Sunday, July 05, 2009

TEACHER EVALUATIONS & THE LAKE WOBEGON EFFECT

by Terry Holliday | Superintendent - Iredell Statesville [NC] Schools [20,000 K-12 students)|2008 Baldrige National Quality Award Recipient | Posted on EdWeek LeaderTalk at June 28, 2009

I caught Secretary Duncan on NPR this week talking about teacher evaluations and other key issues surrounding education reform. Secretary Duncan talked about several studies that were recently featured in Education Week. The studies major findings that teacher evaluations reflect a Lake Wobegon effect. Almost 99% of teacher evaluations studied reflect teachers were at or above average. In other words, all of the teachers being evaluated are meeting or exceeding standards. Sec. Duncan's question rings true - if all of our teachers are meeting or exceeding standards then there is little to no variation in teacher distribution. Another more distrubing question is this - if all of our teachers are meeting or exceeding standards, then why are many students failing or dropping out of school.

About the same time Sec. Duncan was talking about teacher evaluations and the need to improve evaluations, I was having end of year reviews with principals in our system. Our school system deployed new teacher and principal evaluation instruments this year. During the end of year reviews, the conversation focused on performance of the school in the areas of student learning and how professional development impacted student learning. Also, I asked a great deal about how the principals used the teacher evaluation instrument to analyze the variability in student learning among and between teachers at the same grade level and subject. What I discovered was our principals needed more coaching and support to have these conversations with teachers.

Also, last week I was working with leaders from several different states and discovered that some states do not allow the connection of student learning data with teacher evaluation data. While NC does not prohibit this use, it is certainly only one part of a comprehensive teacher evaluation instrument.

While Sec. Duncan seemed to focus on the need to evaluate teachers and find out those who are low performing, I would prefer we focus on an improvement instrument for teachers and principals, By connecting the key instructional strategies that impact student learning and then providing focused professional development, coaching and support, I believe that 95% or more of our teachers can be successful. Why 95%? That is the core philosophy of a systems based approach. In most cases, it is not the people that are the problem. It is not the people that are creating the variation in a process. It is the system and the process itself. For our system, we will focus on the process of teacher evaluation and provide principals with the coaching and support needed to continue to improve student learning outcomes. NC has focused standards for school boards, superintendents, principals and teachers that are systems based and focus on continuous improvement.

1 comment:

  1. The education profession has extensive research on what good teaching looks like, and how students behave when they are learning (or not learning). Research on class learning time, time on task, levels of questions asked by teachers and students, teacher talk versus student talk, etc., etc provides solid guidance for either providing useful feedback to teachers as they gain in skills or for evaluation. Why are we jumping from student test scores to teacher evaluation? There is/should be a connection, to be sure, but a piece of the puzzle is missing.

    It's easy to collect the in-class observation data on these researched best practices and resulting student behaviors (increased student engagement) using such tools as the eCOVE Observation Software (www.ecove.net). The eCOVE data collection tools will track the fidelity of implementation of sound teaching practices, and the effectiveness of the interventions on student in-class learning related behaviors. It's exactly this data that can help provide an explanation of why teachers with positive 'evaluations' (most often a checklist record of observer judgments -- not objective data) have students who score poorly on outcome measures.

    Having the objective data on what is happening in our classrooms would provide the basis for cause and effect examination of teaching and learning, rather than point fingers in all directions. Without this data, easily gathered, we're trying to link up an observer's opinion about what constitutes good teaching with the hard student test data, and that will never resolve the issues.

    ReplyDelete