There is a large body of research documenting substantial variation in achievement growth among students assigned to different teachers, but little progress has been made in linking teacher effectiveness with factors observable at the time of hire. A major cost associated with focusing solely on teachers' on-the-job performance is exposure of students to ineffective teachers. Any policy which successfully screens out ineffective teachers at the time of hire is likely to be extremely cost-effective. Furthermore, in contrast to the significant financial and political costs of altering teacher tenure or instituting pay-for-performance, new hiring processes can be relatively inexpensive (e.g., online survey and testing) and, because they naturally focus on non-union members, face little organized opposition.
Teacher effectiveness has emerged recently as one of the most important avenues of education reform. A widely read report by McKinsey & Company (2010) stresses the importance of selection, but does not offer much guidance to school or district leaders. This study represents an important step forward in developing a base of research evidence regarding teacher selection.
What are the Research Objectives?
We study the relationship between a host of non-traditional teacher characteristics collected during a typical hiring process in Washington D.C. Public Schools (DCPS) and subsequent teacher effectiveness in raising student achievement. In 2010-11, DCPS developed a new teacher selection system. Roughly 2,500 individuals applied in Spring 2011, 500 of which the district planned to hire.
In addition to providing standard information such as educational background, credentials, prior experience, and teacher exam scores, all applicants were required to complete a test of content knowledge for teaching, the Haberman Online Teacher Screener, and an essay about how they would respond to common teaching scenarios. Most candidates also participate in face-to-face interviews and "auditioned" in front of actual DCPS classrooms during the Spring.
Following the approach taken in by the investigators in prior work (Rockoff et al. 2011), we will regress teacher outcomes on the set of teacher characteristics and control variables. We will examine teacher absences, classroom observation scores and final evaluation ratings for all teachers, along with value-added measures for the subset of teachers in tested grades and subjects.
- What are the relationships between the various measures collected during the selection process (e.g., teacher characteristics, experience, attitudes, etc.) and how do they jointly determine teachers' success in completing the process and finding a job in the district?
- Which measures best predict on-the-job success of teachers subsequently hired by the district?
- Are certain combinations of measures especially predictive - that is, can we identify a set of measures that jointly obtain stronger predictions than one could attain by using measures independently?
- Are certain measures more predictive for a subset of teachers - e.g., is writing ability a better predictor of on-the-job success for English teachers than math teachers?
- Would selection of teachers based on any of the measures that predict on-the-job success disproportionately impact specific groups of applicants?
How is the Study Funded?
This study is funded by a private foundation.
Who will Use the Research?
The findings of our proposed research would directly inform the experimentation currently underway throughout the DCPS district, and will serve as an example of how school districts can take advantage of the opportunity afforded by the selection process to learn more about what drives teacher effectiveness in the classroom.
Brian A. Jacob, University of Michigan
Benjamin A. Lindy, JD, is executive director of new site development for Teach For America and the former Manager of Teacher Selection Design for the DC Public Schools. His interests center on the achievement gap, teacher quality, and the role of collective bargaining in public schools. He has published an empirical study of the impact of teacher collective bargaining laws on student achievement in The Yale Law Journal, which received the John M. Olin prize for the best student paper in law and economics. He graduated from Yale Law School in 2010.
Jonah E. Rockoff is Sidney Taurel Associate Professor of Business at the Columbia Graduate School of Business and a Faculty Research Fellow at the National Bureau of Economic Research. His interests center on local public finance and the economics of education. He has done research on the determinants of property taxation and expenditure in local public school districts, the impact of crime risk on local property values, the importance of teachers and teacher certification in determining student achievement, subjective and objective evaluation of teacher performance, and other educational policies such as charter schools, school accountability systems, class size reductions, and grade configuration. His research has been funded by IES and the Carnegie, Smith-Richardson, and Spencer Foundations.