CCSD's Value-Added Teacher Evaluation: Impossible to Test Reading Fairly

A late August post questioned Mathematica's development of a $3 million formula to assess each student's value-added component each year so that the Charleston County School District can pay teachers bonuses on that basis. Supposedly the algorithm will adjust for outside factors such as poverty. CCSD believes that its teachers need motivation and more money will provide it.

In writing about his support for the Common Core, E.D. Hirsch, Jr., laments that no test can fairly judge the value added by a teacher in reading. In other words, value-added testing does not work, and teachers under the gun to produce will focus on test prep to the detriment of knowledge:

      The first thing I’d want to do if I were younger would be to launch an effective court challenge to value-added teacher evaluations on the basis of test scores in reading comprehension. The value-added approach to teacher evaluation in reading is unsound both technically and in its curriculum-narrowing effects. The connection between job ratings and tests in ELA has been a disaster for education.
      The scholarly proponents of the value-added approach have sent me a set of technical studies. My analysis of them showed what anyone immersed in reading research would have predicted: The value-added data are modestly stable for math, but are fuzzy and unreliable for reading. It cannot be otherwise, because of the underlying realities. Math tests are based on the school curriculum. What a teacher does in the math classroom affects student test scores. But reading comprehension tests are not based on the school curriculum. (How could they be if there’s no set curriculum?) Rather, they are based on the general knowledge that students have gained over their life span from all sources—most of them outside the school. That’s why reading tests in the early grades are so reliably and unfairly correlated with parental education and income.
      Since the results on reading comprehension tests are not chiefly based on what a teacher has done in a single school year, why would any sensible person try to judge teacher effectiveness by changes in reading comprehension scores in a single year? The whole project is unfair to teachers, ill-conceived, and educationally disastrous. [italics added] The teacher-rating scheme has usurped huge amounts of teaching time in anxious test-prep. Paradoxically, the evidence shows that test-prep ceases to be effective after about six lessons. So most of that test-prep time is wasted even as test prep. It’s time in which teachers could be calmly pursuing real education—teaching students fascinating subjects in literature, history, civics, science and the arts, the general knowledge that is the true foundation of improved reading comprehension. 

This piece originally appeared on the Huffington Post; a version also appeared on the Thomas B. Fordham Institute’s Common Core Watch.