Wednesday, December 24, 2014

When Testing Drives the Bus

From my spiritual brother over at Curmudgeon Central comes an instructive tale from East Rockaway about what can happen when testing drives the bus. It's a tale of triumph and tragedy, and it comes in two chapters.

Chapter One: How To Achieve Excellence

As reported in a Newsday story now behind a paywall, but excerpted at Curmudgeon Central, Voula Coyle's method of preparing her fourth graders for the state test was neither new nor novel. In fact, it was exactly the technique that we already know is the best, most effective way of preparing students for a standardized tests:

The arbitrator, in her ruling, noted that fourth-graders took more than 60 practice tests from January 2012 through April 2012, virtually a daily routine. The arbitrator also noted that the elementary school's principal had endorsed the practice sessions.

Yessir. The best way to get good at taking standardized tests is to practice taking standardized tests (because the only thing a standardized test really measures is your ability to take a standardized test). And so the seventeen-year veteran regularly achieved the highest possible rating in the Students Improve On The Test portion of her evaluation.

Chapter Two: The Rewards of Excellence

You might have noticed that the above quote mentions an arbitrator and thought, "Well, that can't be good."  And Coyle did find herself in trouble. She was not in trouble for wasting educational time every day with stupid practice tests, however, or for making a mockery of the educational process by suspending it in order to do straight-up test prep.

She was in trouble for getting good test results.

Most coverage of the story tracks back to this piece from Fios1 news, which explains that Coyle's high scores in fourth grade made it difficult for fifth grade teachers to get any more improvement out of the students. This gave fifth grade teachers low ratings which in turn hurt the school ratings. Coyle felt pushback from her administration.

The instructor said her superiors have regularly encouraged her and her colleagues to avoid overachieving and to keep their scores from exceeding the state rating of "effective."
"One faculty member said our job is not to be optimal, but to be adequate. That underlying message of mediocrity was promoted," Coyle said. Teachers were also told to accept that a third of their students would understand the material, another third would be average and the rest would fail, she said.

First, Coyle was accused of cheating by telling students to go back to particular questions they had gotten wrong. That case (which, among other things, depended on not including testimony from the aid in the room during testing) fell apart, so administration simply re-assigned Coyle to a job pushing paper and replaced her with a teacher who didn't do so well.

So What Have We Learned?

The story has two morals.

First, excellence in test-taking comes from excellence in test-taking practice. Teach to that test.

Second, don't teach to the test too well, because it messes up the VAM sauce. Get test scores that are good, but not too good.

Remember-- it's not about education or learning. It's about generating the right set of numbers to keep the state happy .


  1. I have observed that over the years: you need to improve, but not too much at once or we will accuse you of cheating. Why? Because you exceeded our statistical paramenters. But how were those parameters set in the first place? and if they are accurate, why are teachers punished for not moving really below proficient students to proficiency in one year?

  2. Another one to be filed under "You can't make this sh*t up."

  3. I went from a score of a 1 to a 5 the following year based on test scores. What changed? The students. Interestingly, I promptly got transferred.