The Classification of OECD Countries in Terms of Life Satisfaction Using Partial Least Squares Discriminant Analysis

Authors : Esra POLAT

Pages : 365-376

Doi:10.16984/saufenbilder.632820

View : 22 | Download : 15

Publication Date : 2020-04-01

Article Type : Research Paper

Abstract :Life satisfaction insert ignore into journalissuearticles values(LS); measures how people assess their lives as a whole, not their present emotions. Measuring emotions can be very subjective, but it is still a useful completion to more objective data when comparing quality of life across countries. Many questionnaires are used to measure especially LS and happiness. The Partial Least Squares Discriminant Analysis insert ignore into journalissuearticles values(PLSDA); is a statistical method for classification and includes an ordinary Partial Least Squares Regression, where the dependent variable is categorical that represents each observation`s class membership. In this study, the purpose is to classify 35 OECD countries correctly to their predefined classes insert ignore into journalissuearticles values(above or below the average LS level of OECD); by using year 2017 Better Life Index data. In the analyses PLSDA, a flexible supervised classification method, is used. PLSDA is a preferable alternative method in case of some assumptions not satisfied for classical discriminant analysis. The results showed that PLSDA has a satisfying classification performance and self-reported health insert ignore into journalissuearticles values(SH); is only effective variable in determining the LS levels of countries
Keywords : Better Life Index, classification, life satisfaction, OECD countries, Partial Least Squares Discriminant Analysis