As indicated in Table 11-2, all four hypotheses H1 to H4 are supported. For example, for H1, the data well supported that user satisfaction with a Web portal will be affected by gender and the type of portal used.
The two-way ANOVA also tested the controlling effect of either factor in each pair. For example, in H1, H1b suggests that after controlling for the factor of gender, users of horizontal and vertical Web portals have different satisfaction levels. H1a is not supported; after controlling for the factor of portal type used, we cannot tell that male and female users are satisfied to different extent. In the case of H2, portal type is still a significant factor after controlling for age (H2b), while age is not significant after controlling for portal type (H2a). In H3, both of the independent variables work experience and portal type are significant after controlling for the other factor (H3a and H3b). In H4, the two independent variables are also significant after controlling for the other factor (H4a and H4b).
Once it has been determined that differences of means exist in the groups, the least squares means are calculated and T-tests are run for comparisons of means between the groups. The comparison matrices of means in different groups are listed in Table 11-3a, 11-3b, 11-4a, 11-4b, 11-5 and 11-6.