Geoscience Reference
In-Depth Information
LP has a Help button (
Figure 8
), not identified by users, thus the website
was the least preferred and trusted for this attribute.
Study 2 revealed that non-experts do not trust the colours they like the
most, which contradicts the findings of previous studies, where it is sug-
gested that beautiful interfaces are perceived as trustworthy (Karvonen
2000). This finding shows the distinctive characteristics of online map use.
Easily distinguished colours give confidence in map interpretation (e.g.
Red/Blue/Green map). In Study 3, users liked and trusted the colours pro-
vided by LP the most, although the same green shades were used for the
Green map Study 2, which was the least trusted. It should not be ignored
that this was a comparative assessment of the trustee attributes and the
colours used in the other two applications were highly criticised by the
participants. Several participants suggested using distinctive colours
instead of different shades of the same colour, corroborating the second
study's findings.
6- Conclusions and future plan
All three studies showed that the trustee attributes investigated are impor-
tant in the formation of non-experts trust perceptions; furthermore, for
some trustee attributes (i.e. colours and legend), designing for trust cannot
be achieved in parallel to designing likeable interfaces. Also, all three
studies reveal the importance of trust in this context. Although it can be as
simple as designing bigger maps, focusing on a trust-oriented interface
design to improve the trustworthiness of Web GIS is of particular importance
when designing for non-expert users.
It should not be ignored that since Web GIS applications are used within
wider contexts there are several additional trustee attributes that should be
considered, as described by the trust guidelines in Skarlatidou et al.
(2010a). In addition, this paper used the comparative assessment of the
trustee attributes to help users realise their importance but also to under-
stand how different representations of these attributes influence trust (e.g.
what map size or colours are more trustworthy?). A more reliable approach
would involve a controlled experiment where each trustee attribute would
be tested separately, with the development of different interfaces for the
same application.
