This item first appeared on the Teaching Now blog.
A forthcoming study by University of Missouri researchers finds that accounting for factors like poverty when comparing schools could lead to a more “effective and equitable” teacher-evaluation system.
The study compares a so-called “proportional evaluation” system with two other methods of teacher evaluation: A basic value-added model and one in which the performance of individual students is compared to that of their peers. Unlike the other models, proportional evaluation takes into consideration factors outside of the classroom, including school resources and the socioeconomic backgrounds of students.
Researchers modeled each system using data from the Missouri Assessment Program test, including test scores from 4th through 8th grade students at more than 1,800 schools. The data were drawn from a period of five years, between 2007 and 2011.
Teacher ranking and evaluation systems are often criticized for favoring teachers in low-poverty, resource-rich schools. In particular, value-added models, which seek to isolate teachers’ impact on students’ test scores, have been scrutinized for their alleged potential to overstate the quality of teachers in well-resourced schools, since outside factors in such communities may create more fertile ground for student improvement. This difference in teacher rankings, in turn, can make it more difficult for struggling schools to attract quality teachers.
The Missouri study argues that proportional evaluation would avoid this problem. By comparing “equally circumstanced” schools and teachers, the model aims to regard each school’s performance in context. In a 2014 policy brief on the topic, lead researcher Cory Koedel, an associate professor of economics and public policy at the University of Missouri, notes that while higher poverty rates often correlate with lower value-added measures, that association disappears when each school is compared to like institutions.
Koedel writes that the use of proportional evaluation in schools would accomplish three main goals. First, it would more accurately represent the quality of teachers, allowing evaluations to be used more effectively. The system would also avoid demoralizing teachers, which Koedel says can happen with systems that “place teachers in competition with others against whom they have no hope of winning.” Finally, proportional evaluation would help level the playing field for high-poverty schools, allowing them to recruit quality teachers more easily.
While the researchers note that proportional evaluation could be seen as overrepresenting the progress of struggling schools, they suggest that consideration of schools’ test scores could avoid this potential problem. Such data, combined with proportional evaluations, would allow for a comparison of similar schools, identifying disadvantaged schools that are performing relatively well, while recognizing that these schools may still be underperforming in absolute terms.