How WFP Strives to Ensure Data Quality in Its Assessments
WFP relies on good data to make sure its programs are effective and reach the people who need them most. To ensure this data is as reliable as possible, WFP encourages a comprehensive process that covers the entire assessment cycle – from planning to analysis. The idea is that it is much easier to prevent mistakes from happening in the first place than to try and fix them later!
These are the recommended guidelines WFP aims for:
Careful Planning Before Data Collection Starts:
File Management: WFP encourages organizing data using tools like SharePoint and the WFP Data Library. This aims to ensure that all important files are stored safely, and everyone knows which version is the most up-to-date. Think of it like having a well-organized filing cabinet for all assessment documents.
Questionnaire Design and Programming: WFP aims for survey questions to be clear, relevant, and easy to answer. It is recommend using the WFP Survey Designer to ensure the updated modules are used and testing the surveys on tablets to ensure they work smoothly.
Enumerator Training and Testing: WFP suggests carefully selecting and training enumerators. They should understand the survey questions, how to ask them correctly, and how to record the answers accurately. Testing them after training is recommended.
Communication Plan: WFP encourages a plan for how the Data Analyst can communicate with the field teams. This includes ways for the team in the field to get support quickly and address any unexpected issues that come up.
Monitoring During Data Collection:
High-Frequency Checks (HFCs): While the data is being collected, WFP encourages regularly checking for things like missing answers, strange patterns, or answers that do not make sense. By communicating these back to the field teams, mistakes can be caught and corrected early on.
Cleaning and Organizing After Data Collection:
Data Cleaning: Once all the data is collected, WFP encourages cleaning it by addressing typos, inconsistencies or potential errors. This is done to make sure that the final dataset is as accurate and reliable as possible.
Data Management: WFP recommends creating a package that contains all the information on how the survey was conducted, the raw data, the cleaned data as well as cleaning and analysis syntaxes.
By following these steps, WFP strives to make sure that its data is as reliable and accurate as possible, which helps the organization make the best decisions possible to address hunger and food insecurity.
See the Data Quality Guidance for more details.