Maze experiments
A series of four experiments have aimed at discovering differences between
the spatial representations of virtual environments vs. real world
environments. In general, these experiments teach people about the
configuration of a maze in a VE and test their knowledge of it in a
real-world maze in our lab.
Experiment one: Development of route
representations
Experiment two: Development of route
representations
Experiment three: Survey representations
Experiment four: Factors affecting
representations
[ return to top ]
Distance estimation experiments
Three experiments have examined characteristics of distance
perception in VE’s. Using an ecological approach (in which participants
are able to navigate freely through the VE before estimating distances)
these experiments have tested the effects of several variables on people’s
estimates of exocentric (interobject) distances. A variety of
psychophysical models have been tested. Generally, as with real world
distance estimations, a (Steven’s) power function fits the data best. In
general, exocentric distances are overestimated; yet, exponent estimates
from best-fitting power functions are typically less than one.
The first two experiments have been written-up and are currently
under review at Presence. A postscript file of the manuscript is
available for downloading here. Figures are
available here. Or view the .pdf
file here:
Experiment one: Immersion, perspective cues,
barriers, and gender
Experiment two: GFOV, feedback, and gender
Experiment three: Apparent scale
[ return to top ]
Using VE’s for training a small scale assembly task
Two experiments have examined the utility of VE’s for training
small-scale assembly tasks. Imagine designing a system to train a
mechanic how to assemble an engine. Paper tutorials are perhaps the most
common means of training; however, video and virtual environments offer
promising alternatives. These experiments were designed to test the
relative utility of VE’s, video, and paper-based training. Previous
research has shown little effect of VE’s in training motor skills (Kozac
et al, 1993, but see Todorov, Shadmehr, & Bizzi, 1997). We feel that a VE
trainer will be more effective for training spatial skills that have a
large cognitive component. The task we trained involved using
three-dimensional spatial knowledge of the placements of blocks in a
‘circuit board.’
Experiment one: VE training aids retention
Experiment two: The effect of interactivity
[ return to top ]
Individual difference variables in virtual environment
training
Input device effect on orientation acquisition
[ return to top ]