Thursday, December 13, 2007

Clemens, Cheater

Since today seems to be non-education-related blog post day, let's just roll with that a little longer and discuss the fact that Roger Clemens apparently began using steroids in 1998.

If true, Clemens deserves all the scorn currently being heaped on Barry Bonds. They both did exactly the same thing, at exactly the same time in their careers, for exactly the same reason, with exactly the same result. The prima facie evidence of steroid abuse, in terms of freakishly prolonged greatness, was just as strong. Instead of natural decline in their mid-to-late 30s, each began to rack up additional MVPs and Cy Youngs. Instead of merely being the best position player and pitcher of their generation, they cheated their way into becoming among the greats of all time. Clemens' 354 career wins is, in the context of the modern five man rotation, nearly as impressive as Bonds' all-time homerun record. If Clemens goes to the Hall of Fame and Bonds doesn't, that's racism pure and simple. And Dan Duquette, the Boston Red Sox general manager who infamously said that Clemens was in the "“twilight of his career" before letting him go to the Blue Jays, is owed a lot of apologies right about now.

NCLB Face-Off

The CATO Institute’s Andrew Coulson and I face-off at the Reason Foundation’s Reason Roundtable over whether NCLB should be scrapped or mended.

Coulson argues that NCLB is a federal intrusion on states’ rights and blocks the way for more meaningful, market-based education reforms. I argue that NCLB is critical to getting the kind of information on school performance we need in a market-based system, and that it can help build the supply of quality schools—a first step to any true, equitable school choice market.

Check it out here, and better yet, weigh in with your own comments.

Wyclef for President

I went to the annual RIAA holiday party last night at Ibiza, headlined by Wyclef Jean. Whatever Wyclef gets paid for shows like these, he's worth it, particularly when the job is to engage a crowd full of rythmically challenged Hill staffers and DC types. At one point he invited all the women in the audience to come on stage and dance, and then just let them stay there for the rest of the show. Why don't all concerts work this way?

Getting to the show took me through the new New York Avenue Metro stop and involved walking past the new headquarters of the ATF. Except its not really the ATF, it's the ATFE, as in "Alchohol, Tobacco, Firearms, and Explosives." Apparently they added the "E" during the whole Homeland Security reorganization, but decided to stick with that magical ATF brand so as not to confuse people. The building itself was obviously constructed with an acute awarness of the "E" problem; the actual structure is set back a good 50 yards from the street behind large barriers semi-concealed by landscaping etc. One suspects this the future of governmental architecture in the nation's capital, which is understandable but also a shame.

The Laws of Men

In the past few months I've seen numerous citations of Campbell's Law:


"The more any quantitative social indicator is used for social decisionmaking, the more subject it will be to corruption pressures and the more apt it will be to distort and corrupt the social processes it is intended to monitor."
Given current debates about high-stakes testing, NCLB, higher ed accountability, etc. the relevance to education policy is pretty obvious. David Berliner wrote a whole book flowing from this premise.

Campbell was a social scientist who, in declaring an eponymous, universal statute of sorts, joined a long tradition. The trick is to notice some interesting and fundamental relationship between important things, and explain it in a way that's memorable and easy to understand. Sometimes the observation is explicitly framed as a law, e.g. Godwin's Law: "As an online discussion grows longer, the probability of a comparison involving Nazis or Hitler approaches one."

In other cases the observation is so inextricably tied with the observer that the effect is roughly the same, as with Acton on power and corruption or Santayana on remembrance and repitition of history. Getting your name attached to one these things is one road to minor immortality, albeit a particularly reductive kind.

The form is usually pretty similar: As X, Y. Or in slightly different form: As X, not Y. The historical examples people are trying to emulate, I suspect, are the ageless mathematicians and philosophers--Pythagoras, Archimedes, et al. Newton seemed to exlain the entire universe in three short laws of motion. That's why these laws are so popular. The world is a complicated place, more so all the time, and people are always hungry to accumulate a set of inviolable principles with which to make sense of things.

