23
Although future research needs to verify the above explanation, the findings provide
important implications for both government and commercial organizations that provide health
information on the Internet. For government sources to be effective, the findings of this study
suggest that presenting health information in a highly responsive manner is a key. Considering
omnipresent health information by commercial sources on the Internet, which may have adverse
effects on individuals’ health when the information is inaccurate or biased, a government source
could counterbalance the effects by providing health information in a highly responsive Web
site. For commercial sources, unbiased information without commercial intent, which does not
promote certain products or treatment, is likely to appeal better to individuals who seek health
information on the Internet.
LIMITATION
There are a few limitations in the study. The primary limitation is the failure of the
source manipulation. The source manipulation did not work for at least two reasons. It is
possible that the commercial and the government sources do differ in participants’ perception of
credibility, but our manipulation was not strong enough to induce the perception. It is also
possible that participants’ perception of credibility of the two sources does not differ in an online
environment and we failed to identify a dimension on which the two sources differ. Future
research can address these issues by measuring perception of credibility of the two sources prior
to the manipulation and/or identifying a more relevant dimension of source attributes.
The second limitation is two conditions of responsiveness not only differed in
responsiveness, but also in choice. Under the high responsive condition, participants were given
three questions to choose from to view the next information. Under the low responsive
condition, however, participants were not given choices. While we tried to manipulate the only
responsiveness dimension of interactivity, we also reasoned that manipulation of both
responsiveness and choice dimensions would create a condition where confounding effects may
occur. For instance, when participants were given choices, but the Web site was not responsive
Convention