Dataset of choice experiment surveys assessing consumers' and farmers' preferences for five underutilised crops in different European countries

This submission derives from Task 5.3, "Farmers’ and Consumers’ Preferences and Behavioural Economics Underpinning Underutilised Crops" (T5.3), of the Horizon 2020 project "Realising Dynamic Value Chains for Underutilised Crops – RADIANT" (2021–2025). It consists of six datasets derived from discrete choice experiment (DCE) surveys conducted online with farmers and consumers to investigate their preferences for underutilised crops and derived food products, respectively. We use an ‘XXYYZZ’ sequence to name the files as follows:(1) 'XX' = two letters indicating whether the respondents are farmers (FA) or consumers (CO);(2) 'YY' = two letters indicating the European country where the data were collected (BG = Bulgaria, ES = Spain, IT = Italy, SC = Scotland, United Kingdom);(3) 'ZZ' = two letters indicating the underutilised crop or food product considered (BB = Bere barley, HT = hanging tomatoes, ID = IDEAL tomato, PR = Parmigiano Reggiano cheese, YP = yellow pea [and einkorn wheat]). e.g., 'COITPR' = Consumers data from Italy on the Parmigiano Reggiano product Each dataset consists of one sheet called "survey" (one row per respondent), which reports socio-demographic characteristics and attitudes of the respondent, as well as other information derived from the complementary survey; one sheet called "choices" (18 rows per respondent), which reports their choices in the DCE; and one sheet called "metadata", which describes all the variables (one table per data sheet). The farmers' dataset also includes a sheet called "PGG" (one row per respondent), which includes data on a public good game administered to a randomly selected subset of farmers as part of the survey. The total number of consumer respondents is 299 for Bere barley, 345 for hanging tomatoes, 336 for IDEAL tomato, 346 for yellow pea (and einkorn wheat), and 340 for Parmigiano Reggiano cheese. The number of farmer respondents is 331, of whom 186 participated in the PGG (220 and 108 when considering only high-quality responses based on restrictive checks).