The aim of this study is to identify the logic behind a range of statistical methods used to reveal the structure of social representations. Subjects (N = 317) were asked to answer the following question: “For each category of European person, please indicate which other European he would most like to have contact with”. The results of the similarity analysis lead us to the conclusion that there is an ethnocentric bias, and reveal the central factor of the representation. The representation obtained by factorial correspondence analysis seems closer to current reality and enables us to understand the divisions that have structured Europe and remained embedded in the subjects. Thus, the choice of analytical method is not merely anecdotal, given that representations obtained from the same data can vary immensely.