Primary data collection and cleaning involve highly repetitive but extremely important processes that contribute to high quality reproducible research. DIME Analytics has developed
iefieldkit as a package in Stata to standardize and simplify best practices involved in primary data collection.
iefieldkit consists of commands that automate: error-checking for electronic Open Data Kit (ODK)-based survey modules; duplicate checking and resolution; data cleaning and survey harmonization; and codebook creation.
- DIME Analytics Bootcamp on Reproducible Research.
- Stata coding practices.
iefieldkitcurrently consists of four commands:
- Each of these commands can be used independently in a wide range contexts.
iefieldkitopen-source code is available on GitHub for public contribution and comments.
- To install the package, type
ssc install iefieldkitin the Stata command box.
One of the most important developments in economics over the past two decades has been the rise of empirical research, through primary as well as secondary data collection. The authors of
iefieldkit have developed the package to support data collection by researchers directly in a wide range of fields like agriculture, health, energy and environment, transport, financial and private sector development, gender, governance, and fragility, conflict and violence (FCV).
iefieldkit therefore supports general best practices in primary data collection from start to finish:
- Before data collection:
- During data collection:
- After data collection:
These four commands in this package make sure that inputs and outputs are significantly more human-readable by working with spreadsheets instead of Stata do-files. In doing so, they allow field personnel who do not specialize in code tools to understand and review the tasks involved in primary data collection.
iefieldkit thus recognizes the vital role played by field personnel in supporting data management and data cleaning even if they are not proficient in Stata.
Before Data Collection
In Open Data Kit (ODK)-based electronic survey kits, including SurveyCTO, survey forms (or questionnaires) are typically built in Excel using a specialized structured syntax. Before the research team starts with field data collection, they can use
ietestform to test Open Data Kit (ODK)-based electronic survey forms for common errors, as well as best practices for SurveyCTO-based forms.
Most ODK servers, including SurveyCTO servers, have a built-in test feature that tests the ODK syntax of a form when it is uploaded by the research team.
ietestform complements these built-in tests to ensure that the collected data is in a format that is easily readable in Stata, and warns users who use practices we have learnt are prone to data quality errors.
During Data Collection
While data collection is ongoing,
iecompdup allow researchers to test for, and resolve duplicate
entries in the dataset. The commands combine four key tasks to deal with duplicate ID values:
- Identifying duplicate entries.
- Comparing observations with the same ID value.
- Tracking and documenting any changes made to the identifying variable.
- Applying the necessary corrections to the data.
Together these commands ensure that the collected data will be a correct record of the sample, and can be merged with the master database. Both these commands were previously part of the
ietoolkit package, but have now been moved to
After Data Collection
After data collection is complete,
iecodebook allows the research team to automate the repetitive tasks involved in cleaning data before it can be analyzed. As the name suggests, the
iecodebook command is structured around Excel-based codebooks, which allows researchers to perform and document data cleaning tasks in Excel itself, instead of do-files. Therefore, codebooks allow researchers to document the cleaned data in a format that is both human and machine-readable.
iecodebook implements this through 4 subcommands:
iecodebook applyapplies rename, recode, and/or label commands to a large number of variables in the dataset.
iecodebook appendharmonizes two or more datasets, and appends them. That is, it allows two or more datasets to have the same variable names, labels, and value labels.
iecodebook exportcreates an Excel codebook that describes the current dataset. It can also produce an exportable version of the dataset which only contains the variables used in a particular do-file.
iecodebook templatecreates an Excel template that describes the current or targeted dataset(s), and prepares the codebook for the other subcommands in
- DIME Analytics (World Bank), The