New tools for the construction, analysis and interpretation of social indicators based on ordinal variables

Dealing with ordinal variables in social measurement raises many methodological issues, particularly when consistent and meaningful indicators are to be defined about subjective data. Quite often, ordinal data are transformed into cardinal variables, through scaling procedures, so as to apply quantitative multivariate tools. These procedures are sometimes reasonable, but are often quite questionable, since they alter the intrinsic nature of the data. In this paper, we introduce new tools for addressing the construction of social indicators, based on ordinal variables. The approach draws upon partial order theory, a branch of discrete mathematics providing concepts and tools that fit very naturally the needs of ordinal data analysis. Partial order tools allows for the extraction of information directly out of the relational structure of the data and provide robust results, not requiring binding assumptions. In order to exploit the potentialities of partial order theory in social measurement, we present an exemplificative application pertaining European data (European Social Survey). We show how indicators construction can be dealt with through this alternative approach, getting new insights into the data and allowing for better understanding of social phenomena and better communication to policy-makers.