But here's the thing: just because someone makes an observation and calls it a law doesn't mean it's always true. I've heard people refer to Campbell's Law as if it were etched into the marble facade of the Supreme Court of Social Science, right up there with "correlation doesn't imply causation." It's not, nor are all the rest. Lots of people become powerful while avoiding corruption and make new history while remaining ignorant of the old. Heck, even Newton was eventually overtaken by quantum physics.

It'd be comforting if we could ascertain the world with nothing more than a few nostrums and easily remembered laws, but that's not the world we live in.

Wednesday, December 12, 2007

In Defense of College Rankings

I've spent a fair amount of time over the last year having various conversations and arguments about college rankings, and one of the problems with the discussion is a tendency to intermingle critiques of rankings per se and critiques of specific rankings, e.g. those produced by U.S. News. For example, people often say things like "Rankings are reductive, overly-simplistic and create perverse incentives for colleges to engage in an arms race for money, status, and student." While the first critiques--reductiveness and over-simplification--apply to all rankings, the perversity of the incentives is wholly a function of what you happen to base your rankings on. If you rank colleges based on good measures, then the incentives could be constructive.

In other words, there's nothing wrong with universities trying to climb the rankings ladder if the ladder leads to the right place. An 2006 Education Sector report explaining how to build such a construct is here.

As a rule, generalized anti-rankings arguments are a lot weaker than those levelled at specific rankings like U.S. News, because the arguments tend to stem from a broader aversion to accountability, competition, and public scrutiny. To read more on this topic, see this paper, "In Defense of College Rankings," which I presented last month at the Association for the Study of Higher Education annual conference.

Also, for a lively ongoing discussion of college rankings, check out Morse Code, a blog written by Bob Morse, the U.S. News rankings guru. Even though U.S. News has an obvious interest in this debate, the blog is actually quite thoughtful and non-propagandizing.

Tuesday, December 11, 2007

Winerip on Poverty, Etc.

Micheal Winerip covers a new ETS report in the Times today, exploring the relationship between out-of-school factors like single parenthood, TV watching, reading at home, etc. and student achievement. The report--which I have no quarrel with, Education Sector co-director Andy Rotherham was a reviewer--finds, to the surprise of no one, that these things make a difference, in the way that one would expect they would.

Therefore, what?

Because this is the question that matters. Those of us who work on education policy for a living have endured listened to a seemingly endless series of arguments around these issues in recent years, where one side says "of course schools are important" and the other side says "of course poverty (or whatever other non-school issue interests them) is important," as if these are debatable questions. So let's state, unequivocally and for the record, that educational outcomes are significantly influenced by things that happen both in school and out of school. Anyone who believes otherwise is an idiot, and anyone who implies that someone else believes otherwise is arguing in bad faith.

Therefore, what?

If you're not willing to answer this question concretely, you really doesn't deserve a seat at the table. Winerip is clearly not up to the challenge. He says "What’s interesting about the report...is how much we know, how often government policy and parental behavior does not reflect that knowledge, and how stacked the odds are against so many children." This--"reflects"--is typical; variants include "recognizes," "acknowledges," "takes into account," etc.

But what would government policy that "reflects" knowledge of the ETS report look like? Winerip appears to have no interest in answering this question, and that makes the statement itself essentially meaningless, since it's clearly written in the context of a policy debate.

NCLB, by contrast, reflects an identifiable perspective and set of resulting policy conclusions that goes something like this: Poverty (I'll use this as a proxy for all outside factors, since it's the issue that comes up the most) matters, but estimates of how much it matters are often overstated, because they don't fully account for two things:

1) The extent to which impoverished students get fewer educational resources like money, highly effective teachers, challenging curricula, etc. This underestimation is not function of dishonest or sloppy research, it's a function of the fact many of these things are hard to fully quantify.

2) The extent to which overall educational quality is sub-optimal, a problem that disproportionately affects low-income, low-achieving students, since they're more sensitive than higher-income, higher-achieving students to differences in educational quality even if those differences are evenly distributed (which of course they're not).

