Geoscience Reference
In-Depth Information
When finishing all these tasks, participants switched to another prototype, and
the same procedure was repeated for the next sub-route. In total, the test for each
participant was completed within 1 h.
11.3 Results and Discussion
The field experiment was completed on Nov. 2010. All the participants successfully
finished the tasks. The results of the experiment include two parts: wayfinding
performance and user experience, and spatial knowledge acquisition. Results
assessing wayfinding performance and user experience can be found in Rehrl
et al. ( 2011 ). In this paper, we report the results of spatial knowledge acquisition.
The results on the aspect of spatial knowledge acquisition were analysed by
focusing on sense of direction (the pointing task), sketch maps (topological aspects:
sketched landmarks; errors in sketching turns, i.e., missing/wrong/unnecessary
turns), and sense of distance (marking half of the route). We only considered the
results from participants who were unfamiliar with the sub-routes. In total, we got
24 participant/sub-route pairs (8 for mobile maps, 8 for AR, and 8 for voice). 1 The
male-female ratios were similar in the three groups. In the following, we present
and discuss the results.
11.3.1 Sense of Direction (The Pointing Task)
The sense of direction was measured as the deviation between actual directions and
pointed directions in the pointing task. The deviations were measured in degrees.
Figure 11.3 shows the results of sense of direction, comparing mobile maps, AR,
and voice.
The results show that map users and voice users performed considerably better in
pointing to the start compared to AR users. The results for AR users were not
surprising, because the AR-based prototype suffered from the poor GPS signal (in
both the map-based and voice-based prototypes, some route matching algorithms
can be used to improve the GPS accuracy) and poor compass accuracy, and thus
brought some confusion to the users. However, map users did not perform consid-
erably better than voice users, which is inconsistent with our expectation.
A possible explanation is that map users did not make full use of the map-based
prototype, e.g., according to the usage log, they seldom used the zooming function
to get an “overview map”. We also did a one-way ANOVA (analysis of variance)
1 The spatial knowledge acquisition test was conducted within a framework including many other
tests, in which familiarity was not the only criterion in choosing participants. Therefore, only
24 participant/sub-route pairs were “unfamiliar”.
Search WWH ::




Custom Search