Penerapan Statistika Nonparametrik dengan Metode Brown-Mood pada Regresi Linier Berganda

Abstract

Brown-Mood is a method first developed by GW brown 1950 and AM mood in 1951 with the purpose of the parameters of the multiple linear regression model of the linear regression model of the equation of the median small sample size. This study discusse the application of the method of brown-mood on multiple linear regression with the open unemployment rate (X₁), and growth rate of gross regional domestic product at constant prices (X₂) to the number of poor population (Y) Province of east Kalimantan. If the method ordinary least square in a multiple linear regression is a statistical parametric aims to minimize the average (mean) error, the brown-mood methods as a nonparametric statistical method chose a multiple linear regression model by minimising the median and average weighted. The results of this research to get a linear regression model using the method of brown-mood is Ŷ=-31.11+1.74 X₁ + 1.44 X₂ from the multiple linear regression model obtained are percentage distribution of gross regional domestic product at current prices [without oil, gas and its products] and growth rate of gross regional domestic product at constant prices affect to the number of poor population.