Study urges more teacher evaluations

SEATTLE – Once-a-year evaluations aren’t enough to help teachers improve, says a report by the Bill & Melinda Gates Foundation.

And school districts using infrequent classroom observations to decide who are their best – and worst – teachers could be making some big mistakes, according to the second part of a multiyear study from the foundation.

Preliminary results were posted online Friday at www.metproject.org.

Good teacher evaluations require multiple nuanced observations by trained evaluators. Those results should be combined with other measures, such as student test scores and classroom surveys, to gather enough information to both evaluate teachers and help them improve, the researchers found after nationwide experiments involving thousands of teachers.

The most common teacher evaluation method used by school districts today – a single classroom observation once every few years – has only a 33 percent chance of resulting in an accurate assessment of a teacher, the researchers found.

“This confirms what many teachers have been saying for years: That when high-stakes decisions are being made, school districts should allow for more than one observation,” said Tom Kane, deputy director of the Seattle-based foundation’s education program and leader of the research project.

Teachers across the nation are getting too little feedback and are being left alone to figure out what they need to do to improve, said Vicki Phillips, director of the foundation’s education program. If the nation is serious about improving the quality of its teachers, improving evaluation and feedback should be an important element of that effort, she said.

For the past two years, the foundation has been working to build a fair and reliable system of teacher evaluation and feedback to help teachers improve their craft and assist school administrators in their personnel decisions.

The report comes amid efforts across the country to change the way teachers are evaluated. Most of the new systems are a direct result of a call by the federal government for education reform, and many are finding implementation of the evaluation systems difficult.

The core of the Gates Foundation study was a collection of digital videos of more than 13,000 lessons in classrooms of teachers who volunteered to be studied.

The classrooms are being studied in Charlotte-Mecklenburg Schools, the Dallas Independent School District, Denver Public Schools, Hillsborough County Public Schools in Tampa and St. Petersburg, Fla., Memphis City Schools, the New York City Department of Education and Pittsburgh Public Schools.

The main conclusions of this report are as follows:

• High-quality classroom observations require clear, specific standards, well-trained and certified evaluators and multiple observations per teacher.

• Classroom evaluation is not enough. That information should be combined with student feedback and data on improvement in student test scores. Combining the three kinds of evaluations offsets the weaknesses of each individual approach.

• The different evaluation methods still need to be refined, but they’re better than what most districts are using now.

Student surveys, also being piloted by the foundation in school districts around the nation, are not popularity contests, Kane said. They focus on class experiences and ask students to talk about things like whether they are being challenged and engaged.

College professors have been evaluated by their students for years. Kane, who is also a Harvard professor, said he thinks school teachers could learn to appreciate that feedback as well.

“One thing I’ve learned is once you show people the questions, much of the hesitance fades away,” he said.

Kane emphasized that the main finding of this research is that the more information gathered about any one teacher, the better chance she or he will be given an accurate evaluation that helps improve teaching practices.

Districts that don’t have the money to completely change their evaluation systems can take some first steps that the foundation and the school districts thought would make a meaningful difference. Those ideas include:

• Better training and certification for observers, including videotaping lessons and having more than one person evaluate a teacher.

• Student surveys to supplement other methods of evaluation or as a way to help teachers and their mentors work together.

• Convene meetings between teachers and administrators to start collaborating on improving the evaluation system.

• Look at the foundation’s research results and start a conversation about which parts of a teacher’s practice are most closely linked to student success. Focus professional development on those areas.

Randi Weingarten, president of the American Federal of Teachers, expressed concern that too much emphasis is being placed on evaluating teachers and not on improving their performance.

“Until we make a commitment to develop evaluation systems that are first and foremost about continuous improvement and professional growth, we will continue to struggle in our efforts to provide every child with a high-quality education,” she said in a written statement.