What kind of process for evaluating teachers can possibly be devised by a determinedly reform-minded administration, a stubborn union and plaintiffs in a hostile lawsuit? As it turns out, a better kind than they've had up to now.
A tentative agreement reached last week would for the first time allow the Los Angeles Unified School District to use students' standardized test scores to rate teacher effectiveness, something many other schools across the nation are already doing. But just as important is what it wouldn't do — rely too heavily on how well each teacher's students scored on the tests. Those scores can be particularly volatile from year to year, and although they can help identify the very best and worst teachers, they are poor at differentiating among the 80% in between.
Instead, the evaluations would use a mix of schoolwide scores and the scores of individual teachers' students. That mix would be less specific to each teacher, and would include a factor that individual teachers can't fully control, but it would encourage teachers to work as a team. It's unclear how much the combined scores would count for; the district and the union agreed only that they would not be the "sole, primary or controlling factors" in evaluations. Supt. John Deasy has in the past said they should account for about 25% of a teacher's evaluation. That's reasonable.
In evaluating teachers, it is important to measure how well students are doing on standardized tests, but that's only one factor in the equation. One of the best things about the tentative agreement is its broad-based approach. Graduation rates also would count in some form in the new evaluations, as well as attendance and the use of disciplinary measures. In-class observations of teacher skills would be more heavily weighted than those other factors in measuring a teacher's effectiveness.
If the agreement is approved by teachers and the school board, L.A. Unified might end up with the most sensible and nuanced evaluation process in the country. It would meet the requirements of a judge's ruling in a lawsuit brought by EdVoice, a Sacramento-based education advocacy group, that test scores must be included. At the same time, it would improve on the blunt-instrument approach of the U.S. Department of Education, which has pushed the use of test scores so hard that some states and districts make those scores count for half or more of evaluations.
Under the agreement, experienced teachers might be evaluated as infrequently as once every five years, which seems too long. Just as important as what the evaluations would measure is how they would be used. The district should be helping teachers with the lowest evaluations to become more effective at their jobs and moving to dismiss them when they don't. Improved evaluations aren't worth much if teachers don't improve along with them.