Daily Howler logo
PAY THE TEACHERS WELL! Houston institutes merit pay—and the New York Times flags a key problem: // link // print // previous // next //
FRIDAY, JANUARY 13, 2006

MUST-SEE ABC: So how about those college grads—the ones who scored lower on the 2003 National Assessment of Adult Literacy than their 1992 counterparts? Why did the average score of college graduates drop on this measure? We plan to continue that topic on Monday. Mark Schneider (commissioner of the National Center for Education Statistics) has responded to some questions we sent him, but we’re waiting to hear from a few other hombres. As we have said, this topic gives us a chance to consider the way such surveys tend to get reported. For our money, Lois Romano’s “literacy experts” left a great deal to be desired when they vented about those scores. On Monday, we’ll resume the discussion.

Meanwhile, a related topic—and some must-see ABC! Last evening, John Stossel guested on The O’Reilly Factor. Here’s how the session concluded:

O'REILLY (1/12/06): All right. John Stossel. American kids are dumb—and a lot of them are, unfortunately. We appreciate you coming on in. Friday, special 20/20.
Considering the year that O’Reilly just had, it’s rich to see him shaking his head about how “dumb” our children are. But that will be Stossel’s topic tonight in an hour-long 20/20 report. His program is titled, “Stupid in America—How We Cheat Our Kids.” Our best guess? Some of Stossel’s report will be dumb—and some of it will be worth considering.

PAY THE TEACHERS WELL: In this morning’s New York Times, Ralph Blumenthal presents an intriguing report about a Houston merit pay program “which calls for rewarding teachers based on how well their students perform on standardized tests.” The Houston plan has just been approved; it will “distribute up to $3,000 annually per teacher and up to $25,000 for senior administrators.”

We’re not necessarily opposed to “merit pay.” But here’s the part of Blumenthal’s piece which we found intriguing:

BLUMENTHAL (1/13/06): The 9-to-0 vote at the monthly board meeting of the Houston Independent School District, the largest district in the state, with 210,000 children, opened a new front in the long-running national dispute over teacher merit pay and excited particular emotion in a city bruised by a cheating scandal that called some schools' test results into question.
Merit pay may be OK—to the extent that merit can be determined. But there have been endless cheating scandals in public schools in the past thirty years, and when teacher pay is tied to results from testing programs which the teachers administer, schools seem to be asking for further trouble. It’s a well-established fact—teachers and principals do cheat on tests. They cheat in every imaginable way—and in some ways you likely can’t think of.

In an ideal world, how would this problem be addressed? Ideally, high-stakes tests would be administered by proctors—by people who have no personal stake in the test results. And teachers would be kept from seeing actual items on upcoming tests. (This may be the vase in Houston.) If you’ve seen future test items, there are a hundred inappropriate ways to “prepare” your students for them. In many cases, teachers may do so in perfect good faith, not understanding the basic logic of good testing procedure.

Who knows? The Texas and Houston schools may have built security measures into their testing procedures. But this groaning problem is rarely addressed in reports about merit pay programs. Blumenthal deserves credit for citing the issue, but he doesn’t do it justice. One has to read to the end of his report to get this small bit of background:

BLUMENTHAL: Randi Weingarten, president of the United Federation of Teachers in New York, contrasted the Houston plan unfavorably with the one adopted in Denver. That plan offers the system's 4,300 teachers a choice of enrolling in a merit pay program or accepting standard raises, although new hires are automatically in the merit pay plan. So far, 735 teachers have chosen the merit pay option, said a Denver school spokesman, Mark Stevens.

But in 1999 the Texas Education Agency began investigating Houston and other districts because of suspicious results on the statewide test, then called TAAS. Last year, the Houston school board said it had found evidence of cheating at four schools and testing irregularities at seven more. A half-dozen teachers were fired, and several principals were demoted or reprimanded. [END OF REPORT]

Cheating was still going on last year—so Houston ties teacher pay to the tests! Have they built security measures into their plan? Blumenthal doesn’t discuss that.

Why is this such a troubling matter? Because the public needs to get reliable info from these testing programs. Without reliable testing programs, public school systems can (and will) tell you anything about their vast progress. Especially in the case of low-income schools, we need to have reliable measures of student academic achievement. But uh-oh! Unless security measures are in place, Houston teachers have three thousand new reasons to monkey around with their tests.

WONK RIPS WINER: Over at EduWonk, Andrew Rotherham hammered Michael Winerip for the column we linked to on Wednesday (see THE DAILY HOWLER, 1/11/06). And Rotherham links to previous critiques of Winerip’s work on No Child Left Behind.

We’re not sure we fully understand the complaint about Winerip’s column. Here is Rotherham’s nugget statement:

EDUWONK: In [Wednesday’s] column Winerip returns to his favorite storyline: the school not making adequate yearly progress and the unfairness and horror of it all. He also employs a method familiar to long-time readers: misleadingly conflating state and federal requirements to paint a Kafka-like nightmare. There is no doubt there were hassles for the principal he describes, but they're really because of state and city officials though most readers can't be expected to piece that together on their own in a column basically attacking the federal law.
In his column, Winerip speaks, at various times, about federal, state and local behavior; perhaps he could have assigned responsibility for the problems described in a more precise way. (Whoever is ultimately responsible for it, Winerip surely does describe something resembling a “Kafka-like nightmare.”) But are the state requirements involved in this mess independent of NCLB? Or do they represent a bungling attempt to implement the federal law? After all, when the feds tell the states to do certain things, a degree of state bungling will surely follow. Reading back through Winerip’s piece, we can’t really spot the unfairness which Rotherham alleges.

Has Winerip had his thumb on the scale about NCLB? We can’t spot it in this column. But we plan to read EduWork’s earlier posts. You can too; just click here, then continuing your clicking.

TOMORROW: A short aside on the long Frontline film, Country Boys. Yes, we watched all six hours.