Enhancing teacher productivity with information on students and skills: experimental evidence from Bangladesh

Working paper

Teachers may fail to address learning gaps because they do not know which students and skills need support. While exams are prevalent in every education system, these tests often go underutilized. We evaluate an intervention that provided grade-6 math teachers in 312 schools in Bangladesh with results from two brief tests, reports identifying topics and students in need of support, and an annotated answer key. At baseline, teachers could not estimate their students’ test scores, class rank, or spread, and they were twice as likely to interact with high- than with low-achieving students. The reports redirected instruction toward the needs they identified: flagged topics and students improved more on the intervention assessments. On independent endline tests, the program raised math achievement by 0.23 standard deviations (SDs) and reading by 0.17 SDs, with similar gains across the baseline distribution. It also improved three social-emotional skills by 0.10-0.14 SDs. Teachers lowered their estimates of students’ performance, distributed interactions more equitably, and emphasized more foundational content. At $5.06 per student, these findings show that actionable assessment results utilizing exam data in education systems can help teachers become substantially more productive at low cost.

Authors: Djaker, S; Ganimian, AJ; Angrist, N; Sabarwal, S

• Publication status: Published
• Publisher: Centre for the Study of African Economics
• Series: CSAE Working Paper Series
• Place of publication: Oxford, UK
• Publication date: 2026-10-02
• Paper number: csae-wps-2026-06
• Language: English