Assessing Social Interest in Burnout Using Google Trends Data

Abstract

Burnout is a serious problem in modern society and early detection methods are needed to successfully handled its multiple effects. The latter refer to working well-being, as well as to the affective, psychological, physiological, and behavioral well-being of workers. However, in many countries official statistics on this topic are not available. For this reason, we propose to use Google Trends data as proxies for the interest in burnout and to analyze them through the functional data analysis approach. The latter allows to address the so-called ‘curse of dimensionality’ of big data, enabling an effective statistical analysis when the number of variables exceeds the number of observations. Under this framework, the functional analysis of variance (FANOVA) model is used for testing a macro geographic area effect on search queries for the keyword “burnout” in Italy. The estimation of the FANOVA model is proposed in a finite dimensional space generated by a basis function representation. Thus, the functional model is reduced to a MANOVA model on the basis coefficients.