The Effects of Video Game Experience on Spatial Navigation Performance:
A Study of Males in a Virtual Reality Maze
A. Earls, S. Evans, C. Johnson, H. Louie & T. Wong
McMaster University
VIDEO GAME EXPERIENCE
March 29, 2001
Psychology 3L03
Dr. Hong-jin Sun
Micheal Eckert, Bin Xu
ABSTRACT
Video
game experience has been proposed to show improved performance on spatial
virtual reality test (Dorval 1986; Moffat 1998). In the following study, it was hypothesized that video game
players have better spatial navigation and memory abilities in a virtual
reality maze compared to control subjects, indicated by faster maze completion
times. Eight male subjects were used in
this experiment, and the data associated with two of the subjects was
discarded. Three subjects were video
game players (played greater than five hours per week) and three subjects were
non-video game players who served as controls (played less than one hour per
week). The subjects navigated through a
virtual maze on a stationary bicycle with a Helmet Mounted Display (HMD, V8,
Virtual Research). One three-minute
practice trial was permitted to allow the subjects to familiarize them with
operation of the bicycle through the maze, provided by HMD. After the practice trial, four subsequent
test trials were run. Each trial was
timed until the maze had been completed, with the goal being navigation through
the maze at the fastest speed possible.
Video game players had lower average test times when compared to
non-video game players, which is significant through statistical
observation. The results of spatial
memory comparisons between video game players and non-video game players were
inconclusive. In essence, previous
experience with spatial navigation programs is advantageous for males to
decrease their navigation time in a virtual world. However, the strength of spatial memory in more experienced
individuals cannot be proven to be increased compared to controls under current
experimental conditions.
INTRODUCTION
Current studies suggest that video games have great potential as educational tools and researchers have proposed that spatial skills are the cognitive faculties most likely to benefit from video game practice. Linn and Peterson (1985) suggest that there are three different types of spatial skills: (1) spatial perception, (2) mental rotation, and (3) spatial visualization. Spatial perception refers to the ability to infer the orientation of an object with respect to one’s own orientation, while mental rotation can be defined as the ability to imagine the rotation of a visual stimulus. Spatial visualization is a more complex, multidisciplinary process, characterised by multi-step manipulation of spatially presented information using numerous solution strategies (Okagaki & Frensch, 1994, p. 33-34).
Virtual reality (VR) is a complex technology relying on the integration of a vast array of disciplines and it is currently receiving attention from neuroscientists studying the human brain (Foreman et al., 2000). Spatial ability is often measured by analyzing an individual’s ability to navigate through a virtual maze, a process that requires goal-independent memory of space. In this type of environment, information regarding spatial relationships is provided by visual motion sequences generated by simulated movements in a virtual maze, and the subject’s own movement decisions define the path through the maze (Gillner & Mallot, 1998, p. 445). Gillner and Mallot (1998) found that human subjects are able to learn a virtual maze from a series of local views and movements, and that such learning relies partially on the acquisition of three- dimensional (3-D) configuration knowledge. Foreman et al. (2000) also demonstrated that exploration of a virtual maze prior to real maze navigation involves a clear transfer of spatial information from the previously learned virtual version. These studies provide evidence that virtual maze navigation performance is a valid measure of human spatial ability.
Existing studies examining the impact of video game experience on spatial ability have yielded conflicting results. For example, Gagnon (1985) studied subjects with 2-D and 3-D action-adventure video game experience, and found limited gains in spatial skill performance on static paper-and-pencil measures of spatial ability. The absence of spatial skill transfer from video games to the testing variable in such studies may occur due to the unsuitable measure of the spatial skill under scrutiny, or a lack of similarity between the spatial tasks required by the video game and the experimental measures (Okagaki & Frensch, 1994, pp. 33-34). However, current research provides strong evidence that video game experience improves certain spatial skills, providing that there is an appropriate correlation between the specific spatial skills required during video game playing and testing.
