Classroom Assignment Practices in Urban School Districts Using Teacher Level Value-Added Systems

Recent econometric studies have begun to assess validity of value-added metrics in light of the non-random assignment of students to teachers. Value-added models (VAMs) assume student learning trajectories will remain the same across years, but students often experience changes that influence their learning. As a result, teachers’ ratings may be affected by factors that are unrelated to classroom performance or teacher quality. The concern that teachers may be incorrectly assigned attribution to student performance is at the heart of high-stakes accountability reform criticisms. To further investigate this issue, we conducted qualitative, exploratory research to determine how students are assigned to teachers in K–8 schools across three large, urban school districts from the perspective of principals. Our preliminary findings indicate that most classroom assignment is indeed non-random. However, many schools engage in “balancing” practices in attempt to ensure that the students in classrooms are evenly distributed across classrooms within grades. Within the balancing structure, schools may also “match” particular students to specific teachers (or other students) based on pre-identified individual characteristics or interaction tendencies. Our results suggest that information on school assignment practices could compliment value-added validity verification and will also be used to build two systems to support the correct attribution of students to teachers. Introduction and Background A recent Economic Policy Institute Briefing Paper (Baker et al., 2010) presented a critique of value-added modeling (VAM). That critique asserted that VAM is not yet ready to be used in high-stakes decision making at the teacher level. The paper identified factors that may influence student test score gains, such as team teaching, class size, student mobility, and school 1 Contact information: Sara Kraemer, 1025 West Johnson Street, Madison, WI, 53704 [email protected], +608-265-5624