The drive to rid schools of persistently underperforming teachers will intensify with the start of the fall semester. I agree that teachers who are ineffective don't belong in the classroom. What I find disturbing, however, is that efforts for the most part are based on a punitive approach: Identify these teachers sooner and fire them if they don't quickly improve. But a new study shows that many of these seemingly hopeless teachers can get better when given proper support ("Teacher Evaluations Found to Improve Midcareer Effectiveness," Education Next, Aug. 9).
Researchers Eric S. Taylor and John H. Tyler studied the Teacher Evaluation System that was implemented in Cincinnati in 2000. They looked at 105 mid-career elementary and middle school teachers, and concluded that assessing them by means of an innovative strategy helped them improve their instruction. This in turn translated into better performance in math for their students. According to Taylor and Tyler, the improvement was "equivalent to about 3-4 months of additional instruction or a gain of about 4.5 percentile points for the average student."
What I found most noteworthy is the way the evaluations are carried out. All new teachers are observed and scored four times during the school year. Three of these times are by a peer evaluator and once is by an administrator. Prior to receiving tenure in their fourth year, they are evaluated again. Thereafter, they are evaluated once every five years. Scores for tenured teachers result in eligibility for promotions or placement in a peer assistance program that can lead to termination. Taylor and Tyler found that this approach had the greatest impact on teachers who were the weakest before evaluation. I would be most interested in learning whether similar improvements apply to teachers of other subjects as well.
One of the reasons for the success of the program is that evaluators receive intensive training in how to accurately score teacher performance through the use of a rubric. Although no rubric is perfect, it has been found to be highly reliable. The cost of the Teacher Evaluation System is steep (about $7,500 per teacher evaluated), but it is cheap when compared with what it costs school districts overall to find replacements. The Alliance for Excellent Education places that number at about $4.9 billion every year. Teacher turnover, of course, varies from one district to another.
The strategy in Cincinnati contrasts sharply with the trend across the nation, whereby student achievement scores ranging in weight from 20 to 50 percent of a teacher's rating are replacing classroom observations ("The High Stakes of Teacher Evaluation," Education Week, Jun. 6). It's impossible to ever devise a strategy that will please all stakeholders. But in light of the counter currents buffeting public schools, I think the Cincinnati plan is the fairest way of determining which teachers deserve to stay in the classroom.