Spatial visualization has been observed to be a trainable technique in humans, as shown in previous literature (Dorval & Pepin 1986), and is of particular importance to this study. As previously discussed, this form of visualization can be thought of as “the ability to rotate mentally, manipulate and twist two and three dimensional stimulus objects” (McGee, 1979, p. 159). Dorval and Pepin (1986) demonstrated a positive correlation between consistent video game playing and performance on spatial tests in adults.
Okagaki
and Frensch (1994) also studied the relationship between video game practice
and mental rotation and spatial visualization time. They conducted an experiment in which individuals were instructed
to play the video game Tetris (a game
requiring rapid movement of different shaped blocks in a 2-D plane). Pre- and post-test paper-and-pencil measures
indicated that Tetris improves mental rotation time and spatial visualization
time. These results suggest that
practice with other types of video games may lead to increased performance in
other spatial tasks (such as virtual maze navigation). A number of current studies are subsequently
exploring this relationship.
Moffat et al. (1998) conducted an
experiment that analyzed the effect of previous 3-D video game experience on
spatial navigation performance in virtual mazes, although their study focused
on a comparison between sex differences in spatial ability. Subjects were instructed to complete five
virtual maze trials, and their spatial performance was measured through
analysis of the time taken to navigate through the maze, as well as the number
of errors made during this time. However,
only the time variable was analyzed in determining the effect video game
experience. Results indicated that
males typically commit fewer errors and move more quickly when navigating
through a virtual maze than females, suggesting male superiority in such
spatial tasks. Results also indicated
that maze completion time was significantly faster in video game players
compared to non-video game players, indicating that video game experience
improves spatial navigation performance in virtual environments.
It is necessary to expand and modify the research design undertaken by Moffat et al (1998) in order to isolate the effect of video game experience on spatial ability. Although Moffat et al. (1998) tested for video game experience, no analysis was performed using controls of the same sex. Male versus female spatial ability was the only parameter examined, and by chance, the males used had more video game experience than the females. The use of same-sex control subjects would minimize the effects of sex differences in spatial ability, as it is important to eliminate confounding variables that may alter the relationship between video game experience and spatial navigation performance. It is also important to note that while females rely primarily on landmark information for spatial navigation, males more readily use both landmark and geometric information. Furthermore, the subjects tested by Moffet et al. navigated through the virtual reality mazes using a keyboard. This may have biased the experiment, as some subjects may have had more advanced fine motor ability. Those individuals with this ability many have had an advantage over those lacking these skills with respect to completion time and navigation skills. Finally, a time limit of five minutes was put on each trial. Although this might have helped to alleviate a subject’s frustration in being unsuccessful, it did not allow for full exploration of the maze. This result may not have permitted subjects to view the maze to their full potential on each trial, which could have resulted in subsequent memory formation for the correct route.
In this study, subjects were required
to navigate through a virtual maze while riding a stationary bicycle. All subjects were male in order to reduce
variation due to inherent sex differences in spatial navigation strategy and
ability, and both non-video game players and video game players were
examined. A non-video game player was
defined as one who played less than one hour of video games per week, and a
video game player was a subject who played more than five hours of video games
per week (of 3-D action/adventure games).
Eight subjects were tested, however the data associated with two of
these eight individuals were discarded due to inadequate completion of the
required trials. The remaining six
subjects were divided into two groups: three were rated as video game players
and three were observed to be non-players.
Each player was given a practice session and four learning trials to
navigate the maze. The practice and
test trials were performed on different mazes, and time was measured as each
subject completed the trials. It was hypothesized that video game players would
complete the virtual maze more quickly and efficiently than controls. Verification of this hypothesis would
provide evidence that video game experience can facilitate superior spatial
ability in males, which in turn may be beneficial in tasks performed outside of
a virtual world. The current study
will attempt to analyze spatial ability in terms of spatial navigation, the
ability to move through the maze; and spatial memory, the ability to remember
the maze layout to efficiently complete the trials.
