage, but the summary indicates that the students who received vouchers and the control group to which they were compared were chosen at random from those applications which were eligible for participation. Parental gumption, then, should not be a factor favoring those with vouchers. (Though it may be more subtle: It may be that parental gumption makes vouchers successful (the vouchers would act, then, like a lever or pulley to multiply the effect of parental effort) and so the results would not be generally applicable. But that sort of subtle relationship is not something that the researchers can be expected to rule out, and is something for the study's critics (if there are any, har har) to show, if they can.)

Thanks to Instapundit for the pointer to this exchange of letters between a Danish customer who feels he cannot do business with an Israeli company, "when the Israeli militaire (sic) is behaving so rough (sic) in the Palestinian Areas," and the Israeli company with which they were doing business, Radix. If you want to express disappointment or support to Jens Peter Hansen (guess which one he is: Danish or Israeli?), you may want to drop him an email.

Kimberly Swygert over at Number 2 Pencil has posted a thoughtful response to a laundry list of complaints from the American Evaluation Association regarding high-stakes standardized testing. Her responses are largely on target, and her task is thankless and grueling, since the original list of complaints is an unruly set of attacks without any argument behind them. At a couple of points, however, she makes her point less forcefully than she might. Let me see if I can supplement them helpfully.

From her blog:

AEA statement: “[They] assume that all children, including English language learners and special education students, learn in the same ways at the same rate and that they can all demonstrate their achievements on standardized tests"
My reply: I wouldn't say that testing programs assume this so much as measure to what extent this is true, and that's useful knowledge to have.

I think this gives too much to the AEA. Ms. Swygert is right that testing programs do not assume what the AEA statement says they do. But she is also wrong to say that they measure to what extent this is true. What they do, instead, is test how well our students are learning, in whatever way and at whatever rate they learn. They are useful, as a result, in measuring the success of schools because it is the job of those schools to teach the students what they need to learn. It is the school's job, not the test's, to take into account the different ways in which and rates at which the students learn. That is, the ways and rates of student learning are the concern of those from whom they are learning, not of those who want merely to know what they've learned.

And one can feel that Ms. Swygert's answer is even less satisfactory regarding the test's treatment of the student's ability to demonstrate what he knows on a standardized test. There, one might say, the AEA really has Ms. Swygert: The test measures what the student knows, one might think, only on the assumption that what the student knows is reflected in the number of right answers. But some students, it seems, might have excellent skills and knowledge but for whatever reason express it poorly in the standardized test format. And to the test-makers, such students would be indistinguishable from students who didn't have the skills or knowledge the test was seeking to measure.

It is a smart objection, to be sure, but psychometricians are smarter. The ability to demonstrate your knowledge in a standardized test is not some unmeasurable unknown that simply must be assumed not to exist. Rather, it is possible to measure such an ability by varying the difficulty of the material and tracking the difference in scores. How can you tell if someone has a difficulty with the material or the testing style? Give him the same kind of test with easier material. Do that enough and you can measure the importance of the testing style to the final score.

Take an extreme example. Imagine an ordinary SAT, but instead of filling in bubbles to indicate your choice of A, B, C, D, or E, you are to sing the note of the letter to mark your choice. This is a test where most people would do quite badly. How could we tell whether or not they knew the material? We would notice that, no matter how easy we made the material, their scores would not improve, and that is a dead giveaway.

This is not the case with filling in little circles. Just about everyone has the same smooth relationship between their scores and the difficult of the material. Perhaps Ms. Swygert can provide references and pointers, but the point is that this is not some unknown which is simply assumed not to effect the process. It is known, measured, negligible, and better than the test effects of just about any other method of assessment.

AEA statement: “Tests focus attention on particular students, such as those scoring just below the cut-off score and ignoring those who score well above and below the cut off score”
My reply: This comment I don’t quite get. By “focus attention”, do they mean the tests do a better job of measuring those in the middle? These types of tests usually are more accurate in the middle range. Scholastic achievement often has a normal distribution, with many kids in that middle range, so the test items tend to cluster around that achievement level. Or, do they mean that schools pay attention to or give feedback only to those in the middle? I don’t see how, since a child’s performance at the 95th percentile or the 5th percentile is what is going to garner a lot of attention from the teacher, or the school, if the tests are part of accountability.

I think the point of the AEA criticism is that what matters for schools is "pass rates," and improvement in "pass rates." And that is why we tend to focus our attention on those students who are close to passing, but have not passed. They might be pushed into the passing range by reasonable effort. The students far above and far below aren't going to make the difference; the former won't fail and the latter can't be made to pass. It may well be that concentrating on those students who can be made to learn the skills they need but have not yet done so is the wisest application of limited educational resources. In which case, the effects of the tests on our "focus" would be for the good.

Regardless of that, however, notice it isn't the test that bothers AEA, but the fixation on pass rates. But that isn't a feature of standardized testing. It isn't even typical. The SAT, LSAT, GMAT, GRE, etc., etc., are not tests one can pass or fail. Even in those tests which do have such a score, we could hold the schools accountable not to passing a greater percentage, but to raising the median score, or to raising the median score and shrinking the distance between that score and the 10th percentile score, or anything else. So the AEA's complaint is not really with testing, here. Notice also what the AEA would opt for in the alternative: teachers' ignorance of their success so far in teaching students what they need to know.

