Online, face-to-face and telephone surveys : comparing different sampling methods in wine consumer research

Today, depending on topic, goal and budget, all kinds of sampling methods are being used, in order to collect consumer data for research in the wine business. However, it is questionable which survey method is able to generate data that does represent the entire population. A representative face-to-face survey with 2000 respondents and a telephone survey with 1000 respondents were compared with two online surveys, one based on quota sampling (2000) and the other on snowball sampling (3000) using identical questions. Due to the sampling method, three of the surveys were representative of the socio-demographic structure of the German population in terms of six demographic variables that were selected for the quota sampling. The online survey (based on the snowball sample) had large biases concerning representativeness. Regarding the behavioural characteristics of consumers, the face-to-face data delivered the best results, followed by the telephone interviews and finally the online quota survey. Face-to-face surveys still deliver the most representative results. Telephone surveys may provide a good alternative, but we would advise use of a larger sample. The online quota survey needs to be corrected, while in the case of snowball sampling, one should relinquish representativeness.