ABSTRACT
The chapter describes different solutions to the problem of extracting aspects of well-being as expressed by users on social networking sites (SNS). Four approaches are discussed: the “Hedonometer”, the “Gross National Happiness Index”, the “World Well-Being Project” and the “Twitter Subjective Well-Being” index, which is the novel indicator proposed by the authors of this book. Although all these methods rely mainly on Twitter and Facebook as data sources, the same ideas can be straightforwardly applied to other sources of textual big data.
For the Twitter Subjective Well-Being index, a cross-country comparison between Japan and Italy analysis shows that the perceived well-being is a function of structural conditions (such as the level of income of a country, ageing, average health status, etc.) but also of cultural characteristics of each country, meaning that a single global model of well-being is a bit too ambitious as a goal. A causal inference analysis based on structural equation modelling will shed light on the relative importance of macro-economic and health variables on well-being.