METHOD
Subjects:
Seven male, undergraduate students at McMaster University (aged 21 to 23 years), and one male employee in the psychology Department of McMaster University (between 35 to 40 years of age), were utilized in this experiment. Two potential subjects did not complete the five trials, one due to frustration after the first trial on the test maze, and the second subject, due to an apparatus problem. Their data was discarded. A variety of ethnicities were represented, including Irish-Egyptian, Asian, Indian, Italian, and Caucasian. Of the six subjects who participated fully in the experiment, three were video game players (played more than 5 hours of 3-D action/adventure video games per week), and three were are non-video game players (played less than 1 hour of video games per week). The latter three served as controls. The information regarding each subject’s characteristics and video game experience was obtained upon completion of a questionnaire administered prior to testing.
Apparatus:
Subjects were
required to navigate through a virtual maze on a stationary bicycle. The bicycle could facilitate turning and stopping
motions, but the subject could not move backwards. Subjects were able to adjust gears and the height of the bicycle
seat. Each subject wore the Helmet
Mounted Display (HMD, V8, Virtual Research) which provided visual information
to the subject as they navigated through the virtual maze. The visual scenes
viewed through the HMD were updated by an SGI O2 computer through the inputs
from two sensors mounted on the bicycle: one sensor on the steering column (to
signal movement direction) and another sensor on the rear wheel to detect the
wheel rotation speed (to signal the subject's speed of forward motion). The
error information and route choice were stored on the computer for later
analysis.
Procedure:
Prior to beginning the testing session, each subject was instructed to adjust the helmet so that it sat firmly and tightly on their head, as well as to adjust the eyepieces to obtain a sharp image. Upon pedalling the bicycle, subjects entered and navigated though the virtual maze. The first three minutes of riding consisted of the subject going through a practice maze. It was explained to the subjects that this trial was not important as far as finding the finish, but that they should spend the three minutes focusing on the manipulation of the bicycle, and adjusting to the sensations associated with using VR. The practice maze is illustrated in FIGURE 1.
INSERT FIGURE 1 HERE
The next four test trials were run on a different maze, and no time limit was assigned. This trial maze consisted of five dead end routes that the subject could chose to explore. The goal for each subject was to find the shortest route possible out of the maze as quickly as possible. This test maze can be seen in FIGURE 2. The time was recorded for each individual trial, and a copy of subject’s path was saved on the computer.
INSERT FIGURE 2 HERE
RESULTS
The information from the current study
measures the degree of spatial learning acquired by each subject in each group,
the non video game playing, and the video game playing group. The length of time required for each subject
to complete the maze was used as a measure of spatial learning. This measure was used because there was no
notable difference in the velocity with which each subject traveled through the
maze. Furthermore, there was no notable difference in the number of errors made
while navigating through the maze among all subjects. Errors considered
included the number of times a subject entered the wrong path more than once,
the amount of time the subject spent in the wrong path, and lastly, and the
frequency with which the subject bumped into a wall or paused. These inferences can be made by analyzing
the bird’s eye maze trajectories, as the distance between the dots, which
illustrate the route taken, indicate the relative velocity of movement (FIGURE
3). It was hypothesized that increased
spatial learning would result in a decreased completion time and that the video
game playing group would have a lower completion time than the video game
playing group.
INSERT FIGURE 3 HERE
As
previously mentioned, two subjects were excluded from data analysis. Subject 7 was excluded, as he could not
complete the maze on the first trial, due to frustration. Subject 8 was excluded due to a malfunction
in the bicycle equipment at the time of his trials.
The average times for the non-video
game players for the first, second, third and fourth trails were 96s, 41.3s,
35.7s and 35.3s respectively. The
average times, in seconds, for the video game players for the first, second,
third and fourth trials were 53s, 35.3s, 26s and 23.7s respectively. The average times for each trial completed
by the video game players are lower than the average times obtained from the
non video game players (FIGURE 4).
