LIBRARY
Michigan State
University

 

 

 

PLACE IN RETURN Box to remove this checkout from your record.
To AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2/05 c;7EIRD/Datofia.Im-p. I 5

—...—-—-—————

_ I,{I..A. -7 degree in

This is to certify that the
thesis entitled

COGNITIVE SKILL AND GENDER
IN 2D AND 3D VIDEO GAME PLAY

presented by

SEARLE HUH

has been accepted towards fulfillment
of the requirements for the

___ Cotnntunicaticn _ W

 

JAN 30 2007

 

Date

MSU IS an Affinnat/‘ve Action/Equal Opportunity Institution

COGNITIVE SKILL AND GENDER IN 21) AND 3D GAME PLAY
By

Searle Huh

A THESIS
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
MASTER OF AR'I‘S

Department of COI]1[‘IIUI]I«":1II(,III

2007

ABSTRACT
CONITIVE SKILL AND GENDER IN 2D AND 3D VIDEO GAME
By

Searle Huh

This study examined the contribution of cognitive skill and gender in game
preference and game enjoyment. In particular. this study focused on the notion that
cognitive ability is a stronger predictor of game performance and enjoyment than
gender. By comparing differences of performance and enjoyment in 3D game and 2D
game, this study set three hypotheses. Main findings of this study were categorized
largely into three. First, gender is a significant factor affecting video game
performance. Male gamers are more likely to play better than girls. Second, gender
and dimensionality of video games are related. Males performed better than females,
but the difference was more notable in 3D game play than in 2D game play. In terms
of enjoyment, males enjoy 3D game more while females enjoy 2D game more. Third.

game performance and enjoyment could not go together.

TABLE OF CONTENTS

LIST OF TABLES ............................................................................................................. IV
LIST OF FIGURES ............................................................................................................. V
INTRODUCTION ........................................................................................................... 1
LITERATURE REVIEW AND THEORETICAL FOUNDATION °°°°°°°°°°°°°°°°°°°°°°° 3
Gender and Video games ................................................................................................ 3
Alternative thinking- Cognitive SkIIIS ............................................................................. 5
Variable cognitive SkIIIS .................................................................................................. 8
PURPOSE OF THIS STUDY ........................................................................................... 10
RESEARCH MODEL ....................................................................................................... l 1
HYPOTHESES ................................................................................................................. I I
METHOD .......................................................................................................................... 12
Sample description ........................................................................................................ 12
Experimental CICSIgn ...................................................................................................... 12
Measurement ......................... ' ........................................................................................ l 3
RESULTS .......................................................................................................................... l 7
Gender and video game ................................................................................................. 17
Cognitive Skill and VICICO game ..................................................................................... 18
Gender and cognitive Ski“ interaction ........................................................................... 22
CONCLUSION ................................................................................................................. 26
DISCUSSION ................................................................................................................... 27
APPENDIX ....................................................................................................................... 32
BIBLIOGRAPHY ............................................................................................................. 33

LIST OF TABLES

TABLE 1. Performance mean difference in males and females --------------------------------------- 17
TABLE 2. Enjoyment Mean difference in males and females ----------------------------------------- 18
TABLE 3. Performance mean comparison by dimensionality and cognitive skill ----------- 20
TABLE 4. Enjoyment mean comparison by dimensionality and cognitive skill ------------- 22

TABLE 5. Performance comparison by gender, dimensionality, and cognitive skill ------- 23
TABLE 6. Enjoyment comparison by gender, dimensionality, and cognitive skill ---------- 25

TABLE 7 Statistics ofobjects in 2D and 3D ChCX-QUOSI ............................................... 32

LIST OF FIGURES

FIGURE 1. Research model ofthis study ......................................................................... 11
FIGURE 2. Hypothesis 1 ................................................................................................... 11
FIGURE 3. Hypothesis 2........... ........................................................................................ 12
FIGURE 4. Hypothesis 3 ................................................................................................... 12
FIGURE 5. A sample Of mental rotation test .................................................................... 14
FIGURE 6. Screenshots of 2D and 3D games used for the experiment --------------------------- 16
FIGURE 7. The role of mental rotation ability in 2D and 3D game performance ------------ 21
FIGURE 8. The role of mental rotation ability in 2D and 3D game enjoyment --------------- 23

Cognitive skill and gender in 2D & 3D video game play
Introduction

Computer and video games are one of the most popular leisure time activities for
children, adolescents, and young adults in Western and Asian societies (Hartman &
Klimmt, 2006). They have contributed the formation of current cultural landscape,
finding their way into fiction, movies, television, and music (Sherry, Rosaen,
Bowman, & Searle, 2006).

