Online Program

Return to main conference page

All Times ET

Program is Subject to Change

Monday, June 14
Mon, Jun 14, 1:30 PM - 3:30 PM
TBD
Experiments in Business Surveys: Opportunities and Challenges

Strategies for Embedding Experiments in Establishment Surveys: Examples from the US Department of Agriculture’s National Agricultural Statistics Service (307929)

Heather Ridolfo, National Agricultural Statistics Service 
Jaki S. McCarthy, US Dept. of Agriculture 
*Kathy Ott, National Agricultural Statistics Service 

Keywords: response rates, nonresponse, respondent burden, agriculture

With increasing demand for data on establishment populations, survey organizations continue to look for methods to increase response and reduce respondent burden. Embedded experiments can inform choices on establishment data collection strategies and questionnaire design. USDA establishment surveys include challenges to embedding research, such as skewed populations, small subpopulations, high survey burden, and sample overlap across surveys and time. Conversely, advantages to establishment researchers over household surveys include rich descriptive or administrative data, long response histories, dedicated respondents, etc. Strategies for embedding experiments in establishment surveys will be discussed using examples from the US Census of Agriculture testing. For example, known characteristics about establishments (such as type, size, or expected contribution to key estimates) can be used to select subsamples for inclusion, or strategies can be tailored and tested based on previous response dispositions (such as repeated non-response or the mode of previous responses). These strategies may help others embed experiments in their establishment surveys.