Geography Reference
In-Depth Information
spatial representations of individuals even though the perceptive input differs
significantly.
The present study applied a recently introduced method for analysing thinking
aloud protocols using natural language processing (Sarjakoski et al.
2013
). Its
results reflect the distribution of selected landmark terms in the collected
thinking-aloud protocols. The thinking-aloud method has been extensively used
in probing human thinking processes (Ericsson and Simon
1980
), but doubts have
been presented on its validity in concurrent use (e.g., Smagorinsky
1998
). Thinking
aloud concurrently easily affects task performance if a participant strays to free-
flow speech and forgets the task at hand. However, properly instructed, trained and
undisrupted, thinking aloud has been shown to provide reliable data on the thoughts
of test participants (Ericsson and Simon
1998
; Boren and Ramey
2000
). We took
these prerequisites into account when directing the thinking-aloud tasks and con-
sider the collected data to be reliable with regards to people
s perception of
landmarks along the route. Another shortcoming of the concurrent thinking-aloud
method is the probable incompleteness of the data: participants may be unable to
verbalise all of their thoughts and actions related to the task, particularly in
non-verbal practices, such as spatial thinking (e.g., Whitaker and Cuqlock-Knopp
1992
). In the present study, we instructed participants to memorise the route and to
think-aloud about their perception of it, which directed the participants
'
concentra-
tion to the surrounding features along the route and made them verbalise at least
those features that they saw as important for following the route. Consequently, the
results do highlight those landmarks that are potentially effective in route direc-
tions, but they should be regarded as an explorative rather than a comprehensive set
of all the prominent landmarks.
Technical wayfinding support by geospatial applications is most effective when
given in real time, landmark by landmark, during the navigation (Rehrl et al.
2010
).
Therefore, the landmarks in wayfinding maps and navigators should be selected
based on the real perception in the environment. The found loss of perceptive
differences between conditions in the recall phase highlights the need for
perception-based directions in real-time navigation guidance—human conception
and memory do not always inherently focus on the most perceptible landmarks. The
results from studies like ours in the present paper provide sets of empirically
verified, perceptionally prominent landmarks to be employed in geospatial naviga-
tion applications.
'
Conclusions and Future Work
The present study addressed the open question of the perception and recall of
landmarks in nature under limited light conditions at night. We approached
the issue by means of a thinking-aloud study on a nature trail with participant
groups in the day and night conditions and used previously developed natural
language processing and sketch map methods for the analysis of the
(continued)
