I am deeply troubled that many state laws continue to require consideration of change in annual accountability test score as an indicator of student growth for teacher evaluation. This is a patently indefensible policy for a variety of reasons. Now, don’t get me wrong. I believe it is possible to consider student achievement in the evaluation of teacher performance. But this is not the way to do it! Typically, such tests have not been validated for this purpose and such validation is required by accepted standards of sound assessment practice.
First, these tests typically sample broad domains of achievement in a very shallow manner. Test time constraints severely limit the number of items included, which then limits the range of learning targets that can be tested. Many targets are not tested at all, and those that are receive very thin coverage. As a result, the danger exists of a fundamental mismatch between any individual teacher’s instructional responsibilities and the targets that happen to be tested. When this happens, those teachers have no control whatsoever over student scores. To hold them accountable anyway is simply wrong.
Even more serious is the problem of the length of the pretest/posttest time period for measuring growth using annual test scores. Over the course of a year, an immense array of factors that are beyond teacher control can have immense impact on student growth, such as dealing with special needs students or doing without needed resources. These factors have been studied and their effects are well-documented relative to teacher effects. For this reason too, it is wrong to hold teachers accountable for student growth measured in this way.
Instead, when developing a fair, practical and effective teacher evaluation system, here are the key standards of good practice to be met:
- The specific learning targets to be used in tracking student growth for teachers must receive advance notice of a pending evaluation
- The criteria upon with the evaluation will be based must be explicit in advance, including the performance continua associated with each criterion
- Evaluators must receive proper training in applying those continua reliably
- Teachers’ work must be adequately sampled according to each criterion so as to lead to defensible judgments of performance
- Teachers must have an opportunity to describe any extenuation factors influencing performance
- Teachers are entitled to clear and complete communication of inferences drawn about their performance on each of the evaluative criteria
When student achievement is to be considered as one of those criteria, these 5 additional standards of sound practice must come into play:
- The specific learning targets used to evaluate a teacher’s impact must be specified in advance and agreed to be teacher and supervisor
- Those targets must fall within each teacher’s normal instructional responsibilities
- Each agreed upon learning target must be accompanied by an assessment plan for validly and reliably documenting student achievement of it
- That assessment plan must include a pre/post test evidence gathering process to come as close as possible to showing a cause and effect relationship between teacher’s instructional actions and student growth
- Teachers should have the opportunity to describe any factors beyond their control that may have influenced student achievement.
In any context where these standards of good practice are not met, there is a very real danger that incorrect inferences will be drawn about a teacher’s impact on their students. As a result, indefensible evaluative judgments that will have a detrimental impact on a teacher’s career might occur. When the evidence used to evaluate a teacher’s performance is change in annual test scores, such indefensible judgments are likely. This is just plain wrong. An extensive treatment of this problem can be found in my book, Defensible Teacher Evaluation (Corwin 2014), where I spell out the criteria to be considered in developing sound teacher evaluation programs.