Distance Estimation Experiment 3
Distance Estimation Experiment #3
Egocentric distance estimations in the real world are sensitive to
the scale of the environment. In general, small distances in an indoor
setting tend to be overestimated, and large distances in a natural outdoor
setting tend to be underestimated (Teghtsoonian & Teghtsoonian, 1970). We
were curious whether a manipulation of the apparent scale of the
environment would affect distance estimations in a VE. Participants were
assigned to one of two conditions: a "small-scale" condition in which
participants estimated distances between household objects, and a
"large-scale" condition in which they estimated distances between
airplanes. Except for the identity of the objects, the appearance of the
VE’s between conditions was identical.
Method
Participants 17 undergraduate students (7 women) were
recruited from the University of Washington department of Psychology human
subject pool. All participants were given extra credit toward their
Psychology class for their participation .
Stimuli and apparatus With few exceptions, the stimuli and
apparatus used were the same as the desktop conditions of
the distance estimation experiments reported above.
Instead of a grey cubic room,
however, the VE was an open cube (with three faces removed) composed of a
transparent blue material. The sky was textured and blue and the floor
used a green, grassy texture (see figure DE3-1). Depending on the
participant’s experimental condition, distance were judged between either
1.) a violin and keyboard or 2.) two airplanes. All participants used the
desktop system described in experiment two above.
Procedure Participants were trained on the rudiments of VE
navigation with a mouse. They were then required to complete a ‘virtual
obstacle course’ in less than four minutes before continuing
participation. Participants were then randomly assigned to either the
large-scale (airplanes) or small-scale (musical instruments) condition.
As in previous experiments, each trial began with the two objects being
randomly placed inside the environment. Participants made as many distance
estimations as they could in the remaining time, receiving feedback after
each trial.
Results
Participants completed an average of 96.29 trials (SD = 14.93).
Percent distance error and estimated exponents were calculated separately
for each participant and were subsequently used as dependent variables in
a 2 (scale – large/small) x 2 (gender) ANOVA. This analysis showed no
significant differences between groups for either accuracy measures or
power model exponent estimates. Gender, and its interaction with the
apparent scale was similarly not significant.
The mean percent of overestimation for the small scale group (M =
2.2%, SD = 9.1) was not significantly different than for the large-scale
group (M = 2.5%, SD = 14.4) (F(1,13) = 0.25, p = .88). Similarly, the
mean estimated exponent for the small-scale group (M = 0.84, SD = 0.12)
was not significantly different than that of the large-scale group (M =
0.93, SD = 0.14) (F(1,13) = 0.67, p = .43). The data are shown in figure
*DE3-1.
Figure DE3-1
[ return to top ]