NCLB supporters believe, therefore, that poverty-based barriers to achievement are surmountable, provided that we (A) give poor students more educational resources instead of less, and (B) define "surmountable" in terms of fixed goals, not relative performance. In other words, nobody believes (Richard Rothstein's frequent assertions to the contrary) that schools can, by themselves, make the achievement of poor children indistinguishable from non-poor children, in the sense that they would be equal in all ways. Rather, people believe schools, properly resourced and run, can do enough to help poor children learn essential knowledge and skills. Therefore, it's reasonable to hold schools accountable for that goal. Ergo, NCLB.

The logic is certainly debatable--in the sense that reasonable people can debate it in good faith--and the supporting data is far from conclusive. But it's backed up by significant real-world evidence of the efficacy of schools in general and of some high-poverty schools in particular. And the policy implications are right there on the table, which is what matters most in the end.

On the other hand, Winerip and those like him who have devoted years of their lives to mounting counter-arguments against the current accountability regime are unified in their resolute unwillingness to explain what actual policy conclusions we are meant to draw from their ideas. Again I ask: If not this, what? The NEA and AFT have at least produced some fairly detailed outlines of principles for a different federal law. But when God and the devil are in the details, that's not good enough.

While NCLB in totality is lengthy, the parts that generate most of the controversy are actually pretty short and written in plain English. If you don't like them, fine--serious people can disagree. But if you're not willing to say how they should be different and take responsibility for that position, then you're not, in this debate, a serious person.

Update: Side Effects May Vary, which seems to come at these issues from a libertarian / voucher perspective, complains that post above "seems to imply that there is simply no solution that does not involve the state." So let me clarify: That's not what I'm saying. I'm saying if you don't like the current law, be specific about what you want instead. The author says children are "languishing in a cruel trap guarded jealously by social planners like Carey and his ilk who venerate the contraption while condemning those of us who wish to extricate their unwilling subjects." Sure, okay. "Extricate"? What does that mean? Universal vouchers? I can't tell. Enlighten us.

Race and IQ

Malcolm Gladwell’s latest piece forays into the IQ-race debate most recently scandalized by the declaration by James Watson, co-winner of the 1953 Nobel Prize for the discovery of DNA, that Watson is “inherently gloomy about the prospect of Africa” because of Africans’ general lack of intelligence. Gladwell’s article reviews a new book by James Flynn that shows Watson’s comments are almost certainly wrong.

Gladwell starts with a discussion of the Flynn effect, which has documented a steady but gradual rise in IQ scores over time. In order to compensate for the observed 3-point rise every decade, the main measurement instrument, the Wechsler Intelligence Scale for Children, has been updated four times to ensure that a score of 100 still equals the absolute average of peers. Without a process called “norming,” the average rises over time. And a score of 70 on the fourth version of the WISC would be much higher on the original version. This matters especially when considering arbitrary cutoff scores for admission to special education or gifted programs. The article cites a study on the 1991 introduction of the WISC III that found that the revision doubled the number of American children labeled mentally retarded if all states followed strict IQ score cutoffs and adopted the new test.

Flynn has also compiled a host of studies showing the effects of culture on IQ. Studies of mixed marriages offer Gladwell a compelling point,
If I.Q. is innate, it shouldn’t make a difference whether it’s a mixed-race child’s mother or father who is black. But it does: children with a white mother and a black father have an eight-point I.Q. advantage over those with a black mother and a white father.
He also talks about one popular study that looked at births by American GI fathers and German mothers after WWII. African-American GIs sired children who scored almost identically to those of Caucasian fathers.

In a recent Manhattan Institute debate between Flynn and Charles Murray, author the The Bell Curve, Flynn pointed out that black-white IQ score gaps expand as children age. As infants, they score almost identically; the margin increases to 4.5 points by age 4, and then an additional .6 of a point every year until age 24. Murray may ascribe this widening gap as manifestation that genes matter more in the long run, but a more credible argument posits that cultural differences and the availability of stimulating cognitive environments make the difference.

