Abstract
In this study, we examine and compare the analytical methods of data that are obtained by Internet survey using multi-group question items about practical businesses. We aimed to clarify the method that can derive a suggestion of concrete measures helpful to practical businesses, by applying the following four analytical methods – (1) multiple regression analysis; (2) selection-type multi-group principal component regression analysis; (3) hierarchical principal component regression analysis; and (4) principal component SEM – to actual examples, integrating the characteristics of these analytical methods, and avoiding the multicollinearity problem. As an actual example, we selected an Internet survey targeting the users of a company of third party payment service provider in the Internet trading. As a result of the examination and comparison of analytical methods based on the actual example, it was clarified, from the practical viewpoint of practical businesses, it is possible to derive a proposal based on the analysis result, that the (2) selection-type multi-group principal component regression analysis or (3) hierarchical principal component regression analysis was useful.
