Penerapan Cochrane-Orcutt Iterative Procedure untuk Mengatasi Pelanggaran Asumsi Non Autokorelasi pada Analisis Regresi Linier Berganda Menggunakan Software R

Abstract

Multiple linear regression analysis is often used to determine the relationship between one dependent variable with two or more independent variables. Both the dependent variable and the independent variable are numerical. If the classical assumptions on multiple linear regression models are met then the parameter estimator will be Best Linear Unbiased Estimation (BLUE). These assumptions are the normality of error, non autocorrelation, non-multicolinearity and homoskedasticity. In real data is often encountered violation of assumptions, one of which assumption non autokorelasi not fulfilled. In this research discussed about the handling of autocorrelation on multiple linear regression model using Cochrane-Orcutt Iterative Procedure and software R. Data used is data of Regency/City in East Java 2015 sourced Statistics Indonesia.