19
= 22.64, p < .0001; and prior allergy preventive behavior F (1, 79) = 4.62, p < .05. Neither main
effects of the source and the responsiveness, nor their interaction term were significant: source, F
(1, 79) = .92, p > .34; responsiveness, F (1, 79) = .01, p > .92; and interaction, F (1, 79) = 2.58, p
> .11. Therefore, Hypothesis 1a was not supported.
Effects of Source on Health Behavior Intentions (RQ1b)
Hypothesis 1b predicted that individuals would be more likely to engage in health
promoting behaviors when they receive health information from a government source than a
commercial source. This hypothesis was tested through two ANCOVA models each for
preventive behavior intention and information seeking intention. In the ANCOVA model for
preventive behavior, covariates were prior allergy preventive behavior and age. In the
ANCOVA model for information seeking intention, covariates were involvement with health
information on the Web and salience of the issue allergies. The overall model for preventive
behavior intention was significant: F (5, 123) = 4.45, p < .001, R
2
= .153. While the main effects
of the source and the responsiveness were not significant, the source x responsiveness interaction
term was significant: the source, F (1, 123) = .17, p > .68; the responsiveness, F (1, 123) = .10, p
> .75; and the interaction term, F (1, 123) = .7.69, p < .01. The overall model for information
seeking intention was also significant: F (5, 125) = 14.91, p < .0001, R
2
= .374. Similar to
preventive behavior intention, the main effects of the source and the responsiveness were not
significant, while the source x responsiveness interaction term was significant: source, F (1, 125)
= 1.41, p > .24; responsiveness, F (1, 125) = .29, p > .59; and the interaction term, F (1, 125) =
4.92, p < .05. Since no main effects of the source were found, Hypothesis 1b was not supported.
Effects of Source on Attitude toward the Web Site (H1c)
Hypothesis 1c predicted that individuals would be more likely to have favorable attitude
toward the message channel, i.e., the Web site, when they receive health information from a
government source than a commercial source. This hypothesis was tested through an ANCOVA
model, in which involvement with health information on the Web was a covariate. The overall
model was significant: F (4, 126) = 9.18, p < .0001, R
2
= .226. While the main effects of the
Convention