Pemodelan Regresi Spline Truncated pada Faktor-Faktor yang Memengaruhi Angka Morbiditas di Provinsi Sumatera Utara

Abstract

The morbidity rate is the condition of the population declared sick because they cannot carry out daily activities such as working, taking care of the household, and other normal activities. The higher the morbidity rate indicates the worse the health level of the population in a region. The data used in this study is data on the morbidity rate in North Sumatra Province in 2022 along with the factors that allegedly affect it. The data is secondary data derived from the official publications of the Central Bureau of Statistics (BPS), namely "Statistics on People's Welfare of North Sumatra Province" and "North Sumatra Province in 2023 Figures". Based on the analysis conducted in this study, it was found that the relationship pattern of each independent variable and non-free variable did not follow a certain pattern. Therefore, this study used nonparametric regression analysis is truncated spline regression method. The method used aims to model and determines the factors that affect morbidity rate data in North Sumatra Province in 2022. The results of modeling with the truncated spline regression method obtained the best model with a combination of optimal knots (3,3,1,3,3,3,3) and a minimum GCV of 5.69; obtained 7 variables that significantly affect the morbidity rate, namely population density, percentage of poor people, average length of schooling, percentage of households that have access to proper sanitation, percentage of households that have adequate drinking water, percentage of population have health complaints and seek road treatment and level open unemployment with a coefficient of determination (R2) of 98.43%.