Multi-Use Field Testing: Examples from the 2017 Census of Agriculture Dry Run
Jaki McCarthy
USDA’s National Agricultural Statistics Service,
Many large data collections will run a field test prior to operational survey operations. This acts as a smaller scale dry run of the upcoming survey. The 2017 Census of Agriculture (COA) is the largest data collection conducted by the USDA's National Agricultural Statistics Service, and is conducted once every 5 years. Prior to the 2017 COA, a field test was conducted as part of a multi method questionnaire evaluation. This large field test was used as a dry run of the questionnaire and data collection processes, but also had several other objectives such as producing records for the initial donor pool for imputation, comparing alternative versions of the questionnaire on data quality and response rates, evaluating NASS's ability to accurately identify records for a reduced short form, testing a new online system for web reporting, and testing the ability of a pre-survey contact to help identify census nonrespondents. A field test dry run is a rare opportunity that can be leveraged in multiple ways to make improvements to the subsequent data collections. Results from the 2015 COA field test, how they were used to benefit the COA and how best to use dry runs are discussed.