We at the Center for Inquiry don’t typically address how best to evaluate teacher performance, but we are an educational institution. Moreover, we’re an institution that is committed to basing public policy on sound science. Using student test scores to determine teacher pay or whether a teacher should retain a job is a policy not based on science, but rather politics and deceptive intuitions.
The Atlanta public school cheating scandal has, of course, brought this issue to the front page. It has highlighted the extent to which teachers and school administrators feel pressured to have their students perform well on standardized tests.
This scandal also highlights the fundamental flaws in the predominant model that has been adopted for measuring teacher performance, namely how well their students perform on standardized tests. This model reflects a serious misunderstanding of the responsibilities of teachers and the relationship between teachers and pupils.
Imagine this scenario: you are a factory worker who has been informed that your job depends on the quality of the product you are assembling, as determined by how well the product performs on certain tests. Sound fair? In the abstract, maybe. But there’s a catch. You’re not the only one who is working on the product. There is a second shift, and the persons on the second shift sometimes don’t do any work on the product, sometimes expose the product to harmful chemicals, and sometimes actually undo the work that you’ve completed. Complicating the situation even further is that even on your own shift, each worker is supplied with materials and parts that vary widely in quality. Under these circumstances, one might understand why the workers feel frustrated and demoralized and, if possible, might be tempted to alter test results on their products so they can keep their jobs.
In other words, one might understand how frustrated and demoralized teachers feel when they are held solely responsible for their students’ test scores given that these results reflect many factors other than the quality of a teacher’s work.
Conditioning a teacher’s employment or pay on their students’ performance on tests ignores a critical fact: students’ educational attainment depends on many factors entirely outside an individual teacher’s control, not the least of which is the student’s home environment. Teachers have students for only a portion of each weekday. The “second shift”— to analogize to my hypothetical— namely, the parents and guardians of the students, have control over them for the other portion of each weekday and, typically, the entire weekend. Factors such as the education and socio-economic resources of the parents, as well as their interest in and support of their children’s education, have a significant influence on student achievement. A teacher cannot work miracles on a student who receives no educational reinforcement from the home environment.
Moreover, students are not identical in their natural abilities, nor have they all received the same educational preparation prior to entering the school system. If “no child left behind” implies a commitment to strive to have each pupil perform to the best of their abilities, that’s a laudable objective. If it implies that we can realistically expect all students to perform well on standardized tests, that’s an impractical and unreasonable goal.
But don’t take my word for it. Studies have demonstrated that evaluating teachers based on student performance on tests is not a valid method of measuring teacher competence and effectiveness. Yet too many school systems persist in relying heavily on test scores because supposedly it provides an “objective” result. It does no such thing. Using test scores as the principal method of assessment is based on nothing more than intuitive reasoning motivated by politics. “Vote for me: I’ll hold those teachers accountable!” That line is more pleasing to voters’ ears than telling parents to put down the remote and read a book with their kids.
Proper assessment of teacher performance requires careful classroom observation. Of course, this method is more time intensive than merely reviewing a list of test scores, but we should have learned by now there are no shortcuts when it comes to improving education.
Nothing I’ve said here should be interpreted as condoning the actions of those Atlanta teachers and administrators who intentionally altered test scores. What they did was wrong. Period. But we can expect more such incidents if we continue to impose unfair and unrealistic burdens on our teachers, conditioning their jobs and pay on test outcomes over which they have insufficient influence—except when they supply the answers themselves.