Identifying effective classroom practices using student achievement data


Conducted by Thomas J. Kane, Eric S. Taylor, John H. Tyler, & Amy L. Wooten. 

Purpose of study

To estimate the extent to which observational measures of teaching effectiveness are related to student achievement growth and which observable practices best predict achievement gains.

Research questions

  • Do evaluations based on classroom observations identify teachers who are effective based on the test score gains of their students?
  • Which teaching practices best predict student achievement growth in mathematics and reading?


The researchers used teacher observation and student achievement data from the Teacher Evaluation System (TES) in Cincinnati, Ohio during the 2003-04 and 2008-09 school years. Teacher observations based on Charlotte Danielson’s Framework for Teaching (FFT) were conducted four times during each teachers’ observation year, focusing on Domain 2 “Creating an Environment for Learning” and Domain 3 “Teaching for Learning,” which include over two dozen specific teaching practices grouped into 8 standards of high-quality teaching. Overall, data from 207 math teachers and 16,196 of their students were included in the estimation sample, along with data from 365 reading teachers and their 20,125 students. A robust set of covariates were included to control for teacher and student characteristics.

Major results

  • Observable teaching practices predict differences in student achievement growth in both reading and mathematics. And increase by one point in overall classroom practices, a composite score based on all eight teaching standards in Domains 2 & 3 equally weighted, corresponded with a 1/7 of a standard deviation (SD) increase in reading achievement and 1/10 of a SD increase in math achievement.
  • Among students whose teachers have similar overall classroom practices scores, math achievement will grow more for students whose teachers are relatively better at classroom management (Domain 2) while reading achievement will increase more among students whose teachers are relatively better at engaging students in questioning and discussion (Domain 3, Standard 4).
  • Models focusing on the effectiveness of teachers in the years before and after their TES observations suggested that improvements to instructional practice as a result of participation were strongest for beginning teachers and promoted skills in classroom management especially.


Differences in student achievement based on teachers’ scores in overall classroom practices indicate that efforts to improve skills on all standards measured by the TES will benefit students. Concrete suggestions for improvement can be drawn from the descriptions of “Basic,” “Proficient,” and “Distinguished” classroom practice on all 8 standards of the TES. However, a one-point increase in teacher performance, from “Proficient” to “Distinguished,” for example, is a greater than 2 SD increase, suggesting that a one-point improvement is not as easy practically as the rubric suggests.

Teachers who must focus on a smaller number of practices for improvement due to limited time or professional development opportunities may want to focus on classroom management if they teach mathematics and on asking thought-provoking questions and engaging students in discussion if they teach reading, as results suggest the greatest impact of those practices, respectively, in each subject area. TES describes specific practices that contribute to teachers’ scores, thus providing details that teachers can use to improve their effectiveness at raising student achievement.

FFT focus

The TES observational rubric is closely based on the FFT, indicating similar results (and benefits) of using the FfT to observe teachers’ classroom practice. This study validates that the 8 instructional standards of Domains 2 & 3 of the FFT predict improvements in student achievement and provides preliminary evidence that particular practices are more or less important for promoting student achievement in particular subject areas.

This article is only available through the journal in which it was published. J. Human Resources July 1, 2011 46:587-613