ANALISIS REGRESI LOGISTIK MULTINOMIAL PADA PENGARUH DETERMINAN TINGKAT KONSUMSI KOPI

Abstract

The coffee culture by students is influenced by several factors, such as stress on doing assignments, coffee to just calm down, coffee as a consumption during organizational meetings, and so on. An overview of the level of coffee consumer behavior, especially mathematics students of UNU Blitar, can be analyzed by the multinomial logistic regression method. Multinomial logistic regression is a regression analysis with categorical response variables that can be used to find relationships between dichotomous response variables (nominal or ordinal consisting of two categories) and one or more predictor variables that are category scale or continuous (Hosmer & Lemeshow, 2000). The study was conducted with the aim of knowing significantly the variables that affect the level of coffee consumption. Based on the research conducted, the results found that there are one or more independent variables that have a significant effect on the dependent variables. Taken together, there is at least one variable of gender, age, coffee rates, the influence of coffee liking, and the frequency of visits to coffee shops that affect the level of coffee consumption. The influence of variables partially. The variables of the influence of coffee likes from friends, the influence of coffee likes from relatives and the frequency of visits to cafes rarely have a significant effect on the consumption level of the low category. Meanwhile, the moderate level of coffee consumption category is significantly influenced by the variable frequency of visits to cafes (rarely).