Something tells me...

That Mike Antonucci really wants to win Jay Mathews' "best ed. blog" poll.

Graduation Rates in Context

The new NCES report (“Placing College Graduation Rates in Context”) is worth taking a look at. While, as they note, there are major limitations to the Department of Education’s graduation rate calculations, the range in graduation rates among similar institutions is striking (and sounds vaguely familiar…). And, they found Black and Hispanic students succeeding most at institutions where they were most highly represented.

Some highlights:

- Among very selective institutions classified as low-income serving, graduation rates varied from 25% for institutions at the 10^th percentile to 75% for institutions at the 90^th percentile.

- Graduation rates were actually slightly higher for Black students compared with White students at the 16 very selective bachelor’s degree institutions with large low-income enrollments.

- Similarly, the smallest gap between White and Hispanic students existed among the moderately selective institutions with large low-income enrollments.

Also interesting is that NCES found that the "high performing" institutions were not easy to categorize based on structural differences. This is good news -- there is a lot ALL colleges can learn from institutions that both enroll a high number of low-income students and maintain high graduation rates.

Shameless Plug

Per Craig's post below (see also this and this), part of the problem here is that a lot of the tests themselves suck. Conventional wisdom aside, it's possible to design a standardized test that is a pretty good measure of what we want kids to learn--including higher-order thinking skills. The problem is that doing so is really expensive and time consuming, so a lot of states have decided to use crappy, off-the-shelf exams instead that test mostly lower-level skills and aren't aligned with state standards. These kinds of tests can create incentives for "teaching to the test." One way to address the problem would be by encouraging more states to pool their money and work together to develop higher-quality shared standards and test, like in New England. (Or, if you're Matt or the Fordhamites, National Standards.)

The Wire: Teaching to the Test Revisited

Matt Yglesias makes some thoughtful observations about yesterday's Wire post, generating a bunch of really interesting comments about NCLB and test prep over on his site. It's great to see some actual discourse on this topic—both because claims of "unintended consequences" often are unexamined and because they are sure to loom large when Congress gets around to reauthorizing the law.

I agree with Matt that my evidence doesn't prove drill-and-kill teaching to the test isn't happening in lots of places, but just to be clear I didn't set out to prove that. Instead, I'm arguing that teaching to the test is 1) a choice being made by school systems, 2) a bad choice, and 3) an unnecessary choice. It's a choice because there are other alternatives. It's a bad choice because those alternatives are better for kids and for society. And it's an unnecessary choice because research shows that good teaching—instruction that helps kids master basic skills and encourages deep understanding of math and language—actually produces higher test scores than narrowly teaching to the test.

I also agree with Matt that the solution is not to stop testing. Would anyone argue that Baltimore's elected officials and fourth estate should stop looking at crime statistics and putting pressure on the police system to improve public safety? Of course not. Any politician who argued they should stop doing that because it causes the police department to play tricks with numbers would be voted out of office (or should be). The key is to collect data and hold the system accountable while at the same time expecting the department to engage in good, authentic police work to improve public safety, rather than simply juking the numbers. And we should expect the same of school systems.

Of course, that raises some questions, too. Why are systems choosing to juke the numbers rather than doing their jobs more effectively? I would argue that some schools and districts do it because they believe that poor and minority kids really can't learn the math and language skills in question, so instead they help them game the tests without really mastering the content. Some are probably doing so out of ignorance: They've heard they are "being forced to teach to the test" and similar rhetoric so often that it has become a self-fulfilling prophecy; I fear that many educators truly believe they have no other choices. In both cases, we should do more to "educate educators" that there are better alternatives and that those alternatives work just as well for poor and minority kids.

However, as some have argued over on Matt's site, other schools and districts might know there are alternatives, but feel forced to take shortcuts because they believe they lack the capital--human, material, intellectual--to work more effectively. For example, in the world of The Wire, school system leaders might have issued the test prep directive because they believe their teachers really are not capable of teaching math and reading very well. But even if that is the case, we should focus on that question and look to building capacity, rather than getting rid of tests and accountability, as the next frontier of public policy-making.

