| Summary: | 碩士 === 國立臺灣大學 === 國際企業學研究所 === 88 === Because of the development of e-commerce, many companies have expanded their businesses in the Internet world. On the other hand, it’s also much easier for consumers to gather information of products with Internet. So what is this thesis going to research is the consumer behavior on the Internet. The way to analyze is simple linear regression, and the subjects to be discussed are moderator analysis and interaction analysis.
In order to run regression model, there are three types of variables. First type of variable is the characteristics of Internet users, the second type of variable is the relative importance of products between online shopping and traditional front store shopping, the last type of variable is moderator including transaction security and the influence of peer group.
After running regression model and analysis, we summarize the following two conclusions:
1. Moderator analysis:
There is significant influence of transaction security. And peer groups will influence a person whether to shop on Internet or not.
2. Interaction analysis:
Most variables have interaction and the power of influence in the regression model are also very significant. By the way, if we let the interaction variables and other variables to run step-wise regression, almost all the interaction variables can replace at least one single variable. It shows that interaction could be a very important factor in the regression model.
|
|---|
