Reformers look on education as something like an assembly line: the requisite machinery and bins of gleaming parts are all in place. If the products that come off the line don’t sell, it must be the welders who are to blame.
Under this mechanistic view, no one stops to consider blindingly obvious design flaws. What if the factory is tooled up to turn out gleaming, new … Edsels?
There’s more than a little of this sort of oversimplification evident when the talk turns to teacher evaluation: Kids arrive for the school year well prepared. Teachers have been taught by the university and certified by the state. Principals and supervisors are trained to oversee and evaluate teachers.
If the test scores don’t go up, then it’s because the welders — I mean, teachers — held back or didn’t try hard enough.
Worse, the Christie administration and some legislators think it’s a pretty simple matter to gauge which teachers should stay and which should go.
At this point it’s time to hit the brakes, switch on the emergency flashers, and light a few flares — before totally unproven and potentially destructive standards and practices become state law.
But what’s the problem, exactly?
The federal education department is ordering every state to link student test results to individual teachers as the first step to improve “accountability.” On the face of it, the idea of tying test results to teacher evaluations makes sense. Until the details are inspected.
Stipulation: everyone agrees that the current teacher evaluation-tenure system is close to useless and needs to be improved. When a system reports that 98 perent of its practitioners are “good” or “great,” something is wrong.
Now the Christie administration comes along with its proposal to reform the teacher evaluation-tenure system with a plan that requires that at least one-half a teacher’s evaluation be tied to student test scores.
Without wandering into the psychometric weeds, here are four problems with the plan:
Eighty percent of teachers do not teach a grade or course that is subject to state testing.
Further, no state, school district, bureaucrat, or scholar has a workable suggestion as to how to produce uniform, reliable, and fair measures for the untested grades/subjects. Think about judging a middle school music teacher if some students practice their instruments nightly as required and some don’t.
“Not to worry,” reports acting education commissioner Chris Cerf, “the Department of Education is working on some measures and should have them ready by September.” That is, we’ll do in eight months what has not been done elsewhere after years of effort.
Standardized tests should only be used for their intended purposes. State tests are designed to gauge student mastery of some state academic standards, not to evaluate teachers.
Take the SAT, which is designed to predict if a student can handle the academics in the first year of college. College-bound seniors in 2011 from Mississippi, the nation’s poorest state, made mincemeat of seniors from New Jersey, one of the wealthiest. Mississippi students scored 568 on the reading part, compared with 495 for Jersey seniors. But recognize that 86 times more students took the SAT in NJ and that only 30 percent of them were in the top tenth of their high school class (vs. 58 percent of Mississippi of students).
Conclusion: the SAT is not a valid measure of high school quality or of a state’s education.
New Jersey state tests can be used to identify possible instructional problems to help both students and teachers; they should not be used to make key personnel decisions like tenure or dismissal.
The proposed “reform” relies on exactly the same folks who currently find 98 percent of teachers good or great, to evaluate teachers under the new scheme.
Proposals from both Christie-Cerf and Sen. Teresa Ruiz (D-Newark)
give principals nearly absolute authority to evaluate teachers. Unless the teacher can prove that the district evaluation system was not followed, there is no appeal or grievance to the superintendent, board of education, or NJDOE.
Nothing has changed from the current broken system, except that teachers are left defenseless in the event of arbitrary, spiteful, uninformed, or undocumented evaluations. Such things are known to happen (except in the reformer’s world).
Further, the education commissioner is expected to review and approve all 600 district-created evaluation systems. Again, reformers are relying on exactly the same institution that oversees the current mess, to suddenly obtain greater knowledge and discernment.
The search for improved teacher evaluation methods, underway in a few districts and states, has been inconclusive. Why not wait?
Even Michelle Rhee, the superstar reformer, accepted limits on using test results in Washington D.C.’s revised evaluation system. Called IMPACT, now in its third year, the approach acknowledges that teachers of untested subjects cannot be evaluated with “quantifiable measures of student performance.” Instead, the Rhee plan relies on five annual observations by principals and specialist “master educators.” Under Rhee, the D.C. schools attracted extensive outside funding to pay for the costs of the new system and for the merit bonuses it triggers, funding not likely to flow to New Jersey.
Already, IMPACT has made significant adjustments — struggling teachers are given more time to work with mentors, and teachers with two years of “highly effective” ratings are excused from two of the five observations. In addition, it is too early to declare victory or to be certain that what works in D.C. could be applied to New Jersey.
Back at the assembly line, the efficiency experts with their pocket protectors are all over those Edsels, toting up spot welds and seams and completely missing the point. It would be funny if it were only metaphor, but New Jersey’s education reformers don’t have an evaluation system that has proven to be reliable, consistent, and fair — and yet they’re ready to imperil the professional lives of countless teachers and principals.
It’s time for them to take their foot off the gas before it’s too late.