Gender difference is one of the main issues in video game study. Generally, it
has been found that video games are liked more and played more by males than by
females (i.e., Buchman & Funk, 1996; Griffiths, 1991; Kaplan, 1983; Philips, Rolls,
Rouse, & Griffiths, 1995; Wright et al., 2001 ). Some argue that majority of available
games on the market have content that is more desirable for male players (AAUW,
2000; Cassell & Jenkins, 1999; Kafai, 1998; Miller, Chaika & Groppe, 1996) and
these games may make female players reluctant to play (Kafai, 1998). Other scholars
focus on gender differences in perceptions, social stereotypes, and circumstances
encountered by gamers (Funk & Buchman, 1996). Finally, biological sex differences
in cognitive skills have been suggested as contributing to the differences for video
game preference and play between the genders (Chu, et al., 2004).

Recently, however, Nielsen’s 3rd annual Active Gamer Benchmark study (2006)
showed that 64 percent of online game players are female, although 70 percent of
total video gamers are still male. Even admitting some differences between general
video games and online games, it seems to be implausible to insist absolute male

dominance in video game use. With increasing numbers of female players, gender is

losing its explanatory power in gender differences in video game. Additionally, more
game manufacturers are making games targeted at the female market (i.e., The Sims
and Final Fantasy X), suggesting that male orientation of content may no longer
explain gender differences. Instead of arguing that video games are liked more and
played more by males than by females, perhaps researchers interested in how gender
realtes to video game appeal should focus not on why one gender likes video games
more than another, but should examine the on differences in the types of games found
appealing to males and females and explanations for these differences.

Because video games are puzzles that involve a variety of cognitive skills (e.g.,
mental rotation, color recognition, object location memory, etc.), different native
abilities in cognitive skills areas are likely to predict game playing ability (Sherry,
2004). Some of these cognitive skills show large differences by sex. For example, on
average, males are a full standard deviation better than females at 3D mental rotation
(Rahman & Wilson, 2003; Jones et al., 2003; Terlecki & Newcombe, 2005; Levine et
al., 1999; Bosco et al., 2004). Players with higher level of 3D skill significantly
outperform players with lower levels of 3D skill on games demanding 3D ability
(Sherry, Rosaen, Bowman, & Huh, 2006). Due to the innate differences in cognitive
skills. a male or female player who has unique sets of advantageous skills over the
other is more likely to preference for game genre that require that skill and to perform
better in the corresponding games.

This study examines the contribution of individual cognitive skill and gender in

game performance and game enjoyment. Specifically. the study advances the notion

that cognitive ability is a stronger predictor of game performance and liking

regardless of gender, as predicted in media flow theory (Sherry, 2004).
Literature review and theoretical foundation

Gender and video game

It is widely believed that the violent nature of many video games alienates girls
(Malone, I981; Greenfield, 1996), turning them off video games in general (Sherman,
1997). In addition, an increasing degree of realism in the majority of violent computer
games would seem to aggravate the gender gap (Hartmann & Klimmt, 2006).
Although violence in video games may be one of the touchstones of gender
differences in video games, there are many different ways to approach what the
violence in a video game is and how to measure it (i.e. Sherry, 2001; Smith, Lachlan,
& Tamborini, 2003). The boundaries and components of violence may not always be
identical in most studies.

Rather than exploring the degree of violence in video games, Kafai ( 1996), tried
to explore what children wanted in video games by studying what they produced
when given the skills and opportunities to create their own games. She found that
boys designed games having greater resemblance to commercially available violent
games such as Nintendo while girls did not. In addition, she concluded that the girls
had “no particular interest in pursuing video game playing because they did not like
the games, their content, and their violent aspects” (p.62). Thus. existing (violent)
games generally reflected boys”, but not girls’ tastes (Cassel & Jenkins, 1998).

However, violent games are not universally preferred by boys and rejected by

girls. The different preference for the violent content is likely to be reflected on

different genre preference. The preferences for different types of violence appear to
be generally consistent with the findings of genre preference studies. Morlock et al.
(1985) reported that males prefer games with fast action and aggressive themes, while
females are likely to prefer more “whimsical” games. The result of genre preference
studies can be summarized that male generally like shooters, fighters, sports,
racing/speed games, fantasy/role playing. action-adventure, strategy, and simulations
while females prefer board games, arcade, quiz & puzzle, and kids games (Roberts et
al., 1999; Sherry, Lucas, Rechtsteiner, Brooks, & Wilson, 2001). Kafai (1996)
showed consistent gender differences, especially different preference for kinds of
games, game characters, and game worlds. Further, several studies showed that
violent games remained consistently popular and were chosen as favorites for both
boys and girls (Funk & Buchman, 1994; Buchman & Funk, 1996). It seems to be a
contrasting argument, but it is not. While the mean scores of violent games liking for
females is lower than mean scores for males, there remain girls that like playing
violent games Additionally, gender differences have also been found in the nature of
the preferred violence: boys were more likely to prefer games with more realistic
“human violence.” while girls were more likely to prefer games with cartoon or
“fantasy violence" (Buchman & Funk, 1996; Funk. 1993; Funk & Buchman, I994,
I996; Griffiths, 1991).

