| Author(s) |
Barbara Bruns (Center for Global Development, Visiting Fellow) Leandro Costa (World Bank, Economist) Nina Cunha (Stanford University, PhD Candidate) |
| Date |
2016-05-13 |
| Country |
Brazil |
| Language |
English |
| Description |
This paper presents the initial results of a randomized evaluation of the Ceará program. Section 2 describes the context, the intervention, and the research questions. Section 3 describes the instruments used and the sample. Section 4 presents the results. Section 5 analyzes threats to the experiment and analysis we carried out to check the robustness of our sample and our results. Section 6 summarizes our conclusions and their implications for education policy in Brazil and other settings. |
| Abstract |
This study evaluated a program in the northeast Brazilian state of Ceará designed to improve teachers’ effectiveness by using an information “shock” (benchmarked feedback) and expert coaching to promote increased professional interaction among teachers in the same school. We show that the program significantly increased teachers’ use of class time for instruction (.29 - .35 SD), by reducing the time spent on classroom management (-.25 - -.28 SD) and time off-task (-.21 - -.24 SD). The program also increased teachers’ use of questions during their lessons, consistent with the coaching program’s goal of encouraging more interactive teaching practice. The treatment schools also registered an increase in student engagement. Finally, consistent with the program’s strategy of promoting greater interaction among teachers, the improvements in schools’ average results were achieved by reducing the variation in teacher practice. These preliminary results are probably confounded due to the likely Hawthorne effect; thus, when would be able to measure a sizeable effect on the students test scores, we can confirm the effectiveness of this intervention. |
| Table of contents |
1. Introduction
2. Intervention and Experiment Design
3. Instruments and Data
4. Results
5. Experiment Threats and Robustness Checks
6. Conclusions |
| Download |
http://datacatalog.ihsn.org//catalog/13139/download/107831 |
