The idea of evaluating preservice candidates by their students' academic achievement is jarring the world of teacher education. Such policy has become a requirement in state approval protocols, The National Council for Accreditation of Teacher Education (NCATE) standards, and is even showing up in federal legislation as a proposal that professional programs be evaluated on the academic achievement of the students of their graduates.
In our state, Washington, teacher education programs are under a mandate to demonstrate that their candidates make a positive impact on student learning. The phrase "positive impact on student learning" occurs in no less than 19 different places in the Washington Administrative Code governing the approval of professional programs. It is a requirement for the mission of teacher preparation, a condition of program approval, an element of accountability, and a separate condition in each endorsement program. Several of these requirements specify that programs demonstrate impact using multiple methods of measurement over time. Another requires programs to retain candidate work samples to show positive impact. A new state pedagogy assessment will evaluate candidates by focusing primarily on what students are doing during lessons rather than on teacher behavior ("Performance-Based," 2004).
At the federal level, proposed requirements of states under Title II of the Higher Education Act make the remarkable step of asking that states rank teacher education programs by the test scores of the students of their candidates. At least one state, Louisiana, appears to be moving in this direction (Archer, 2004). And our own recent NCATE accreditation review explicitly emphasized the requirement that institutions demonstrate how "teacher candidates ... have a positive effect on learning for all students" (NCATE, 2002, p. 16).
In this article, we document our own unsettled responses to what we refer to as the "positive impact mandate." Although a host of research efforts today aim to illuminate how teacher education might be more evidence based, we offer an inside account of how one program responded to a specific state policy. We report on informal fact finding we did to determine how other teacher institutions in our state were responding to the positive impact mandate, we describe our efforts to meet the demands of NCATE evaluators, and we highlight dilemmas we continue to face in connecting the complex learning of beginning teachers to the learning of their students. As a documentary account, we emphasize our own grounded responses to issues teacher educators will continue to face as accountability reforms inexorably bear on teacher education.
FRAMING AND REFRAMING
The positive impact mandate differs in substance from long-standing practice in teacher education emphasizing the evaluation of instructional methodologies, management of the classroom, and professional behavior. Teacher evaluation, in other words, has typically been grounded in the behaviors, qualities, and growth that teachers exhibit rather than in changes exhibited by students--although supervisors and mentor teachers also have looked to student response, especially lesson engagement or time on task, as a measure for effective teaching.
On one hand, a positive impact emphasis seems reasonable enough. Not only are teachers hired to affect student learning, but the new policy language makes explicit what many reformers have long wanted--an emphasis on student understanding rather than on the act of teaching. The literature on teaching for understanding, for example, is replete with stories of sound pedagogical methods or curricula that yield conceptual misunderstandings among students (e.g., Schneps & Sadler, 1987). In addition, constructivist conceptions of learning, widely supported in preservice education literature, "involve fundamental shifts ... from focusing on dispensing content to placing students efforts to understand at the center of the educational enterprise" (Windschitl, 2002, p. …