It has been suggested that girls seek different types of stimulation from video
games than boys (Bruner, Bennett, & Honey, 1998; Kafai, 1996). Funk, Jenks, &
Bechtoldt (2001) showed that even when the game content seemed violent, girls

defined them as "action" or "adventure" games. Research suggests that girls do not

find this violence appealing (Cassel & Jenkins, 1998). The empirical evidence
confirms that boys are more likely to play games requiring aggressive competition
(Heller, 1982, cited in Morlock et al. 1986; Keisler et al., I985; Lin & Lepper, 1987).
Kafai (1996) found gender differences in feedback resulting from a player’s action
during the game. She reported that the feedback in boys’ games was overwhelmingly
violent, whereas the feedback in girls’ games was overwhelmingly nonviolent. In
general, girls do not specially enjoy ‘shooting bad guys and monsters” (Klawe, et al.,
2002, p.21 I).

If game content deters females from game play, what about playing non-violent
games? Can females play better than or at least equivalent to males? This study

purports to answer to the question.

Alternative thinking: Cognitive skills

Funk and Buchman (I996) noted that although studies have consistently found
gender differences in video game play, “The origin of gender differences in game-
playing habits has not yet been established” (p. 27). If these differences are not
entirely due to content, what else could they be due to? Gender differences have
always been explained in television research as content based, but games are different
from television in that they are more than simple content. Games require interaction
of the user. Therefore, differences in how individuals interact with games may
determine differences in game genre preference. Several studies have shown that one
of the main experiential features that games provide for users is the intellectual

challenge of beating the game (Vorderer, Hartmann. & Klimmt, 2003; Sherry, 2004).

Therefore, gender differences in how people experience the challenge of the game
may improve explained variance in studies of game preference.

One compelling possibility is in the area of cognitive abilities. There are a
variety of cognitive skill sex differences that would facilitate rapid success in
particular genres and not in others (Sherry, 2004). The main sex differences that
cognitive scientists have studied include spatial rotation, color memory,
disembedding, field independent spatial perception, object location memory, targeting,
verbal fluency, and verbal memory (Kimura, 1999). Male advantage is found in
spatial rotation, disembedding, field independent spatial perception, and targeting,
with females being generally superior on the other cognitive tasks (Kimura, 1999).
The differences have been supported partially by the studies that found the gender
differences in visual and spatial tasks (i.e. Battista, 1990; Macccoby & Jacklin, 1974;
Keisler et al., 1983), cognitive strategies to play a video game (Blumberg & Sokol,
2004), and object memory and targeting (Kimura, 1999).

Out of many cognitive skills, gender differences in spatial rotation abilities have
been most frequently reported and studied (Bosco, et al., 2004; Caplan, et al., 1985;
Levine et al., 1999). It is commonly believed that males’ spatial rotation abilities are
superior to females’ (Caplan, et al., 1985). Although some meta-analyses showed that
the most robust sex differences are found on spatial tasks (Linn & Peterson, 1985;
Voyer et al., 1995), it is interesting that the gender effects in visuo-spatial tasks are
not homogeneous (Bosco et al., 2004). Rather, the magnitude, or even the existence,
of the differences between genders varied over the characteristics of the task. For

example, Vercci and Girelli ( 1998) showed that active spatial skill (transforming) has

induced much larger gender gap than passive spatial skill (3D memorizing). In
relation to these variations, it is reasonable to think that spatial skill has several
dimensions in it, as Linn and Petersen (1985) argued in their meta-analysis. They
found three dimensions in spatial skill: they are spatial perception, mental rotation,
and spatial visualization. The mean effect size of each was 0.44 (p<.05), 0.73 (p<.05),
and 0.13 (p>.05) respectively. Therefore, they showed that sex differences in spatial
perception and mental rotation are robust (Voyer & Bryden, 1995). Spatial perception
was defined as ‘the ability to determine spatial relations despite distracting
information (Linn & Petersen, l985)’ and mental rotation was defined as ‘the ability
to rotate quickly and accurately two- or three dimensional figures in imagination
(Linn & Petersen, 1985)’.

