Religiosity, neutrality, fairness, skepticism, and societal tranquility: A data science analysis of the World Values Survey.

Abstract

Quantitative models of social differences have not only made major contributions to the fields of cross-cultural anthropology, psychology and sociology, but also have allowed for interdisciplinary studies that bring together engineering, life sciences, and social sciences. In this study, the authors use a data science approach to discover a set of quantitative social dimensions based on the World Values Survey, a nationally representative survey covering 60 countries and 90,000 individuals. Five national social dimensions, representing 198 questions and 56 countries are discovered using multidimensional item response theory (MIRT). They are (1) Religiosity, (2) Neutrality, (3) Fairness, (4) Skepticism, and (5) Societal Tranquility. This approach is unique from previous quantitative models because it groups responses by country and analyzes binary, nominal, and ordinal survey questions. It is possible today due to recent advancements in computing power and programming. Furthermore, this methodology tests the validity of previous quantitative dimensions and finds that some of the existing social and cultural dimensions are not clearly discernable. Therefore, this model provides not only more a rigorous methodology but also new social dimensions which more accurately quantify underlying differences across countries in the World Values Survey. Like other quantitative cross-cultural models, this model is a deeply simplified representation of national social differences. However, it is a useful tool for modeling national differences and can be used to help us understand the impacts of social preferences and values on different political, economic, and development variables.