Abstract
The four most recent sub-datasets of the World Value Survey (WVS) database (i.e., WVS3, WVS4, WVS5, and WVS6) contain a total of 25 non-numerical variables of environmental protection values and cover the period from 1994 to 2014. This study utilized these datasets to obtain the spatiotemporal distributions of the values and provided a preliminary analysis of the environmental protection values in different cultural districts. The work includes 4 parts. First, the information from the values included in the 25 variables is subjectively classified into action values and attitude values according to the meanings of the variable labels. Then, quantitative clustering is used to verify the results of the first step. These two steps consistently classify the 25 variables into "action" and "attitude" families. At the third step, all variables are processed as horizontal distributions in terms of the country using the arithmetic mean of the serial numbers chosen by the respondents because these numbers reflect the grade of the behavior or attitude toward environmental protection. A clustering procedure is also included in this step to reconfirm the classification results of the previous two steps. Finally, the two families are quantified using their common factors, which are the first leading modes of the empirical orthogonal function for each family. The multiyear averaged cultural district mean "action" and "attitude" indices are analyzed according to the World Culture Map. The results show that districts with different cultures have very different environmental protection values.