March 29, 2006

My letter to Jay Mathews

Last month, Jay Mathews wrote an op-ed, Let's Teach to the Test, wherein Mathews defined teaching to the test as not drilling. That struck me as rather odd, so I wrote him. I didn't make this letter public until now because I wanted to give him a chance to respond and because, well, I've been rather busy. But Mathews responded with a brief kind note, so here's what I wrote (on the continuation):


Mr. Mathews,

As usual, your column today made me think ... about how malleable language is, in particular. To you, inappropriate drilling doesn't come under "teaching to the test." I'm not so sure, though I agree with your conceptual argument that there's a difference between teaching to a curriculum that's the putative basis for a test, on the one hand, and distorting that curriculum to game the system, on the other.

Let me start with where I agree with you. Next week, my 10- and 13-year-old children and their million closest friends will be taking the Florida state tests. As I've told them for some years, I don't care whether a homework assignment has the word "FCAT" in it. I know that enough pressure pushes teachers and principals to call lessons FCAT related regardless of whether it's drilling. The key question for me as a parent is whether it's legitimate instruction if you take the word FCAT off the whiteboard or the worksheet.

But I disagree on a few points. First, I have seen and heard about curricular narrowing to test subjects in ways that not only is teaching to the test in the way you discuss (and Lauren Resnick, too, if you remember her arguments on that point about 15 years ago), but that violate the state's own curricular framework. Schools aren't supposed to deemphasize history or ignore art and music so they can teach the tested subjects, but they do.

Second, the pressure to raise test scores in Florida has distorted other practices in a way that can firmly be called teaching to the test. Let me identify the shifting of the school year as a clear example. You may or may not know that most Florida schoolchildren start the year in early August—August 4 in Tampa, I think. And my wife, who teaches in the Tampa schools, returned to work the last week of July. There is one legitimate reason for doing so and one illegitimate one. The legitimate reason for shifting the schedule is to have a full semester for high school students before winter break so they can take the finals before the midyear break. Of course, high school students have asked for that change for years, but it wasn't until the current regime put pressure on educators that superintendents and school boards realized that they wanted as much of the school year before the late-winter testing as they could manage. The school year has shifted three (and for some districts four) weeks in the last decade. I suppose the reductio ad absurdum schedule would be to adjourn schools in mid-March, right after testing, and then resume in mid or late May. In reality, I hope you'd agree with me that shifting the summer vacation around (as opposed to having interventions during the summer) does not really change what children learn. (My wife thinks that having the long vacation in August and September would make more sense in hurricane-vulnerable Florida, especially after St. Lucie County's schools were closed for five consecutive weeks in 2004-05.)

Third, and perhaps most importantly, excluding inappropriate drilling from teaching to the test is misleading, because drilling is the main technique abused by schools in response to high-stakes testing. Publishers—sometimes the same publishers as produce the tests—sell millions of test-prep booklets to the schools every year. My children do have a legitimate beef when they complain about the weeks of drilling before the state testing, and it's become worse every year. To omit drilling from your discussion of teaching to the test ignores what upsets children and parents most, I'd gather.

If you think I'm wrong on this last point, I'll challenge you in two ways here that I think are decent rule-of-thumb tests of how much test-prep cheats kids of a real education and how many resources test-prep absorbs:

1) Find two sets of test-prep booklets sold as specific to any single state that has any redeeming values as main instructional tests when measured against that state's published standards;
2) Find five public high schools or middle schools anywhere in the country this year which spent more money buying class sets of Shakespeare than it spent on test-prep booklets. [And now, blog readers, you know the origin of the challenges.]

You may be able to meet these challenges, but I suspect it'll take you quite a bit of time. Consider it a gedankenexperiment, at the very least.

Listen to this article
Posted in Education policy on March 29, 2006 9:35 PM |