Data Processing
Data Cleaning
Mathematica worked closely with a local data collection partner to train interviewers and closely monitor the data collection effort to identify challenges and address them while data collection was ongoing.
Upon receipt of the complete datasets, Mathematica conducted additional cleaning to code "other" responses, correct out of range responses, address item nonresponse and inconsistent skip patterns, and conducted merges between different datasets if necessary.
Data Entry Process
Data entry was conducted manually at the data collection firm's headquarters using a standardized Excel-based data entry template for surveys to reduce errors from out-of-range values and incorrect skip logic. Double data entry was conducted for all surveys and traffic counts, followed by correction of the final datasets to correct errors from the original entry.
Documentation in PDF