Nonparametric truncated spline regression model on the data of human development index (HDI) in indonesia

Abstract

The standard measurement for country's human development is Human Development Index (HDI). Several factors allegedly affect it, such as life expectancy, gross domestic regional product (GDRP), the number of poor people, and the percentage of an illeterate people. If the HDI's data and those factors is plotted, then the plot do not form a specific pattern, so the HDI's data can be applied with nonparametric truncated spline regression model. The best of nonparametric truncated spline regression model is influenced by the order and knot points selection. In this article, nonparametric truncated spline regression model on the data of HDI in Indonesia is applied at first order with 3,4 and 5 knot points. Based on the result, the best nonparametric truncated spline regression model on the data of HDI in Indonesia is obtained by the combination of 5-5-5-4 knot points. There were 2 of 4 allegedly factors significant to HDI in Indonesia based on parameter significance test followed in this research.