14

The second factor could be viewed as the industry structure, which

specifically refers to the scale of the service industry including educational

sectors. The development level of the educational industry is reflected in the

number of schools and student enrollment in the area.

The third factor is related to urban people’s wealth. The more money they

make, the more money they will spend on telecommunication. We noticed that

the rural residents’ income is not categorized in this factor; rather, it is more

correlated with the region’s overall development level, which was categorized as

“cosmopolitanism” in the first factor. It indicates that the rural resident’s income

and expenditure do not significantly affect Internet adoption at the present stage.

Correlation does not provide prediction of dependent variables. Since

most of predictors are highly correlated with each other, stepwise regression is

adopted to build regression model. Stepwise regression is typically used to

develop a subset of IVS that is useful in predicting the Dependent variable, and

to eliminate those Independent variables that do not provide additional prediction

to the predictors already in the equation. As one methods of statistical

regression, forward selection adds one variable at a time provided they meet the

statistical criteria for entry (Tabachnick &Fidell, 2002)

Table 7 presents the results of OLS regression models on Internet

population in 2001. Rate of home computer ownership is one of the indicators for

the factor of “level of cosmopolitism” (table 4), number of schools is one the

indicator for the factor of “industry structure”, and the income of urban residents

is one of the indicator of “Urban residents’ wealth”. The first model shows that the