A New Measurement Method on Correlation Coefficient for Attribute Fuzzy Interval Data and Its Applications

Abstract

In general, the classical Pearson`s correlation coefficient shall be applied to evaluate the relationship for two continue variable and the Spearmen correlation coefficient could be employed to measure the casual for two attribute variable. However, as the data are composed of qualitative attributed fuzzy interval values, it is hard to adopt above the traditional scheme calculation the attribute correlation coefficient. In this study, we present the simple calculation of attribute fuzzy interval correlation coefficient via the central and length of the attribute fuzzy interval observation and the attribute correlation coefficient of two variables. As such, this investigate is able to offer an improving measure of attribute fuzzy interval correlation from attribute fuzzy interval observation. Furthermore, we apply simulation study and empirical study to proof the validity of our proposal on attribute fuzzy interval correlation coefficient using data via the adult population survey (APS) of the global entrepreneurship monitor (GEM) in Taiwan.