ANALISIS KEPUTUSAN KONSUMEN MEMILIH BAHAN BAKAR MINYAK (BBM) MENGGUNAKAN MODEL REGRESI LOGISTIK BINER DAN MODEL LOG LINIER (Studi Kasus SPBU 44.502.10 Ketileng Semarang)

Abstract

Fuel oil is a fuel derived and/or processed from petroleum. Fuel is often used for motor vehicles among others premium and pertamax. Some recent times has happened several times increase and decrease in fuel prices, even at the beginning of 2015 has happened  a new policy on the elimination of fuel subsidies. It affects on fuel consumption, especially consumption of premuim and pertamax. Many factors influence the consumer's decision in choosing a fuel, therefore needs to be analyzed to find out factors influencing consumer decision in choosing a fuel. This study was conducted to determine the factors that influence consumer decisions in choosing a fuel with a binary logistic regression model and the factors that influence the relationship with log linear models. Binary logistic regression is a method of data analysis used to find the relationship between the response variable (Y) that is binary or dichotomous with some predictor variables (X). Log linear models were used to analyze the relationship between categorical variables. Of a binary logistic regression model obtained influential variable is employment, vehicle age and income variable, with the biggest opportunity is 0,78862, is premium consumers with private employment, the age of the vehicle mote than 5 years and the income less than 1.500.000. for log linear models got the biggest opportunity is 0,91259, is premium consumers to the work of civil servant, the age of the vehicle mote than 5 years and the income less than 1.500.000. 

Keywords : fuel, binary logistic regression model, log linear models