that the demographic variables could not be said to be predictors of community
network use regardless of the presence of other variables.
The Predictors of Use in Two Subgroups Divided by Alternative Service
Accessibility
In the test above, attitude was found to be the only predictor of use among
all eight psychological and demographic variables. When the test was conducted
separately on people with an alternative ISP and on people without,
would this finding be replicated? Research question 2 relates to this question.
For this purpose, the sample was divided into two subgroups to conduct
the same hierarchical regression analysis. In the regression model,
the four demographic variables were entered first, followed by the four
psychological variables. Tables 3 and 4 present the test results of people
without and with an alternative ISP, respectively.
As table 3 and table 4 show, the results from the two separate tests
uncovered that community network use was predicted by a different set
of predictors for each group. First, table 3 shows the result for users who
did not have an alternative access. When four demographic variables were
entered into the model in step 1, they explained use statistically significantly
at the .05 level (R2 p .123, p ! .05). The addition of four psychological
variables in step 2 did not increase the variance in use statistically