To the Editor:
Last month in New York City, the media fought for access to information about “the job performance of public employees,” yet what they won, the public release of teacher value-added scores, is only part of the overall picture of teacher performance. Though the newly released teacher value-added scores are considered “the best available measure,” we wonder why they are not being offered in conjunction with other data points that illustrate a fuller picture of teacher performance.
The danger of the public release of this data is that its authoritative identity as a numerical, statistical wonder-number may make it seem as if the number itself carries meaning and authority on its own, disconnected from the context in which teaching occurs and the scores themselves were presumably generated. While advocates of value-added data have never claimed the scores should be used as a singular measure of teacher effectiveness, the scores are often presented to the public, in court battles and online databases, without context or direction on how to interpret or situate them.
As we recently wrote (“Public Displays of Teacher Effectiveness,” edweek.org, Dec. 15, 2010), “The way we choose to write and talk about [value-added measurement] may make all the difference.” When we talk about value-added scores as being only one part of an overall picture of teacher effectiveness, we are immediately prompted to ask additional questions and to engage with the process rather than just make a first-pass diagnosis of effective teaching. We can ask how such scores align with other student-achievement measures, teacher observations, trajectories of professional growth, and particular contexts and circumstances.
We hope that the public reads the scores in ways that spark conversations about other sources of information and other features and layers of the story of student achievement and effective teaching. As Lee Cronbach, the father of a method for determining reliability in educational and psychological testing has written: “The majority of studies of educational effects—whether classroom experiments, or evaluations of programs, or surveys—have collected and analyzed data in ways that conceal more than they reveal.” Cronbach was referring to studies conducted 35 years ago, but his comment holds so long as single data points are allowed to dictate a version of reality without the context and layers of information that act to situate and enrich their meaning.
Value-added is a starting point for conversations about and further examination of effective teaching, but on its own it may conceal more than it reveals.
Rachael Gabriel and Jessica Lester
University of Tennessee
Knoxville, Tenn.
