As mentioned yesterday, Ohio is looking to transform how teachers are paid by basing pay raises on a teacher’s evaluation, an evaluation system that bases half a teacher’s score on gains in student achievement. Similar proposals are popping up all around the country. However, only a handful of states and districts (such as Houston, Denver, and Washington, D.C.) have actually implemented such a system.
When it comes to transforming teacher evaluation systems, what gets most of the attention is how measures of student achievement are included. In most proposed systems, student achievement measures typically account for between 25 and 50 percent of a teacher’s overall score, similar to what is proposed in Ohio. But almost no attention is given to the other measures that account for the rest of the teacher’s score.
For those of you who have checked out our Building a better evaluation system report, you know that teachers should be evaluated using multiple measures to provide a more accurate assessment of a teacher’s true effectiveness. Such measures include teacher observations, lesson plan reviews, and teacher portfolios along with measures of student achievement. Currently, teacher observations are the most common method by which teachers are evaluated, and our report found these observations (as practiced) are neither very accurate nor helpful.
However, Ohio doesn’t have to look far to find a great example of how using teacher observations can be an effective tool in evaluating teachers, if done right. A new study in this month’s Education Next found that Cincinnati Public Schools’ Teacher Evaluation System (TES), which evaluates teachers using a comprehensive observation system, reliably predicts the change in student achievement. That is, students of those teachers who had the highest evaluation scores made significantly greater gains than students who had teachers who earned lower scores. So if done correctly, teacher observations can be an effective tool to determine how effective a teacher is at improving their students’ achievement.
Which raises the question, why not just use teacher observations? The answer is that no measure of teacher effectiveness is perfect. By incorporating multiple measures into evaluations, the chances of misidentifying a teacher is much lower than if just one measure was used.
Furthermore, different measures provide different feedback. For example, student achievement data enables teachers to determine which students they had the greatest impact on, while teacher observations provide teachers with information about what instructional techniques they need to improve.
In addition, a high quality comprehensive teacher observation system like Cincinnati’s requires significant resources. Hence, Cincinnati teachers are not evaluated every year as a system like that proposed in Ohio would require. An evaluation system that requires annual observations is not likely to provide as accurate results as Cincinnati’s current system, so even then other measures would be needed.
Based on this study, there will be those who will argue that only observations are needed to evaluate teachers. Others will argue for going with student achievement data since it requires fewer resources. But both would be wrong. The best way to evaluate our teachers is to use multiple measures. – Jim Hull