The Influence of Information Search on Preference Formation and Choice (INSPiRE)

Abstract

In this stated choice experiment respondents choose among bottles of wine described using seven attributes: 1) Country of origin, 2) colour of the wine, 3) alcohol by volume, 4) grape variety, 5) characteristic of the wine, 6) whether the wine was organic, and 7) price. The attribute levels were established after scrapping the websites of three large supermarket chains in the UK. Using the web-scraped wine data, we calculated a set of probability weights to establish the likelihood of each experimentally designed wine bottle being available in the supermarket. We used these weights to sample individual random profiles from the restricted factorial each time an individual entered the survey. The survey was programmed in Shiny, which is an R package, and the data was gathered during January 2020. In total, 4,121 respondents were randomly allocated to one of 10 treatments, each designed to test a specific aspect of search and preference learning. Example tasks and the instruction videos can be found at https://choice-tasks.inspire-project.info/. Respondents were randomly recruited from the UK population aged 18 and over. We did not use any quota sampling to ensure that our samples were representative of the target population.