IN THE POWER OF TESTING! David Brooks believes in tests. But how much does he know about schools? // link // print // previous // next //
WEDNESDAY, MARCH 18, 2009
The Gaels and the French and the jam and the carrot: As she opens this mornings column, Maureen Dowd has Barack Obama speaking a bit of Gaelic. Before long, though, we see Andrew Cuomo borrowing from the French:
If we may borrow in turn from Steve Martin: Those French! They have a different word for everything! (The paragraph we cite is from Dowds hard-copy column.)
Dowd, of course, is writing about the bonuses at AIG. But along the way, she showed her skill with a third languageupper-end journalese:
Poor Dowd! Shed been off, and hadnt had a chance to complain about the earmarks Congress jammed into that bill. Like all languages, journalese is acquired through repetition. Dowd rattled one there for the team.
(Unnecessary earmarks? Quick translation: Dowd doesnt live near a pig farm!)
Many languages marble through this mornings fist-waving column. But Dowds first love is for We Irish. As it turns out, the president isnt quite Irish enough, unlike the columnists dad:
Good advice! As we see in that example, no one reasons quite so clearly as Dowds Gaelic, non-journalist clan.
We were struck by one last passage, though we cant make it fit our theme:
Set aside your views about Geithner. Does this scribe know a bit too much about the way people start fires?
For those who prefer their hypocrisy themes: Reciting from upper-end press corps scripts, Dowd complains when phony Obama recites from a teleprompter. (See Dowds opening graf.)
PART 2IN THE POWER OF TESTING: Big pundits sometimes tend toward omniscience when they discuss public schools. In last Fridays column, David Brooks surrendered to this impulse as he rated the various parts of Obamas education agenda:
Assume Obamas agenda is enacted. How does Brooks know that merit pay and teacher dismissal will turn out to be more important than early education provisions? We dont have the slightest ideaand the omniscient rarely explain.
At any rate, Brooks believes in merit paynot that theres anything wrong with it. Its possible that some form of merit may might turn out to be helpful somewherebut there are certain obvious problems with such proposals too. One of the major possible problems involves the possible use of test scores in determining teacher pay. But then, David Brooks believes in the power of testing. Despite his apparent omniscience, he doesnt seem to have heard about one problem which lurks in some uses of tests.
Brooks believes in the power of testingnot that theres anything wrong with it. We believe in annual testing ourselves; indeed, such testing was already the norm when we taught our first fifth-grade class, in 1969-70, here in Baltimore. (In those days, every Baltimore student took the Iowa Tests of Basic Skills, a widely-used set of standardized tests. The ITBS came with detailed technical manuals, unlike todays statewide tests.) Absent some sort of annual test, big school systems can tell you anything about the spectacular state of their progressand some big schools systems surely will. Tests can be put to unfortunate uses. But we cant imagine running a major school system without them.
But Brooks doesnt just believe in testsDavid Brook really believes. That said, it doesnt take long before he makes some odd remarks about testingthe type of remarks big pundits make when they may not really understand schools. In our view, Brooks is smarter than most major pundits. But sorrythis doesnt make sense:
Sorrythats just strange. Who wouldnt have guessed that progress on tests between the third and eighth grades would correlate with high school graduation, four years later? Who did Brooks think was likely to graduatethe kids who kept going downhill on tests from the third grade through the eighth? And Brooks isnt content, in this slightly odd passage, to gape slack-jawed at a correlation any eighth-grader could have predicted. Its also strange to see him declare that the predictive power of Kleins data represents a clear demonstration of the importance of these assessments (these tests). In reality, those assessments only become important if they somehow help teachers know how to reverse the decline of the kids who seem to be failing. Do Kleins assessments do that? The answer is far from obvious, but theres no sign it occurred to Brooks to ask. But so it sometimes tends to go when pundits proclaim about schools.
Indeed, Brooks odd statements about Kleins data run a bit deeper than we have shown. As this longer passage shows, Brooks seems to think that these miracle tests are something we have only now today. As Brooks explains what todays tests can do, experienced teachers, slightly embarrassed, will look off, averting their gaze:
Today, tests can tell you which students are on track? Which teachers are bringing achievement up by two grades? Did we mention the fact that we administered the Iowa Tests to Baltimore kids in April 1970? Sorry, but those tests were really quite good at telling you which students are on track. Are Kleins tests more powerful in some way? We doubt it, but theres little chance that Brooks knows the answer to that.
Uh-oh! Unless were grading on the curve, wed have to say Brooks is floundering here; his column just isnt on track. And as he continues, a mustachio-twirling demon of a familiar old story appears. That demon is the evasive, unhelpful education establishment, of course. (As Brooks closes, liberal orthodoxy on school reform presents as a demon too.) Well only review the scribes first complaint, leaving more for tomorrow:
Earlier, Brooks praised merit pay for good teachers. Here, he complains that his demonthe education establishmentis refusing to use todays wondrous new data to reward good teachers this way.
For ourselves, we have no huge view about merit pay. But we do understand that there are a few down-sides to this ballyhooed reform idea. Brooks believes in the power of tests. But he doesnt seem to know about the possible problem we will now reveal:
Many reformers are now suggesting use of test scores to help determine teacher pay. Conventional experts tend to agree, before they knock off for their three-hour luncheons; for that reason, major pundits tend to swear by this widely-affirmed idea. One example of such a proposal: In Washington, new superintendent Michelle Rhee has proposed paying truly big bucks to teacher who give up tenure and produce good results in the classroom. In this recent news report, Bill Turque cited Rhees plan to pay teachers as much as $135,000 a year in salaries and bonuses.
This isnt about Michelle Rhee; her proposal may turn out to be well-crafted. But well now ask a question Brooks didnt offer: Do you know how much some teachers will cheat to snag a high salary like that? Do you realize how worthless todays tests will be, if teachers engage in this cheating? And yes, we ask about cheating here, not about teaching to the test.
Today, tests can tell you which teachers are bringing their students achievement up by two grades in a single year? Thats only true if the tests are administered in an appropriate manner. But those tests are normally administered by classroom teachersby the very people who will get that large pay if the results are strong enough. And yes, some teachers will cheat their keisters off to earn $135,000.
Yes, Virginia (and well get to that state on Friday): For decades, teachers and principals have frequently cheated when pressure is put on these testing programs. (For a recent, front-page example, click here.) This has been recorded again and again, but experts and journalists rarely seem to incorporate this into their world view. Were always struck by the omission when experts and journalists discuss merit pay without discussing this obvious problem. As the piles of money at stake grow larger, the obvious hole in such discussions becomes more remarkable still.
Brooks believes in merit pay, which is fine with us; we would assume that some such plan could be constructive in some situations. He seems to believe in the power of testing, quite deeplybut he doesnt seem to know about the pitfalls which sometimes lurk. In his column, Brooks praises Obama, who has broken with liberal orthodoxy on school reform. Obama will build on a Bush program that gives states money for merit pay so long as they measure teachers based on real results, Brooks exclaims, believing truly. Does he know that teacher cheating is one way those results get unreal?
Sorrywe hate to ruin a novel. But its crazy to talk about basing huge salaries on test resultsif the teachers who give those tests are the people who end up getting those salaries. Well assume this thought would occur to most teachers; this may explain why some demons in the education establishment arent quite as high on merit pay as Brooks reformer luncheon guests. Any teacher could spot the shape of this problem. Routinely, the omniscient do not.
Brooks believes in reformin the power of tests. Truly, theres nothing wrong with that. But reading his column, a question came to mind: How much does he know about schools?
Tomorrowpart 3: In higher standards