Penggunaan Metode Backward untuk Menentukan Persamaan Regresi Linier Berganda (Studi Kasus: Jumlah Penyalahgunaan Narkoba di POLRESTA Medan)

Abstract

ABSTRACT Stands for Narcotics Drugs, Psychotropic Substances, Materials addictive, which is the name of a class of natural substances, synthetic or semi synthetic. Factors that are considered influential to the increasing abuse of drugs such as negative parent, make friends with peers, as well as increased economic pressures of the evidence. Formulation of the problem in a study titled "Backward Method for Determining the Use of Multiple Linear Regression Equations (Khasus Study: Number of Drug Abuse in Medan Police)" This is to determine the appropriate equation to estimate the amount of drug abuse in the resort city of Medan police. To obtain the regression equation using the method the author is Backward backward method which is a step backwards, in which all the variables Xi regressed with the dependent variable Y. pengeleminasian variable Xi Fpar based on the value of each variable Xi is a variable that has the smallest value Fpar tangkah principal and co-absence of these variables in the model is based on Ftab. Estimators obtained is: y = 1317 + X3 + X4 0293. With the Y represents the number of drug abuse, X3 is the economic pressure, and X4 is barangbukti and the percentage of variation (correlation coefficient of determination) is described by the probe is at 55.1% so that we can conclude that the models are quite good estimator is used as an estimate of the number of Police crime in Medan resort.