Analysis of the relationship between lifestyle and life satisfaction using transparent and nonlinear parametric models

Abstract

This study aims to analyse the relationship between lifestyle and life satisfaction in the UK using a class of transparent and nonlinear parametric models, based on which the dependency of the response variable (happiness or life satisfaction) on a set of independent (or explanatory) variables relating to life styles can easily be interpreted. A forward regression orthogonal least square (FROLS) algorithm, initially developed in control and systems engineering area, is employed to build the transparent models. Based on over 10000 surveyed samples with 42 variables from UK Understanding Society Data, a number of significant variables of lifestyle have been selected first according to the rank of their contribution to explaining life satisfaction. As an example, a reduced simple nonlinear model with only 12 model terms is presented. This model is meaningful in that it provides a parsimonious representation of the relationship between happiness and lifestyle, reveals how life satisfaction quantitatively depends on lifestyle, and how the lifestyle variables interactively affect life satisfaction.