Advances in Data Science

4On the “Complexity” of Social Reality. Some Reflections About the Use of Symbolic Data Analysis in Social Sciences

4.1. Introduction

The theme of the “complexity” of the social world has long been present in methodological and epistemological reflection on the specificities of social sciences (e.g. [SIM 22]). Passage of textbooks or theme of a more or less vague essayist discourse that, on the contrary, we can consider it as an important issue if we wish, as is the order of the day, to enhance the dialogue between natural sciences and social sciences.

This theme brings into play the very notion of modeling and its limits when it is transferred more or less mechanically from physics, chemistry, or biology to the “social” world, in the name of the proven effectiveness of the mathematization of these disciplines.

In this text, we begin by discussing various strategies developed in social sciences research, particularly since the advent of the “statistics for researchers” [ROU 98], to rigorously treat this “complexity”, which can be defined in various ways (section 4.2). This allows us to address, section 4.3, some of the characteristics of Symbolic Data Analysis (SDA), which make it a promising area of research from the perspective of the study of complex phenomena in social sciences. An exploratory case study, based on the exploitation of European EU-SILC 2013 cross-sectional data, is presented and discussed for this purpose. It leads us to conclude with a research program ...