In the field of social sciences, certain tasks, such as the identification of typologies and the characterization of groups of individuals according to a set of questions, tend to pose a challenge for researchers. Further complications arise if the chosen rating scale is from 0 to 10, since the responses can be treated either as metric or categorical variables. This paper shows that neither treatment is able on its own to capture all the inherent properties of this type of data, and goes on to propose a bicriteria clustering approach, which captures both perspectives and enables the simultaneous analysis of mixed data using multiple tables.