Evidence supports this latter hypothesis. Flynn began dissecting the “model minority” hypothesis for Asian-Americans—that they were naturally smarter and thus became successful—by looking at data from a widely circulated 1975 study. It turned out that research relied on an intelligence test that hadn’t been updated in 20 years, meaning they were taking an easier version and biasing their scores upward. Flynn actually found they scored lower and achieved success despite the disparity. Even more interesting, their children, apparently absorbing a much more enriching childhood, scored 3% higher than average, reversing the prior numbers. This means that effort and the availability of enriching educational opportunities dictate future success. More importantly, as opposed to IQ, these are things educational policymakers can actually address.

The last point to glean from this topic is on the subject of immigration. In the 1920s and 30s, immigrants from Southern Italy scored more than one standard deviation below Americans and Western Europeans on IQ tests. Their scores, in the 70s and 80s, were comparable to those of blacks and Hispanics and fueled discussion about supposed inferiority of Italian genetics. Their scores have risen, and they are no longer the source of isolationist fervor, but maybe the debates of today should be injected with the lessons of the past.

--Guestblogger Chad Aldeman

Monday, December 10, 2007

Undeserved Publicity for Harvard

The Wall Street Journal is flashing the headline "Harvard Cuts Undergrad Prices" at the top of its Web site at the moment, along with an article that begins:

Harvard University sweetened its financial aid for middle class and upper middle-class families, responding to criticism that elite colleges have become unaffordable for ordinary Americans.
That's almost right, except the correct word would be unaccessible not unaffordable. It's good that an unimaginably wealthy institution like Harvard is only going to be charging rich people full freight. But that doesn't mean all that much when your student body is made up primarily of rich people. In the University's prepared statement, President Faust says "This is a huge investment for Harvard." Is it? Compared to what? How much does this increase the university's annual need-based aid expenditures? Will it increase the percentage of student eligible for such aid, and if so by how much? How much more aid will students receive, in percentage terms, compared to what they're getting now?

Harvard's statement concludes by noting that "With the new initiative fully in place this coming year, more than 90 percent of American families will be eligible to benefit from Harvard's exceptionally generous financial aid." I'm sure that's true, but it's also essentially meaningless, since only a tiny minority of that 90 percent attend high schools like these, identified just last week by the WSJ as those that offer the best odds of getting graduates into school like Harvard, often in exchange for tuition in excess of $25,000 per year.

Harvard has been getting a ton of great publicity over the last five years by cannily staying one step ahead of the curve in announcing new programs to cut tuition for it's small number of low- and middle-income students. But it until it actually does something to admit more of those students--not just help them once they arrive--it won't deserve headlines like these.

UPDATE: I see walking into work this morning that this is making the front page of all the big newspapers. Sigh. Some have reported, per above, the actual cost of the change, $22 million. That is (A) less than 0.4% (four-tenths of one percent) of what Harvard earned on its endowment last year, and (B) a small price to pay for this kind of publicity.

According to InsideHigherEd, "because Harvard officials said that they hoped the plan would attract new, less wealthy applicants, the share of undergraduates eligible could grow over time." Similarly, the New York Times reports that "Harvard’s dean of admissions and financial aid, William R. Fitzsimmons, said [previous changes targeted to families earning less than $60,000 per year] had increased the number of low-income students by 33 percent in three years."

Both of these statements are premised on the idea that the economic makeup of Harvard's student body is essentially a function of the economic makeup of the student who apply. That's nonsense. Harvard, like all selective colleges, decides who to admit. It rejects 90% of applicants, so it could easily put together a freshman class of smart, well-qualified students who better represent the economic diversity of the nation if wanted to. It just doesn't want to, because a lot of the admissions spots are reserved for athletes, legacies, children of the rich, famous, and powerful, and students who come through the quasi-aristocracy of east coast private schools.

The makeup of the overall applicant pool, moreover, isn't outside of the university's control. What if Harvard took $10 million and used it to create an office of people who'se only job was to identify and recruit the best and brightest low- and moderate-income students from across the nation, with a specific goal of substantially increasing the number of slots for such students and thus decreasing the number set aside for the children of wealthy donors? THAT would be front-page news.

Financial Aid and a Get Away Car

"Two college students say the high cost of tuition led them to rob a bank." That's a new one.

