Biomedical Engineering Reference
In-Depth Information
of scope, the review focuses on the effect of different components of body-based
information (rotation vs. translation), rather than different cues, because during active
walking users are provided with a full set of body-based cues. The criteria for inclu-
sion in the review were that a study: (a) involved users changing both position and
orientation as they navigated, and (b) investigated different components (rotational;
translational) of body-based information, not just different cues (proprioception;
vestibular; efference copy). Low-level studies that focused exclusively on rotational
movement (e.g., [
21
]) or distance perception are omitted (e.g., [
22
]).
5.4.1 Review Framework
The studies that are reviewed are divided into four groups (see Table
5.2
), which are
dictated by the type of navigation participants performed while acquiring knowl-
edge of the environment (single-route vs. whole-environment) and the scale of the
environment (small vs. large). Single-route acquisition is where participants only
navigated one specific route. Whole-environment acquisition is where participants
either freely explored the environment or navigated to find target locations that were
distributed around the environment, in specific but changing orders. The distinction
between small- versus large-scale environments is explained above.
Spatial
extent
is classified as either small (room-sized; a maximum of approxi-
mately 10
×
10m) or large (building-sized or greater). The richness of the
visual
scene
is classified as low, medium or high. Low corresponds to environments where,
apart from target landmarks, variations in the visual scene were just designed to
provide optic flow. High corresponds to rich visual scenes that contained a surfeit of
visual detail of deliberately added landmarks (e.g., at each junction in a building),
and medium corresponds to scenes that did not belong clearly to either of the other
categories.
The experimental results summarized in Table
5.2
are divided into
navigation
performance (time taken and distance traveled metrics that show how efficiently
participants moved between places) and survey knowledge (
direction
estimates and
straight line
distance
estimates). These survey metrics are the basic information
people need if they are to know the location of places in relation to each other, or
take shortcuts [
30
]. Absolute direction estimate errors were used, rather than signed
errors that indicate response biases (e.g., see [
31
]) and, in all except the triangle
completion studies (single-route acquisition/small-scale environments), the distance
estimates were estimates of relative straight line distance, which are accurate if people
have a well-developed cognitive map [
9
]. For a discussion of distance estimation
methodologies, see [
32
].
Each type of results is subdivided into four columns:
Vis
,
Rot
,
Tran
, and
Full
.
Vis is where participants were only provided with visual information (e.g., a desk-
top VE). Rot and Tran are where participants were provided with the rotational
and translational component of body-based information, respectively, in addition to
visual information. Full is where participants were physically walking through the