A paired t-test (two sample for means)
was used to compare the average times for the second, third and fourth trials
performed by the video game players versus the non video game players. This statistical test shows that the
difference between the two groups was significant. (tstat=5.49, tcrit=2.20 one tailed,
p=0.016). Thus,
on average, the video game players completed the trial maze more quickly in
comparison to the non video game players.
INSERT FIGURE 4 HERE
The difference between the learning
trial (the first trial) and the average subsequent trials can be used to
measure spatial memory. The difference
between the first trial and the averaged subsequent trials of subject one and
subject five (non video game playing group) was 75.7s. The difference between the first trial and the
averaged subsequent trials of subject four (video game playing group) was 63s.
This difference suggests that the video game playing group had learned the maze
better, in terms of memory (FIGURE 5).
Subject six (non video game players) and subjects two and three (video
game playing group) were excluded from this particular analysis, as all three
of these subjects navigated through the maze without error on the first trial,
and as such is not a reliable measure of spatial memory.
INSERT FIGURE 5 HERE
REFERENCES
Dorval M. & Pepin M.
(1986). Effect of playing a
video game on a measure of spatial
visualization. Perceptual
Motor Skills, 62, 159-162.
Foreman N., Stirk J., Pohl J., Mandelkow L., Lehnung M.,
Herzog A., & Leplow B. (2000). Spatial information transfer from virtual to
real versions of the Kiel locomotor maze.
Behavioral Brain Research, 112(1-2), 53-61.
Gagnon, D.
(1985). Videogames and spatial
skills: An exploratory study. Educational
Communication and Technology Journal, 33, 263-275.
Gillner S. & Mallot H.
(1998). Navigation and
Acquisition of Spatial Knowledge in a virtual maze. Journal of Cognitive Neuroscience, 10(4), 445-463.
Moffat S., Hampson E., & Hatzipantelis M. (1998).
Navigation in a “Virtual” Maze: Sex Difference and Correlation With
Psychometric Measures of Spatial Ability in Humans. Evolution and Human Behavior, 19, 73-87.
Okagaki L. & Frensch P.
(1994). Effects of Video Game
Playing on Measures of Spatial Performance: Gender Effects in Late Adolescence. Journal of Applied Developmental
Psychology, 15, 33-58.
Figure Captions
Figure 1
Practice Test Maze
This maze is a
simple virtual reality maze. Utilized
to allow the subjects to become familiar with the maze; the goal of this maze
was not for an individual to reach the end, but to coordinate navigation
through the maze with the bicycle and the helmet. Subjects were allotted a maximum of three minutes to explore the
maze.
Figure 2
Test Maze
This maze was
more challenging than the practice maze.
Subjects were timed during their navigation through the maze, and four
trials were performed by each subject.
Figure 3
Subject-Specific Maze Trajectories
There was no
notable difference in the velocity with which each subject traveled through the
maze. Furthermore, there was no notable difference in the number of errors made
while navigating through the maze among all subjects. Errors considered
included the number of times a subject entered the wrong path more than once,
the amount of time the subject spent in the wrong path, and lastly, and the
frequency with which the subject bumped into a wall or paused. These inferences can be made by analyzing
the bird’s eye maze trajectories, as the distance between the dots, which
illustrate the route taken, indicate the relative velocity of movement.
Figure 4
Non-Video Game Players
This graph
depicts the test trial times for all non-video game players.
Video Game Players
This graph
depicts the test trial times for all video game players.
Figure 5
Spatial Memory
This graph
compares the learning (practice) trials of video game players with non-video
game players. The average of learning
times for video game players was subtracted from the average of learning times
for non-video game players. These
totals were subtracted and their result represented in a line graph. Subject six, a non-video game player, and
subjects two and three, video game players, were excluded from this
analysis.
FIGURE
1
FIGURE
2
FIGURE 3
Subject 1
Subject 2
Subject 3
FIGURE 3 cont.
Subject 4
Subject 5
Subject 6
FIGURE
4
FIGURE
5