Yes, the high cost of tuition, difficulty finding financial aid, and stress over taking out loans is enough to break anyone, but robbing a bank just isn't the answer. Of course, they might qualify for some free college classes in prison...

Exposing the Teachers Unions' Corporatist Pro-NCLB Agenda

The new issue of Phi Delta Kappan is well worth reading, and not just because it reprints an article I wrote about high-performing community colleges early this year. There's also a priceless debate (not online, unfortunately) between Susan O'Hanian, self-styled "educational activist," and Joel Packer, head lobbyist for the NEA, wherein O'Hanian--along with University of Alabama professor Philips Kovacs--accuses the NEA of selling its members down the river by not being anti-NCLB enough. Apparently, this is because the union is just a lapdog for its corporate masters.

No, really.

Packer reaffirms that the NEA opposes NCLB and wants to gut its core accountability provisions, but also notes that NCLB is, in fact, just the name given to the most recent version of the Elementary and Secondary Education Act, which contains a lot of worthwhile programs and also provides billions of dollars in funding to help educate poor children. So just standing on the fringe yelling for NCLB to be "dismantled," as O'Hanian does here, doesn't do much good; groups like the NEA are much more effective when they engage in the political and legislative processes in a substantive way. That involves a certain level of moderation and willingness to talk and compromise with those who disagree with you--although you'd hardly know it from the NEA's current scorched-earth campaign on the Hill, which includes demanding that Democrats sign an anti-NCLB loyalty oath, and their California affiliate running ads attacking George Miller and Nancy Pelosi.

But if you think critiquing Speaker Pelosi from the left represents the extremes of political discourse, you clearly haven't read this month's Kappan. O'Hanian and Kovacs respond by grasping for the moral legacies of, variously, Woody Guthrie, Rosa Parks, women's suffrage, abolitionists, Galileo, and Malcom X while offering a combination of anecdote, childish sentiment and gratuitous insult, such as:


We call on union leaders, members of Congress, and their Business Roundtable allies to do something radical: we ask them to listen the highly qualified teachers who work with children every day.
and

the NEA leadership's decision to side with corporate reformers rather than with the teachers who pay their salaries--but certainly not their dinner bills--says a great deal about the priorities of the organization.

Packer responds with remarkable restraint, reiterating the NEA's anti-NCLB talking points before concluding that "we will not apologize for our decision to step up to the plate and actually do the hard and painstaking work work of directly influencing the policy makers who will write the next version of ESEA."

This type of exchange actually has some value beyond entertainment purposes. First, because the fact that Kappan--the second-largest education publication in the nation by circulation, after Educational Leadership--chose to print it suggests that there are a substantial number of people who actually think this way. You see it sometimes in the more leftish/academic blogs as well. Second, because, like the writings of those who believe the principal failure of the Bush administration has been insufficient war-mongering and imperialist zeal, it serves to establish boundaries of seriousness in the NCLB debate. That's useful, if nothing else.

Another Way to Give Green This Holiday

You probably know about Nicholas Negroponte's One Laptop Per Child initiative, aimed at getting his MIT-created super-sturdy, super-cheap, super-cool (happy green kid-friendly design, no glare in the sun, impenetrable by water or dust) XO computer design to the world's poorest children (w/WiFi). Well, some people and some companies don't like this idea.

I get that it might not work technically (can't say, I'm not that fluent in computers). But I don't get the "give them food instead" critique. Sure, they need food. They need a lot of things, mostly contingent on systemic change that no bag of rice will change either. And you know the saying, give a fish... eat for a day, teach to fish...lifetime. Maybe it's not a good analogy- fishing (although Eduwonk might like it). And we can't know if the XO computer will be well-received, or a tool that results in any sustainable change. But I think these kids and families should get that chance and that experience. So if it works- if those computers and WiFi really get to those kids and creates opportunity and not some unfortunate waste (computers arriving but not WiFi, or vice versa), it seems like an investment that developing countries should at least be able to consider.

BTW, until Dec 30th, for $200 you can buy one of these XO computers, as they're called, for your own kid or for any other kid you know or for yourself I guess, and OLPC will send another one to a child in a developing nation. See OLPC's give one, get one campaign here.