Relationship between job and smoking habits based on the CART method

Abstract

Smoking is still a large public health problem, especially for people in developing countries. Tobacco Atlas, a map published by the World Health Organization (WHO) to monitor the spread of tobacco, shows Indonesia at number 4 in the average number of cigarettes consumed per person. A study conducted by WHO claims that Indonesia has the third largest smoker population in the world after China and India. The growth of cigarette consumption has had an impact on the spread of diseases caused by smoking and increases in death related to tobacco-consumption. Tobacco-related deaths are predicted to increase to 10 million people, with 70% of these in middle-income countries. Data from Riset Kesehatan Dasar (Riskesdas) Kementerian Kesehatan RI in 2013 shows that 44.5% of fishermen, farmers, and factory laborer are active daily smokers. Based on those findings, this study seeks to determine the relationship between employment field, employment status, and smoking habit using data from the National Socio-Economic Survey 2016. This paper use the Classification and Regression Tree (CART) method, a statistical method that can predict the value of dependent variables based on the values of several independent variables. CART consists of two methods, the classification tree (used when the variable is categorical) and the regression tree (used when the variable is continuous). This results of this CART analysis suggest there is a significant relationship between employment and smoking habits.