In many school districts, classroom observations make up as much as 75 percent of teachers’ evaluation scores, according to a new study published in Education Next. And these scores predict a teacher’s ability to raise student test scores the following year, as measured by value-added models. With that in mind, analysts did a deep dive into the observation practices of four school districts. They found that score stability and the quality of the information gathered improved as the number of individual observations increased, and that evaluations by trained, independent, outside evaluators (instead of principals) were more predictive of the following year’s value-added gains. The report recommends that districts observe teachers at least two or three times annually, using an outsider at least once. Moreover, the study draws attention to the latent bias against teachers with lower-achieving students or who teach in struggling schools. Although value-added models are careful to control for students’ backgrounds and achievement levels, there’s often no such adjustment for classroom observations. (Simply put, it can be much harder to teach a great lesson when the kids are below grade level or unruly.) Unchecked, this can push teachers to avoid these roles, widening the achievement gap. The authors suggest that states conduct statistical analyses to control for these variables. If districts hope to retain and improve their teaching force, making the most of their teacher observations is a good place to start.
SOURCE: Grover J. Whitehurst, Matthew M. Chingos, and Katharine M. Lindquist, “Getting Classroom Observations Right,” Education Next (Winter 2015).
