4.2.2 Multiple Regression Analysis
Sometimes two or more variables have a major impact on the dependent variable. In this situation, the dependent variable multiple regression is used. The linearity assumption in multiple regressions of relationships between variables is also confirmed and thus the regression equation is defined as follows:
y = a+b1x1+b2x2+b3x3 +... [27]. (1)
The results and output tables obtained from the multiple regression analysis test by SPSS software are presented in the table 5.
The significance level (sig) for the components of pace of providing the products and services and satisfaction is more than 0.05, which indicates no linear relationship exists between the above factors and competitive advantage in e-commerce. On the other hand, the significance level (sig) of components of knowledge from customers, knowledge for customer, knowledge about customer, recording and spreading customers' knowledge, quality of the products and services and reasonable prices is less than 0.05, which indicates a linear relationship exists between the above factors and competitive advantage in e-commerce. The determining coefficient is 72.2% between the above components and the competitive advantage in e-commerce and this indicates that 72.2% of the change in component of competitive advantage in e-commerce is affected by the components of the knowledge from customers, knowledge for customer, knowledge about customer, recording and spreading customers' knowledge, quality of the products and services and reasonable prices. Multiple regression line equation and the obtained results are summarized, as follows:
Table 5: Results of multiple regression analysis of 8 factors and competitive advantage in e-commerce
Competitive Advantage in E-commerce = 0.933 + 0.076 Knowledge from customers + 0.234 Knowledge for customer + 0.222 Knowledge about customer + 0.116 Recording and spreading customers' knowledge + 0.107 Quality of the products and services + 0.057 Reasonable prices