Data for: Information, Technology, and Market Rewards: Incentivizing Aflatoxin Control in Ghana

Published: 3 March 2021| Version 1 | DOI: 10.17632/y3tys8bfrd.1
Contributors:
Nicholas Magnan,
Vivian Hoffmann,
Nelson Opoku,
Gissele Gajate Garrido,
Daniel Kanyam

Description

Replication data and code for the paper "Information, Technology, and Market Rewards: Incentivizing Aflatoxin Control in Ghana"

Files

Steps to reproduce

Main analysis 1. Put STATA data file "groundnut_data_replication.dta" in working directory. 2. Set working directory and output director in .do file "groundnut_replication.do". 3. Manually select PDS LASSO (main paper) or ANCOVA (appendix) throughout the .do file (PDS LASSO is the default, search "ANCOVA" to find where to make changes". 4. For Lee bounds, manually set confidence interval on line 819 (90% is the default). Heterogeneity analysis (using R code from Chernozhukov et al. 2018, https://www.nber.org/papers/w24678). The Heterogeneity analysis was conducted using the file “groundnuts_ML.R” in R version 3.6.3 on a Mac. Some of the packages may not run using version 4.0 and newer. 1. Set destination directory for “gower” and “caret” packages at lines 21 and 23. 2. Set working directory at line 35. 3. At line 77, choose the appropriate STATA dataset to use (generated at the end of "groundnut_replication.do"). At line 109, select appropriate outcome variable. 4. At line 112, set appropriate treatment variable. 5. At line 124, set appropriate vector of control/heterogeneity variables.