If you add the scores on standardized tests for five years in a row, can you tell who the best and worst teachers are?


But that’s the theory behind value-added assessment.

The idea is that an “effective” teacher raises test scores every year. The computer predicts what the test scores are supposed  to be, and the teacher who meets the target is great, while the one who doesn’t is ineffective and should be shunned or banished.

But study after study shows that value-added assessment is rife with error. As this paper from the National Academy of Education and the American Educational Research Association shows, value-added assessment is unstable, inaccurate and unreliable. Teachers who get high ratings one year may get low ratings the next year. Teachers are misidentified. Data are missing. The scores say more about which students were in the classroom than the teachers’ “quality” and ability to teach well.

Teachers of the gifted are in trouble because the students are so close to the ceiling that it is very difficult to “make” them get higher scores.

Teachers of special education are in trouble because their students have many problems taking a standardized assessment. A teacher wrote me last year to tell me that her students would cry, hide under their desks, and react with rage; one tore up the test and ate the paper.

Teachers of English language learners are in trouble because many of their students don’t know how to read English.

A superintendent in Connecticut wrote me to say that his state department of education is pushing the Gates’ MET approach. I urged him to read Jesse Rothstein’s critique. In fact, the MET study won the National Education Policy Center’s Bunkum award for research that reached a conclusion that was the opposite of its own evidence.

For a fast and accurate summary of what research says about value-added assessment, read this article by Linda Darling-Hammond.

VAM is junk science. Bunk science.

Just another club with which to knock teachers, wielded by those who could never last five minutes in a classroom.