Linear regression model as an approach to analyze spatio temporal data to know the influence factors of the number of crimes in North Sumatera

Abstract

Crimes in Indonesia tend to increase every year, both in variation and in number. Crimes in this study include murder, torture, rape, theft, fraud, and other cases that make people feel insecure. This study aims to know what factors affect the number of criminal acts in the Province of North Sumatera and determine the influence of time and location in 33 districts/ cities in North Sumatra. This study uses secondary data collected by POLRI, BPS, and Dinsos in North Sumatera Province in 2013, 2014, 2015 and 2016. The independent variables used are the number of unemployed people (X1), the number of poor people (X2), population density (X3), human development index (X4), the number of places of worship (X5), the economic growth rate (X6) and the number of families with a social psychological problem (X7). The linear regression model is used as an approach to analyze. The results of this study show that the independent variables significantly influence the number of crimes acts in North Sumatera Province. The independent variables are unemployment (X1), the number of poor (X2), population density (X3), and human development index (X4). There is a significant location influence on the occurrence of the number of criminal acts in 33 districts/cities of North Sumatera Province. Based on the descriptive statistics, Sibolga, Tebing Tinggi, Padangsidempuan, Tanjungbalai, Samosir, and Gunungsitoli are the six districts with the highest proportions in criminal actions, where those districts are all cities.