Statistical methods´ application in Comprehensive Sustainability Index and its application in regional sustainability measurement

In this paper, we apply efforts to explore the insights of sustainable development of states and to develop a model based on factor analysis aiming at evaluating sustainability in a qualitative and quantitative method, thus distinguishing more sustainable countries which can help elaborate sustainable development plans, and evaluate the effectiveness of such plans. We use the term Comprehensive Sustainability Index (CSI) to measure sustainability and such index is determined by a linear combination of a set of indicators. Two primary challenges of establishing such sustainability model lie in the selection of indicators and determination of weights of indicators. We select 10 representative indicators coming from three aspects: social development, economic development and environmental protection. Considering it´s difficult for us to figure out the insight relations among sustainability indicators, we use factor analysis to convert observations of correlated variables into values of linearly uncorrelated variables called factors thus determining the weights of indicators. In the evaluation section, we select 15 countries to form the sample-5 least developed countries, 5 developing countries and 5 developed countries. By calculating the value of CSI of each country, we find out that there are clear differences among the CSIs of the three kinds of countries. And based on such differences, we reach classification of CSI. Furthermore, we define sustainable development by comparing CSI (whether the CSI is greater than certain value). By comparing the contribution of each indicator to the CSI and the variation trend of the indicator, which is represented by the gradient of the regression curve, we detect disadvantageous aspects of certain countries´ efforts toward a more sustainable future. Based such analysis, we put forward our 20 Year Development Plan. We also conclude with concrete values of indicators that are required to be achieved.