Using Teacher Performance Assessments For Human Capital Management

April 10, 2010

Tony Milanowski

Tony Milanowski

What aspects of teaching most influence student learning?

What specific local strategies most improve student achievement?

States or districts designing systems for assessing teaching practice need to answer these questions. In doing so, they begin to develop a competency model to guide teaching assessment design, and more broadly, their entire teacher human capital management system. Competency models can reflect both what is common to good teaching across states and districts, and the local vision of instruction and local improvement strategies. A competency model for teacher performance usually specifies key performance domains. Within each domain are specific behaviors. For example, most competency models include a domain pertaining to delivery of classroom instruction. Within that domain will be numerous instructional behaviors such as using assessments in instruction. Examples of other common domains are instructional planning, classroom management, interactions with staff and parents, and professionalism.

Developing a teaching assessment for use in a strategic human capital management system is a complex undertaking. The paper by WCER researcher Tony Milanowski and colleagues can help districts and states think about how to get started and what resources to gather.

Milanowski and colleagues set out to identify the teacher competencies common to seven state-of-the art teaching assessment systems (see below).

  1. The PRAXIS III teacher licensing performance assessment 
  2. The Performance Assessment for California Teachers (PACT) 
  3. The Formative Assessment System Continuum of Teacher Development (University of California, Santa Cruz)
  4. The Framework for Teaching (Charlotte Danielson) and its implementation by the Cincinnati Public Schools
  5. The Teacher Advancement Program (National Institute for Excellence in Teaching)
  6. The National Board for Professional Teaching Standards (NBPTS) assessment system
  7. The Classroom Assessment Scoring System (CLASS; University of Virginia)

They assessed the similarity of the systems in two ways. First, they counted how many performance dimensions in each system referenced eight important competencies that underlie the kind of teaching that is often regarded as most likely to improve student achievement. Second, they laid out the system rubrics side by side to identify competencies that were mentioned in at least three of them. The two approaches yielded substantially similar results.

Milanowski found that the most comprehensive assessment systems were the Framework for Teaching and the Formative Assessment System Continuum of Teacher Development (University of California, Santa Cruz).

Less comprehensive, but arguably more focused on the specific aspects of instruction likely to influence student achievement, were the National Institute for Excellence in Teaching’s Teacher Advancement Program and the Cincinnati school district’s system (itself based on the Framework for Teaching). Milanowski says these systems may represent a worthwhile trade-off of breadth for depth and focus on competencies key to the developers’ strategies for improving student achievement.

Based on this comparison study, Milanowski and colleagues propose a preliminary set of specifications that states or districts may consider when deciding how to assess teaching. The specifications attempt to present what a state-of-the-art teaching assessment system would look like, based on the best features of the systems reviewed.

Milanowski and colleagues do mention a few cautions. First, no matter how well designed the competency model and the assessment processes, if the assessment system is not implemented as intended, it is unlikely to realize the desired benefits. Second, there is unlikely to be one best data collection approach for all of the uses of teaching assessment. Third, while there probably should be multiple methods of data collection customized to specific uses, all of the methods should be based on a single competency model in order to preserve the alignment of the system.

With these cautions in mind, these WCER researchers outlined the following eight specifications for a state-of-the-art district teaching assessment system:

The system should be based on a competency model that includes the drivers of student achievement and the actions teachers need to take to implement district strategies for improving student achievement.

The emerging state of the art in teaching performance assessment suggests that states or school districts can reliably evaluate teaching practice and use these assessments as part of a coherent strategy for identifying, developing, and retaining effective teachers. A review of seven assessment systems by WCER researchers describes the emerging consensus on the important dimensions of teaching performance and the procedures that contribute to reliable assessment.

The competency model and the basic structure of the rubrics need to apply to all grade levels, career levels, and subjects. However, to ensure that grade- or subject-specific instructional strategies or skills are included, the model should use customized language in performance dimension definitions and rubrics when needed.

If the intention is to assess teacher content knowledge and pedagogical content knowledge for consequential purposes (e.g., decisions about tenure or career ladder progression), the assessment system should include standardized performance tasks that ask teachers to demonstrate this knowledge. These tasks can be planned to assess the same key content-related knowledge for all appropriate teachers.

Content-knowledgeable assessors with experience in the relevant grade levels should be used to judge teaching performance. The use of such assessors promotes more valid judgments and adds credibility to ratings and to any coaching provided using the assessment.

When assessment is done for consequential or summative purposes, the system should include features that promote reliable and valid measurement.

The system should include features that promote teacher learning.

The assessment process should be standardized and documented.

The system should use technology to reduce workload and improve administration.