Foregone and Forgotten Learning: COVID-related Learning Losses from National Assessments in Bangladesh and experimental evaluation of a learning takeup program
How to cite
Ramachandra, Shrihari; Djaker, Sharnic; Sabarwal, Shwetlena; Angrist, Noam; Chang, Andres Yi. 2022. "Foregone and Forgotten Learning: COVID-related Learning Losses from National Assessments in Bangladesh and experimental evaluation of a learning takeup program." AEA RCT Registry. AEARCTR-0010184. https://doi.org/10.1257/rct.10184-1.1
Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.
* Primary investigator
Abstract
Bangladesh witnessed an overall 1.5 years of school closures due to the COVID-19 pandemic. Research from several countries provides evidence that school closures have generated steep learning losses. We conducted a large-scale learning assessment in Bangladesh comprising 7,218 grade 9 students to determine the current learning levels of students after 14.5 academic months of school closures. This was based on the National Assessment of Secondary Students (NASS) 2019 framework and covered a nationally representative sample of 300 schools. This study tracks 4997 students who took NASS 2019 and administered psychometrically linked tests in Math and Language (Bangla). The study includes Grade 8 and NASS 2019 grade 6 questions to estimate forgotten and forgone learning. The initial results show that for grade 9 students, 14.5 months of COVID-related school closures led to 26 months of learning loss among secondary students. Around 69 percent of learning losses were students failing to learn new skills (learning forgone), but 31 percent of learning losses were students forgetting skills they had already acquired (learning forgotten). There was also a sharp increase in learning inequality. The poorest students lost 50% more learning than the most affluent students.
The study also comprised a phone-based Interactive Voice Response (IVR) lesson to assess the demand for additional instruction. The students were randomly assigned to three groups (one control and two treatment arms) and provided a dedicated call number. The control group students received no message. Treatment 1 - The students were informed about the returns to education. Treatment 2 - The parents were informed about the returns to education. The lesson's content was carefully selected to be suitable for delivery over IVR.
Intervention
The study is a large-scale learning assessment in Bangladesh comprising 7,218 grade 9 students to determine the current learning levels of students after 14.5 academic months of school closures. This was based on the National Assessment of Secondary Students (NASS) 2019 framework and covered a nationally representative sample of 300 schools. This study tracks 4997 students who took NASS 2019 in grade 7 and administered psychometrically linked math and language tests (Bangla). The question booklet included LASI 2017 and NASS 2019 Grade 8 questions (approximately 20 questions) and NASS 2019 grade 6 questions (approximately 20 questions). This is to estimate forgotten and forgone learning separately. Learning inequality estimates were designed by benchmarking learning losses for those with low socio-economic status and by gender using a short survey at the end of the assessments.
The study also comprised a phone-based Interactive Voice Response (IVR) lesson to assess the demand for additional instruction. The students were randomly assigned to three groups (one control and two treatment) and provided a dedicated number to call. The control group students received no message. Treatment 1 - The students were informed about the returns to education. Treatment 2 - The parents were informed about the returns to education. The lesson's content was carefully selected to be suitable for delivery over IVR.
Experimental design
The novel data collection was from 300 schools in Bangladesh (sampled randomly from 1000 schools that were part of NASS 2019). Learning level data was collected from 7218 grade 9 students (4997 students had taken the NASS 2019, and 2221 new students were selected randomly).
Further, after the exam, the students were considered for a phone-based Interactive Voice Response (IVR) lesson to assess the demand for additional instruction. The students were randomly assigned to three groups (one control and two treatment) and provided a dedicated number to call. The control group students received no message. Treatment 1 - The students were informed about the returns to education. Treatment 2 - The parents were informed about the returns to education. The lesson's content was carefully selected to be suitable for delivery over IVR.
Design details
Cited by
No citation data available for this registration yet.