When I arrived as the superintendent of the New London, Conn., schools in 2009, the local board of education made it clear to me that its primary goal was to improve student achievement and close achievement gaps.
Over the past five years, we’ve worked hard to improve student reading and math skills. Our efforts include the following targets: Eighty percent of students should achieve at “goal” level (above “proficient” and below “advanced” on the state’s standards-based achievement tests) in English/language arts and mathematics; and fewer than 5 percent of students should drop out.
This month, U.S. News & World Report gave New London High School a “bronze” ranking for performing better than expected on state tests and better than the state average for the least-advantaged students.
As a small, urban school district located along six square shoreline miles in southeastern Connecticut, New London has approximately 3,300 students. Fifty-eight percent are Hispanic, and 27 percent are African-American. Eighty-two percent of our students live below the poverty line.
Throughout my 40 years in public education, I have come to believe that the key to improving student achievement is improving instruction and changing adult behavior in schools.
Therefore, one of our first steps on the way to turning around our district was to create a new teacher-evaluation system. For many educators, enduring the evaluation process is viewed as falling somewhere between contracting polio and the bubonic plague.
Our previous teacher-evaluation system was full of vague categories that included “planning,” “lesson development,” “classroom management,” “student performance,” and “professionalism.” In interviews, several teachers said they had not really been evaluated in years. Principals reported to me that they had been unwilling to pursue termination because too often previous district leadership told them to back off in the final stages of the termination process.
With the district’s objective of improving student achievement, I knew we had to overhaul our evaluation system.
As a guidepost, we used the components of Connecticut’s Common Core of Teaching—research-based instructional behaviors that have been shown to improve student achievement. A team of teachers and administrators developed our evaluation system over the course of a year. Over the four years since then, we have regularly clarified, modified, or tweaked the process, following feedback from teachers, administrators, and monthly meetings with the New London Education Association.
Our revised evaluation contains expanded guidelines on teacher performance so that we can link the evaluation to more than test scores. We built the performance evaluation around observable behaviors known to improve student achievement, including set schedules for classroom observations, professional-development standards, and access to growth opportunities. We use tiered levels of intervention to encourage the professional growth of teachers. Parent and student feedback is built in.
In 2013, the Connecticut State Department of Education recognized the work we had done in refining our teacher-evaluation system and, in an unusual move, granted the district a waiver from the state’s evaluation system. We are one of the few districts in the state able to use our own.
The key to improving student achievement is improving instruction and changing adult behavior in schools.”
New London’s teacher-evaluation system received perhaps the rarest of praise, when Christopher Teifke, a Connecticut Education Association representative, recently told me that our evaluation system, when done properly, had succeeded in connecting improved instructional practice and student achievement.
In the course of my career in education, I have learned what is essential for creating successful teacher-evaluation systems.
Useful and effective evaluation systems need several basic building blocks. They require clarity and observability of standards in practice, a common language, calibration, consistency, documentation, a check for teacher comprehension, and an understanding of the “reasonable person” rule often expressed in state administrative hearings and state courts in connection with teacher- and other personnel-evaluation cases. Administrative-hearing officers and judges frequently ask the following of administrators evaluating teachers and staff: “Could a reasonable person have been expected to know what you wanted them to do based on your description of what was expected of them?”
Standards must be observable. Having worked with emerging teacher-evaluation systems in Florida, Massachusetts, and Connecticut, I have learned that the process too often goes undefined, and that this is a cause of great consternation to classroom teachers. Standards for performance are often not identified clearly enough so that teachers and observers know what they look like in the classroom.
A common language must drive teacher evaluation. In addition to clarity and the ability to observe teacher behavior, it is equally important to build a common language. Building such a common language often takes three to five years of continuous professional development.
Active listening is key. In court and administrative hearings, state hearing officers and judges often find that the evaluator never checked to see if the teacher understood what he or she was expected to do differently.
In many direct conversations between administrators and teachers, the phrases “you know what I mean” and “OK” are used as code to check for people’s understanding.
Don’t fall into the trap of assuming that someone understands what you are talking about. You are much better off saying something like, “Please tell me in your own words what you think I’ve just said.” You can then use the individual’s response to clarify your intention.
Be consistent and calibrate. This gets to a simple question that addresses a very challenging set of problems in evaluating teachers: If two people were to observe the same classroom at the same time, would they see the same thing? The answer too often is no. A common language can build consistency.
Validity and reliability are musts. An observation of a teacher would pass the validity test if she were being judged on the basis of a clear understanding, by both parties, of what’s expected of her in the classroom. Reliability is defined by consistent behavior.
Failure to understand these essentials of the teacher-evaluation process can lead to educators’ not knowing what is expected of them, resulting in ineffective instruction and foundering student achievement. Clearly, that’s a result none of us seeks.
A version of this article appeared in the May 21, 2014 edition of Education Week as How to Build a Better Teacher Evaluation