This week, State Supreme Court Justice Michael C. Lynch ruled that the New York State Board of Regents incorrectly interpreted a new law on teacher evaluations, the New York Times reported today. The ruling invalidates parts of the Regents’ vote on evaluations and will delay the introduction of the law.
The teachers union had sued the Board of Regents in June, arguing that the use of student test scores to evaluate teachers went beyond what the law permitted. The judge ruled that the Regents failed to consider the law’s collective bargaining requirements or the law’s requirement that test scores alone cannot determine a teacher’s performance review.
The ruling is a hollow victory for teachers, as they may still be evaluated based on student test scores. The law says that 20% of their evaluation must be based on state tests, while an additional 20% may be based on locally-derived tests. 60% must be based on subjective measures, like observations by administrators. The state attempted to circumvent the law by basing 40% of their evaluations on state standardized exams, rather than the 20% allowed. In fact, they could have even gone up to 40% total, so long as half was based on local tests.
Because student test scores measure student ability, they can only tell us how well students perform on standardized tests. They cannot tell us why. For example, students with test anxieties tend to perform poorly on high stakes exams, even when they know the material. This has no connection to the quality of their teachers. Likewise, numerous studies indicate that students’ socioeconomic backgrounds correlate much more strongly with their test scores than do their teachers’ quality.
While it might seem counterintuitive, teacher quality has very little to do with student test scores or even their ability to improve on tests. Good teachers may be able to help some low performing students improve, but they cannot reverse the effects of a lifetime of poverty, malnutrition, lead poisoning, low birth weights, absenteeism, lack of access to reading during the pre-K and toddler years, or lack of access to enriching extracurricular activities. A super teacher in a low income school will still see lower overall test scores and weaker improvements than a mediocre teacher in an affluent school. Therefore, student test scores should not be used at all to assess teachers.
Unfortunately, there is no reliable or fair method to quantitatively evaluate teachers. We could regularly test the teachers, rather than their students, but this would only tell us what they know about their content, for example, and nothing about how they actually deliver it in the classroom. Because teaching is a social endeavor, teacher evaluations must be based substantially on observations of their interactions with students. Are they communicating grades and performance well with students and parents? Are they setting fair and reasonable expectations and classroom norms? Are they designing curriculum and lessons that not only cover the content standards, but that incorporates students’ interests and questions?
Good teachers do these things regardless of the skill and successfulness of their students. Some students benefit from it and some do not. A student who is reading at the 4th grade level is unlikely to do well on an 11th grade state English exam, though a good teacher might help them significantly improve their reading comprehension over the course of the year. However, a student who is suffering financial insecurity at home, violence, chronic illness or depression will be much less likely to benefit from good teaching.
It is as much of a fantasy to think that good teachers can make all children succeed in school as it is to believe that welfare reform or slashing unemployment benefits will force people to get jobs.