Carol Burris, principal of south side High School in Rockville Center, New York, writes here about the multiple flaws of test-based teacher evaluations.

At an Ed Trust celebration, Duncan told the crowd, “But we can’t let the perfect become the enemy of the good. We can’t let the utopian become the enemy of the excellent. And we can’t let rhetorical purity become the enemy of rigorous practice.” I do not have any idea what the third admonishment means, but I doubt Arne needs to fear that his rhetoric is pure.

So it came as no surprise that when he spoke of Tennessee’s teacher evaluation plan, Mr. Duncan praised the state for “not letting the perfect become the enemy of the good”. The teachers of Tennessee, however, are not seeing the new system as “the good”—they are, for the second time, suing the state because the system is, in their eyes, arbitrary and flawed. And it is.

When it comes to the new teacher evaluation systems, it is not a dispute between perfect and good. We are now forbidding the good to be the enemy of the lousy. The use of students’ scores is becoming more and more indefensible. In New York State, teachers despise APPR, and it is equally unpopular among principals who, for the most part, see it as a headache that does nothing to improve teacher performance. Teacher and principal scores, by district, were supposed to be released in the winter. It is the end of July and they have not appeared. That is not a surprise. If they were released, it would be an embarrassment, especially for districts that actually tried to engage in the Las Vegas pursuit of predicting student growth from pre-tests to post-tests. The New York State Education Department is stalling, and Governor Cuomo is letting it happen.

There was one state, Massachusetts, that created a plan that was more sensible than most. It did not use numbers, but rather was rubric based. It was phased in over time and applied to everyone, including central administrators. But now that the time has come to phase in the test scores, the trouble begins.

In his July 17 memo to Superintendents and Charter School leaders, Commissioner Mitchell Chester states he is pleased that the Bay State has not chosen “an algorithmic approach,” only to later explain in detail the algorithm by which teachers should be evaluated by test scores. To go further down the path of the lousy, he explains how the state will generate growth scores from PARCC exams for participating schools, and then attempt to show “growth” from the prior year student MCAS scores. Please say it isn’t so. That is not a growth measure. That is comparing students with similar scores on one test with each other the following year on an entirely different test. New York did the same thing last year. Can you do it? Of course you can—there is very little that you cannot do with numbers. It is easy to create a formula that is intimidating enough that eyes will glaze over. But that does not make it valid, reliable, fair or useful. It will be one more silly system that will result in a lawsuit, no doubt.

Chiefs for Change, including State Superintendents Huffman and Skandera, took the NEA and AFT to task for having the guts to back away from the test-based teacher evaluation systems they once supported. They accused them of ‘evading accountability’ like horse thieves running from the posse. They wanted union leaders to sit compliantly with hands folded, in the face of mounting evidence that the test-score evaluation systems are not working. These Chiefs for ‘change at any cost’, do not understand. True accountability means having the courage to speak the truth when facts come to light, even when it contradicts what you once supported. To keep one’s mouth shut as the lousy marches forward is wrong.