Modeling the Happiness Index in Indonesia Using a Multiresponse Spline Nonparametric Regression Analysis Approach

Abstract

Regression analysis is a statistical methodology that utilizes the relationship between two quantitative variables so that the response variable can be predicted. Multiresponse spline nonparametric regression is a regression model that has more than one correlated response variable where the regression function or curve is not assumed to form a certain pattern and is approximated by a spline function. Spline is a segmented polynomial model that able to handle increasing or decreasing data patterns with knot points indicating changes in data. The aim of this research is to obtain a multiresponse nonparametric spline regression model on the happiness index with the predictor variables used being literacy rate, open unemployment rate, population density, human development index, and unmet need for health services. The best model produced in this research is a linear spline model 1 knot points with a GCV values is 6,227930 and a coefficient of determination is 78,6021.