--- Guestblogger Craig Jerald

AFT: He Who is Not the Enemy of My Enemy is My Enemy, Or Something.

AFTie Ed faults Education Sector's new Connecting the Dots about the Walton Family Foundation's support of the charter school movement for not being what it isn't--a report about Wal-Mart's anti-union activities. Why? Because, says Ed:

Wal-Mart attacks unions and those workers who want to form one. There is no reason to see their support for charter schooling as anything other than part of that attack.

Really? "No reason"? "Anything other"? There's basic aversion to empiricism here that runs throughout a lot of what gets written on the AFT blog. It's the main reason teachers unions are increasingly getting left out of real debates about education policy.

Look, Wal-Mart is, without a doubt, a stridently anti-union corporation. I think this is an inherently immoral position and Wal-Mart deserves pretty much all the blame it gets on this front. People have a right to organize and bargain collectively, whether you like it or not. Taking away that right is wrong, period.

So it's perfectly reasonable to wonder if Walton Family Foundation is pursuing an anti-union agenda through its philanthropic activities. It makes sense to use a gimlet eye in examining their motives and funding choices.

But having asked the question, you still have to answer it. Fairly, using actual facts and data. Particularly when it comes to an issue like charter schools, which are supported by people and organizations of all political stripes, including those who are as pro-union as they come. People should read the report and draw their own conclusions, but personally I find the idea that it's all just a stealth anti-union campaign to be ludricous and unserious. Choice, markets, entrepeneurship, innovation--these are really issues that deserve to be debated on their own terms. There's a certain narcissism to the belief that all your enemies think about is you.

Similary, AFTie Beth recently said of the North Carolina AYP growth model experiment:

What’s the point? Apparently, in NC, no schools made AYP just because of their growth model. It’s a lot of work for states and the feds to submit and analyze plans. If it is not helping, why continue with the process?

The point of growth models is to identify schools that are missing absolute performance standards but are making a lot of growth. What if the North Carolina didn't identify any such schools because there, you know, aren't any?

To say that it's not "helping" is to give the game away -- to AFTie Beth the issue isn't what's true or false. The issue is what helps the cause. If the facts support the cause, great. If they don't, that's their fault. This is why far too many education policy issues are hopelessly politicized, and why students get stuck with irrational systems as a result.

The Wire, Week Nine: Ain't No Jukin' It!

This week the writers ramped up the education subplot in a scene where Tilghman’s principal and vice principal order teachers to spend the next six weeks preparing students to take an upcoming standardized test. The word has come from downtown: Teachers are given test preparation materials and enjoined to drill students in how to answer “clone questions” that closely parallel the questions that will be on the test—and only the reading portion of the test, so Prez’s slowly improving math lessons must be put aside. Prez draws a parallel to his experience in the police department, where top brass, under pressure from politicians to improve public safety stats, periodically sent word to “juke the numbers” to create the impression that crime was going down and/or arrests were going up.

One way to read what’s happening is this: The No Child Left Behind law is putting pressure on educators to improve test scores, so they are “being forced” to substitute dumbed down, drill-and-kill test prep for real teaching and learning. That explanation has pretty much become part of the national discourse over the past five years. “Teach to the test”—It’s such a suggestive, seductive, even alliterative (four t’s!) phrase, one that critics of NCLB use quite effectively to conjure up outrage on demand. Fortunately for them, it’s also one bit of conventional wisdom that is very seldom questioned.

But it should be. Let’s break it down. In The Wire’s scenario, Prez and his colleagues are being told they must teach to the test, but not by No Child Left Behind. The directive is coming from the school’s administrators, who are, in turn, passing on a directive from downtown. Okay, perhaps the district administrators are being forced to force principals to force teachers to teach to the test. But that would be true only if one of the following conditions is true: 1) There are no other strategies Baltimore's schools could use to raise test scores, or, 2) If there are other strategies, teaching to the test is the only strategy that can raise scores sufficiently to meet the demands of NCLB.

