This paper helps to clarify a recurring difficulty in qualitative methodology, namely the comparability of the corpus on which the analysis is based. This paper addresses this question by examining a case arising from the use of focus groups. With the aim of studying the attitudes of citizens toward European integration, we conducted a comparison between France, French-speaking Belgium, and the United Kingdom. We also specifically privileged artificially constituted groups segmented along a typology and consisting of socioeconomic groups reflecting a simplified model of social stratification. Because the intervention of the researcher is particularly salient in this type of interview, critical issues are raised for the comparison. Yet these are rarely explained when a qualitative approach is adopted. Among these issues are how far we should go in the construction of comparability, that is, in the harmonization of data, and on what basis, and what place should be given to the uniqueness of the social within each field. In the light of our research, we show that typicality is a variable-geometry structure that cannot be regarded as absolute, that is, in our case, uniformly applicable across social categories or countries. However, these obstacles are not insurmountable as long as we adopt a sufficiently reflective stance toward our practices. Overall, we aim to highlight the flexibility needed in the design of the comparative methodology implied in such a qualitative approach.