Primary data

Especially if you work with primary data, you have to ensure the data is de-identified as early as possible:

• Consult with the data protection officer before collecting/accessing primary data to ensure data protection regulations can be fulfilled
• Check the DIME and J-PAL guidelines for de-identification of primary data
• Check the section on random variables to ensure you create reproducible, unique identifiers
• Check out the ietoolkit with the command ieboilsave to perform checks on de-identification

Also, it is even more important to ensure that the data is well-documented and prepared. Some helpful tools for data cleaning are provided by the iefieldkit . Check out the section on interaction with the system for code to automatically produce codebooks.