Replication Materials for “Early-Childhood Exposure to Cyclone Sidr and Later Foundational Learning in Bangladesh”
Description
This repository contains replication materials for the manuscript “Early-Childhood Exposure to Cyclone Sidr and Later Foundational Learning in Bangladesh.” The study examines whether developmental timing at Cyclone Sidr is associated with later foundational learning outcomes among children in Bangladesh. The empirical analysis draws on the Bangladesh Multiple Indicator Cluster Survey (MICS) 2019 foundational-learning module and covers 19,008 children aged 7–14 across all 64 districts. Survey information is combined with official district-level Cyclone Sidr damage classifications, 2007 monsoon-flood classifications, and district-level pre-Sidr characteristics derived from MICS 2006. Twelve districts are classified as Sidr-affected, comprising four severely affected and eight moderately affected districts. Developmental exposure is distinguished for children who were in utero or aged 0, 1, 2, or 3 at the time of Sidr. Outcomes include standardized composite learning, reading, and numeracy measures and indicators for foundational reading, foundational numeracy, and mastery of both competencies. The statistical workflow uses survey-weighted cohort-by-location models with district and birth-month fixed effects, adjustment for 2007 flood exposure, district-clustered inference, and extensive robustness analyses, including wild-cluster bootstrap inference, false-location permutation tests, multiple-testing adjustments, migration restrictions, alternative weighting, and exposure-severity analyses. The public replication archive contains Stata 14-compatible analysis code, documentation, district-level supporting data, reproducibility checks, statistical tables, figures, and non-restricted derived outputs. Original and child-level MICS microdata are not redistributed and must be obtained directly from the authorized data provider under the applicable access conditions.
Files
Steps to reproduce
Download and extract the replication ZIP while preserving the folder structure. Read 00_READ_ME_FIRST.txt before running the analysis. The replication code requires Stata 14.0 or later. Install the community-contributed Stata commands identified in the README. The underlying Bangladesh MICS microdata are not redistributed with this repository; researchers wishing to rerun the complete analysis must obtain authorized access directly from the UNICEF MICS data portal and follow the data-preparation instructions included in the package. Place the authorized analytical input in the location specified in 01_input/main/README_RESTRICTED_INPUT.txt. Run 00_VERIFY_PACKAGE.do to verify the project structure and required files. 00_RUN_QUICK.do may be used for a technical diagnostic run, whereas 00_RUN_FINAL.do reproduces the full analysis. The final analysis uses 9,999 wild-cluster bootstrap replications and 4,999 false-location permutation assignments. After completion, confirm that FINAL_REPLICATION_STATUS.txt reports PASSED and compare the generated outputs with OUTPUT_CROSSWALK.txt.
Institutions
- Jagannath UniversityDhaka Division, Dhaka