Methods used to conduct the pan-European Union survey on consumer attitudes to physical activity, body weight and health

The purpose of conducting this survey was to identify data on consumer attitudes towards and beliefs about physical activity, body weight and health among the 15 countries of the EU. A cross-sectional study to get a picture of the attitudes to physical activity, body weight and health in the EU. For this, it was considered important that samples be nationally representative so that inferences drawn from the data could be applied to the population in each country as well as to the EU population as a whole. Using a non-probability sampling method employing quota controls (and the national weight) we obtained large sample sizes from each country which were nationally representative in terms of the variables age, sex and regional distribution. To ensure samples were truly nationally representative a national weight was used when analysing the data using the same characteristics as those used to define quotas. When examining pooled estimates for the total EU sample a population weight was applied. In total, 15,239 subjects aged 15 years and upwards in the EU completed the survey. This article gives details on the methods used in carrying out the survey from design of the questionnaire to sample selection, questionnaire administration and analysis of the data. The methods and their limitations are discussed.