But according to the best research available, neither of those conditions holds true. For example, researchers with the Consortium on Chicago School Research conducted a three-year study analyzing classroom assignments and student gains on standardized tests across more than 400 Chicago classrooms in almost 20 elementary schools. They found that students whose teachers assigned them more “authentic intellectual work”—tasks that called on disciplined inquiry, complex thinking, and deep conceptual learning—logged much higher test score gains than students whose teachers relied on “drill and kill” assignments normally associated with teaching to the test.

That means good teaching produces higher test scores than “teaching to the test”! Not coincidentally, it also produces better educated citizens equipped with a more powerful set of intellectual skills increasingly in demand in today’s workplace. Of course, a little test prep, such as helping students become "test literate" so they understand how tricky questions can prevent them from showing what they really do know, can be healthy in small doses. But there is no legitimate reason to spend months on the kind of soul deadening test prep materials Tilghman's teachers were given in this episode. (More on all this here.) Prez is right to question the school system’s choice about how to handle NCLB and standardized testing, even more than he realizes.

The police department has made a similar choice to juke the stats rather than to engage in authentic police work—such as the investigations conducted by the Major Crimes Unit over the first three seasons. (I won’t spoil it, but McNulty gets some screen time next week in small sub-plot that very effectively underlines the choice between juking the stats and doing real police work.) Like authentic instruction, authentic police work is better for better for the professionals as well as the citizens they serve, and better for the city of Baltimore as a whole because it promotes real improvements in the quality of life.

This makes for a much more complex storyline than the “forced to teach to the test” narrative that many viewers are reading into The Wire’s education subplot, and it leads to a much more compelling question: Why do these systems make the choice to game the numbers rather than promoting good teaching or good police work, especially when the latter will produce better results over the long term? To some extent, that is the single biggest overarching question The Wire has been exploring for the past four seasons. Unlike Tilghman’s teachers, the writers are unlikely to spoon feed us the answer.

--- Guestblogger Craig Jerald

Disappointment abounds

When I read the headline of this New York Times article (Leaving the City for the Schools, and Regretting It), I was excited – perhaps there was finally an article extolling some of the virtues of urban public schools. Yes, these schools have plenty of troubles and there are schools in NYC that any parent would run from (assuming they could), but there are also schools in NYC that provide a good education, and with a diversity of perspectives and backgrounds that can be an educational asset. Unfortunately, that is not what this article was about.

I won’t go on too long, since the article is clearly intended to fuel the anxieties and “beat the Joneses” impulses of upper-class parents. Suffice it to say that the concerns highlighted do not impact most people. For example, one parent’s pressing problem that “her husband’s Ford Focus had become something of a joke parked alongside his co-workers’ Mercedes-Benzes and BMWs, as the family has forgone fancy cars and vacations to afford the tuition [of an expensive private school].” Or the horror that “public schools, powerless to choose who fills the desks, often had less motivated students.” Which, by the way, is not true – those high property taxes parents in the article complain about work as a pretty effective filter for who gets to fill the seats at suburban high schools.

Is this 'trend' a new problem of failing suburban schools? I doubt it. It is more likely that there is a new echelon of parents looking to give their kids yet another edge over their neighbors, and that means forsaking public schools that would be the envy of most parents in this country. Any parent would do this. It is, in fact, very reasonable for parents to give their children the best education they can afford. The unfortunate thing about this article is that it feeds into the notion that public schools are inherently inferior - only meant for the average kid (and what parent would say that their kid is average?). And it wastes news space - those words could have been better utilized focusing on the similar plight of parents in the boroughs around Manhattan, where the choice can be much more difficult - to risk their child's future at a failing public school, or find financial aid and make sacrifices (beyond giving up their Mercedes) to give their kids a chance at a good education.