Then there is the kicker:

AEA statement: “[They] measure, for the most part, parental income and race, and therefore perpetuate racism, classism and anti-working class sentiment"

It is an interesting question when one would say that the tests measure those things and when one would say that the tests measure something correlated with those things. Another interesting question is: Why would it be that these "measure, for the most part, parental income and race"? Is it because of questions on the SAT about what kind mileage a yacht gets? Is it because of questions on the GRE about what kind of wood burns best in a cross? When questions like that are found, where some plausible case can be made for bias, those questions are removed.

Perhaps these tests "measure, for the most part parental income and race" because they measure, for the entire part, what the students have learned, and that is correlated, for the most part, with educational opportunities correlated to some degree with parental income and race. Is that something the AEA folks wouldn't want to know? Wouldn't they want to know who was falling behind, so that we could concentrate our effort on them?

Of course, the real "racism, classism, and anti-working class sentiment" comes from the folks who oppose testing on these grounds. Why wouldn't they want to concentrate our attention on those who need the help, whatever their parental income or race? It is because people like those at AEA don't think it will do much good. The poor and the black and brown, they think, form an underclass which is more or less permanent. And all that measuring the amount of work that needs to be done can do is more deeply entrench class- and race-hatred. The best thing these folks think we can do is ignore and cover over the differences that exist, for surely there is nothing we can do about them. That's a hard and bitter view that I (and I suspect Ms. Swygert) do not share.

Drudge refers to this story that the bomb threat that closed a downtown Washington branch of Bank of America (and, according to Drudge earlier in the day, other Washington banks) was made by a Dutch kid. So given that "[t]he FBI's Washington field office said in a statement that it did not have any reason to assign a "high degree of credibility" to a threat that was received via telephone on Sunday from the Netherlands routed through a Canadian telephone operator," this kid does not seem to have taken any (successful) steps to make this a "successful" hoax. Does this mean that the US is vulnerable to this sort of call at any time from any kid in the world? What kind of sorting and screening procedures are there? And, most importantly, how can the US protect itself from the delivery of pizzas it didn't order?

I just got around to reading Netanyahu's April 10th speech to the Senate. Classic Bibi. True, clear, and powerful, it is masterful both as rhetoric and as analysis. Don't miss it.

Also from Steyn: "The only question now is whether the US is a member of the Kofi set in good standing or whether it's a member mainly in the sense that Saudi Arabia is a member of the coalition against terror." This formulation, with the supporting chronology and priceless quotation from Colin Powell ("A week ago, asked to define what Washington meant by Israeli withdrawal "without delay", Colin Powell replied that the Administration "does expect something to happen soon with respect to bringing this operation to some culminating point where you can start to see a movement in the other direction"") gives me some hope that we are no worse than Saudi Arabia is good. And that ain't bad.

Mark Steyn is, as usual, terrific. "Of the 30 ongoing conflicts in the world today, the Muslims are involved in 28 of them." If "Islam" didn't mean "peace," I betcha it'd be 30 out of 30!

Thanks to Little Green Footballs for the pointer to the fine Boston Globe article by Jeff Jacoby, Let Israel Fight Its Way to Peace. The ideas of the article are two. One, Israel is at the front lines of the war on terrorism, taking strong (and internationally unpopular) action that advances the interests of the civilized world in the long run (the key comparison, which I have heard too little of lately, is with Osirak.) Two, the Israeli-Palestinian conflict is a conflict between good and evil. Both of these are true. And they show, respectively, why the US is being foolish and immoral in its opposition to strong Israeli action. What is astonishing is that Jacoby's is a minority view.

Jonathan Rauch's piece in The Atlantic, about computer construction of artificial societies and the laws and structures that such societies seem inescapably to express, is well worth the read.

As Glenn Reynolds over at InstaPundit claims, the departure of Cornel West from Harvard is a big win for Harvard. It is in the same way a big win for Larry Summers, though I do not know enough about the man to conclude whether this was something he wanted, or even if it was, if it was something he pursued intentionally. It may have just been the natural byproduct of better thinking at the Harvard helm.

George Will's column today is a real gem. Go read it. If you still favor the tarriffs, go read Bastiat's classic essay. If you still favor the tarriffs, go read Hazlitt's terrific book. If you still favor the tarriffs, there is no hope for you.

The main thing to think about the little drama going on in Israel of the recent bus bombing, the postponement of Powell's meeting with Arafat, the latter's "denunciation" of terrorism, and the word that the meeting will go ahead tomorrow, is that it is an absurd farce. What, was Powell surprised that a bomb dared to go off when he was there? Are we to think that he was outraged, or American policy was effected, in a special way by the latest bombing? And what are we to make of a country (or a Secretary of State) who still, still!, is moved in any way whatever by this man's words? Can anyone seriously think that his statement, produced under these circumstances, means anything whatever? If this little drama reflects anything genuine in US policy or in this administration, I am frightened. I am also bewildered about how I could have been so wrong in believing this administration to possess both moral clarity and political realism. If it doesn't reflect anything genuine, then I am puzzled by what purpose it could serve to play this role in this tired immorality play.

Debka.com reports that Powell has had his eyes opened by an Israeli briefings regarding the Isreal-Lebanon border, and that this has become the focus of his visit. Perhaps he should have read Charles Krauthammer's fascinating, terrifying piece about the border situation and what it means.