Estimates of carbon dioxide (CO2) emissions are currently an important issue due to the enhanced greenhouse effect which leads to the phenomenon called global warming. This study focuses on ASEAN countries and applies regression analysis to examine interrelationship among the determinants of CO2 emissions. Stepwise multiple regression models are developed using seven indicators and the generated interaction variables are also included in the entire possible models. There are four main phases involved, namely data collection and preparation phase, model identification and estimation phase, model refinement and selection phase, and model validation phase. These four main phases are jointed and summarized as an algorithm of regression model-building. By taking the best model into account, it is found that energy consumption, GDP, population and income group have statistically significant effects on CO2 emissions. From the findings, there exists interaction effects in the best model obtained.
展开▼
