Are Shoppers Representative of the Population? Using Geofenced Grocery and Convenience Stores to Represent the Population

• Activity Number: 365 - SPEED: Innovations in Survey Sampling Designs: Administrative Data, Record Linkage, Non-Probability Samples, and More
• Type: Contributed
• Date/Time: Tuesday, July 31, 2018 : 10:30 AM to 11:15 AM
• Sponsor: Survey Research Methods Section
• Author(s): Davia Moyse* and Matt Jans and Ronaldo Iachan and Lee Harding and Scott Worthge and James Dayton and Yangyang Deng and Tracy Visconti
• Companies: ICF and ICF and ICF and ICF and MFour and ICF and ICF and MFour
• Keywords: nonprobability surveys; mobile phones; innovative data collection methods; nonprobability benchmarking; paradata

Abstract

General population surveys traditionally sample people from households because people are easy to locate where they live. However, mobile phone technology and geolocation advances have made it simple to locate people in other places. For example, most people visit a grocery or convenience store to purchase food and household items. Can sampling at these locations provide estimates comparable to traditional sampling? This paper discusses a piloted method using geofenced grocery and convenience stores in Washington State. We sampled mobile panel members when they entered geofenced areas around these stores, asking them to answer a few questions, and take a picture of an alcohol, tobacco, or sugar-sweetened beverage display. We benchmarked survey responses against population control totals and a probability survey on the same topic and population. We also used paradata to assess feasibility and respondent burden. The benefits of this innovative method over traditional phone or mail surveys are quick and inexpensive administration, and the ability to capture images as data. We discuss findings in the context of panel representativeness and respondent burden.