Testing, Testing 1-2-3
On Tuesday, January 9, 2013 the Bill and Melinda Gates Foundation released their final report on its three years research project: Measures of Effective Teaching (MET). AP reported conclusions of “test scores or principal evaluations are not enough on their own” and that “it takes multiple measures to most accurately judge educators.” Later I saw the LA Times’ article on the same report which led off with “Student standardized test scores can accurately identify effective teachers”. I am familiar with the Times’ bias as their release of raw test scores caused a teacher to commit suicide. Using student test scores makes it easy to blame the teacher, so I went directly to the MET report at to see what I could find.
In the report I found that the researchers adjusted their “estimates to account for student differences in prior test scores, demographics, and other traits.” In the interest of fairness they randomly assigned students to different teachers. They found that reliability increased with more observers as opposed to fewer observers over longer periods of time. In addition to testing and teacher observations, the report found student perception surveys important.
One interesting part of their findings was that through the responses of multiple observers, it was easier to identify both good and bad teachers. Using a videotape of a lesson ensued that each evaluator witnessed the same lesson. Because of individual biases and different ranking levels, there was a need for many different observers. The report found that teachers could be just as effective as administrators at evaluating good teaching. However, the teachers rated their fellow teachers closer to the average while administrators’ standard deviation was 50 percent larger. ”Compared to peer raters, administrators differentiated more among teachers.“ One explanation for this is that the administrators were more discerning, but another way to look at it is that they were more biased. Since the purpose of evaluation should be to help teachers improve, I believe that teachers should be evaluated primarily by other teachers. Teachers can offer better advice on how to improve, since as teachers themselves they are fully aware of the problems involved in teaching. Through the process of observing other teachers, the teachers observing could also pick up tips for making themselves better teachers. Ultimately, teacher evaluation should lead to improved teachers, which in turn will lead to better education for students.
After spending three years doing research, the Gates report concluded that testing alone was not enough to reliably identify effective teaching. The report stated that student surveys and teacher observations were also needed. What is the agenda of those who want to make student test scores the sole basis of teacher evaluation, employment, and pay? Some may claim that test scores provide accountability, but whose accountability? The accountability of the students taking the test? The accountability of the parents of the students? The accountability of the politicians who provide the funding for the schools? Or the accountability of society itself? It seems to me that too many people want to use test scores to blame the teacher.