How Can We Encourage Teachers to Improve the Skills of Low-Achieving Students? Experimental Evidence from Bangladesh

How to cite

Djaker, Sharnic; Angrist, Noam; TM, Asaduzzaman; Ganimian, Alejandro; Sabarwal, Shwetlena. 2022. "How Can We Encourage Teachers to Improve the Skills of Low-Achieving Students? Experimental Evidence from Bangladesh." AEA RCT Registry. AEARCTR-0010203. https://doi.org/10.1257/rct.10203-1.1

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Sharnic Djaker* , Noam Angrist , Asaduzzaman TM , Alejandro Ganimian , Shwetlena Sabarwal

* Primary investigator

Registry ID
AEARCTR-0010203
DOI
https://doi.org/10.1257/rct.10203-1.1
First registered
Last updated
Status
On going
Trial dates
2022-05-01 – 2023-01-31
Countries
Private
JEL codes
C93, I21, I22, I25
Registry record
http://www.socialscienceregistry.org/trials/10203

Abstract

Researchers have argued that teachers in developing countries do not devote enough attention to low-achieving students primarily because they face incentives to focus on their high-achieving peers. Building on recent evidence from South Asia showing that primary- and middle-school teachers in South Asia overestimate their students' performance, this randomized evaluation aims to measure the impact of providing teachers with accurate information on their student's academic skills. The study has two treatment groups and one control group. Each group consists of 156 schools, totaling 468 schools. The schools in the treatment groups, 312 in total, will be handed over a report card immediately after baseline and midline data collection. Specifically, the business-as-usual control group will not receive any interventions; the external feedback group (T1) will receive two formative assessments administered by the research team and reports summarizing their results at baseline and midline, along with training for teachers on how to interpret the assessment results; and the internal feedback group (T2), will receive two formative assessments administered by teachers at baseline and midline, training for teachers on how to administer and interpret the written assessments and on how to administer one-on-one oral assessments to the three lowest-ranked students in the written assessments who are identified in the report cards. The study measures the impact on students' learning levels, teachers' instructional strategies, and the accuracy of teachers' beliefs at the endline.

Intervention

The external feedback group (T1) will receive two formative assessments administered by the research team and reports summarizing their results at baseline and midline, along with training for teachers on how to interpret the assessment results. The internal feedback group (T2) will receive two formative assessments administered by teachers at baseline and midline, training for teachers on how to administer and interpret the written assessments and how to administer one-on-one oral assessments to the three lowest-ranked students in the written assessments. These students are identified in the report cards.

Experimental design

Our evaluation of these interventions will be conducted in a representative sample of 468 schools in six districts selected from the eight divisions in Bangladesh. Schools in these districts will be assigned to one of the three experimental groups, one for each treatment group listed above or the control group that operates business as usual. Each group consists of 156 schools. Comparisons of outcomes for these groups will provide policy-relevant evidence on the efficacy of each individual program. The random selection of districts is stratified by their average performance on 2019 National Assessments (NASS) so as to draw three districts that have above-median aggregate NASS scores and three districts that have below-median aggregate NASS scores. Within each of the selected districts, 39 public-private-partnership schools (PPP) are selected at random after excluding Madrasas, non-PPP schools, and schools that have either very small or very large (< 5th percentile or > 75th percentile) class sizes. The 39 schools in each of the sample districts are grouped randomly into each experimental arm with 13 schools each. Within each sample school, one grade six section is selected at random, and 20 students are further identified at random for participation in the data collection for the RCT. The teacher responsible for teaching math for the selected grade six class is identified as the sample teacher who will receive the intervention (report cards).

Design details

Randomization method
Randomization was done using the statistical analysis software STATA.
Randomization unit
The unit of randomization is the school.
Clusters
468 schools
Observations
468 teachers and 9360 students
Arms
Control: 156 schools External feedback: 156 schools Internal feedback: 156 schools
Primary outcomes
We will evaluate the impact of the two interventions on: (a) Students' learning levels in math and reading; (b) teachers' instructional strategies; (c) Teachers' knowledge of student skill levels

Cited by

No citation data available for this registration yet.

Keywords