Implementing ordinal regression model for analyzing happiness level in Indonesia

Abstract

Happiness can be an indicator of social progress achievement in developing a country. Understanding factors affecting the level of happiness in a country become important in the study of subjective well-being. This study aims to determine factors that influence happiness in Indonesia using the fifth wave of Indonesian Family Life Survey (IFLS) data. From the national data of 29,336 participants in 13 provinces were reported that 13.4% of the participants responded they were very happy, 78.6% were happy, 7% were unhappy, and 1% were very unhappy. Ordinal regression models in particular proportional odds models were implemented on analyzing factors of the four levels of happiness. The significant variables associated with happiness were age, marital status, job status, education, satisfaction, economic level, health, and religious variables. Sex, hopeful, personality, and social relation were not significantly associated with people's happiness. Happy people have higher probabilities of happiness than others. Very happy appeared to have lower probabilities of happiness as their ages increasing, while in contrast to unhappy and very unhappy people.