Following Linn and Peterson’s (1985) categorization, a majority of studies
dealing with spatial skill and its gender differences have focused on mental rotation,
rather than spatial perception, which has less robust results in gender differences
(Levine, et al.. 1999) and less clear measurement. In addition, with reference to
Vercci and Girelli (1998)’s active-passive distinction, mental rotation seems to be
equivalent to the active ‘transforming’ while spatial perception to the passive
‘memory’. Therefore, it is reasonable to think that mental rotation needs more active
cognitive participation and correspondingly induces a larger gender gap than spatial
perception. This is linked to several scholars arguments. Levine et al. (1999)
mentioned that the manifestation of sex differences in spatial skill would be on tasks
that involve mental rotation. Quaiser-Pohl et al. ( 2006) also argued that one of the

largest and most reliable gender differences in favor of males can be found in mental

rotation. Especially a significant difference between males and females in 3D rotation
ability has been reported while little difference found in 2D tasks (Hyde & McKinley,
1997). Therefore, gender gap would be more conspicuous in 3D tasks than in 2D
tasks.

For video games, this difference is more likely to be meaningful. Due to males’
superior 3D mental rotation ability, it would make it easier for males to achieve flow
state, which is a state of the pleasure found by immersion in an activity, in playing
games that rely heavily on 3D graphics, such as fighters, shooters, and sports games
(Sherry, 2004). Contrasting to males, females are more likely to have more
difficulties playing the 3D games and feel more frequent frustration than males. For
example, most paths in 3D games need to be translated into 2D bird-eye’s view for
finding the right way to exits or to gates to next level. If a user has low mental
rotation ability to translate the 3D blocks, he or she would get lost easily. Therefore, a
user with high mental rotation ability is more likely to show better performance than
one with low ability. In addition, due to the gender difference of 3D mental rotation

ability, males are more likely to perform well in 3D games than females.

Variable cognitive skills

Although there are many studies that clearly show gender differences in
cognitive skills, the causes of these differences still remain unclear (Quaiser-Pohl et
al.,2006). The question is linked to the nature of the cognitive skill, innate or trained,
or nature-nurture controversy. Some scholars have tried to explain these findings

focusing on biological factors like hormones and genetic influences (Kimura, 1999)

while some emphasize more recently socio-cultural factors on women’ performance
in visio-spatial tasks (Richardson, 1994; Bosco et al., 2004). Beyond the controversy,
what is more important is that experience has a clear effect on spatial skill (Quaiser-
Pohl et al.,2006). That is, gender differences in cognitive skill, especially spatial skill,
is not static over time but changeable. The preexisting sex differences, including
cognitive skills such as mental rotation and maze navigation, can be attenuated
through practice (e. g., DeLisi & Cammarano, I996; DeLisi & Wolford, 2002; Lawton
& Morrin, 1999) or differential exposure to this medium (Subrahmanyam &
Greenfield, 1994).

Accordingly, gender may not be sufficient as an explanatory variable for finding
the difference in video games. In other words, individual differences in cognitive
processing may be one of the factors explaining the relationship we see between the
sexes in genre preference (Sherry, 2004). One flow study showed that individual
differences in cognitive skill predict game-playing ability (Sherry et al., 2004).
Importantly, this relationship remains after controlling for sex and amount of game
play (game addiction) (Sherry et al. 2006). This may lead some alternative
possibilities in gender differences in video games. First, individual variation of
cognitive skill may not ignorable. Although many explanatory variables for different
game uses have been reported, individual variables like personality, experiences, and
cognitive skills could be more influential as an explanation for differences in video
game play and enjoyment than grouped variables like gender, nationality, and
occupation. Second, gender may only have a significant influence differences in game

play only when interacting with another variable like cognitive skill.

Purpose of the study

There is strong evidence in the literature for gender differences in game genre
preference. The content of a video game then may be a critical variable to
differentiate unique game preference and play of males and females. Especially, the
violence of most video games is believed to induce gender differences in video game
play. However, there were not many studies that focused on the gender difference in
non-violent game plays. If content of video games matters, gender difference in
playing non-violent games will be narrower. For female gamers, performance and
enjoyment could be bettered by reducing violent contents that were believed to
impede their video game play. For male gamers, game performance and enjoyment
could be reduced by diminishing violent components that were believed to facilitate
their video game play.

Content may not be the only explanatory variable. Especially due to specific
characteristics of video game, such as need for active participation and high-
interactivity, individual cognitive skills may be much more significant than with any
other entertainment media. It is also possible that cognitive skills explain video game
preference and play more than the content does. If so, individual differences can be
more influential than gender differences. Otherwise, (individual) cognitive skill may
interact with the gender difference. The purpose of this study is to test the
contribution of cognitive skills have in predicting game play ability and enjoyment in

2D and 3D games. especially with a non-violent game.

l0

Research model

Figure 1. Research model of this study

 

 

Independent Variables Treatment Dependent Variables
5/ "w TTT‘\\ ,x/T'w "I" x
7/ ' \\
, \
l Gender I ( Pertbrntance )
\
\ X /
.\ \\ \w’ / \.\~\\—’ //<
/'J_'-\ ~\ //....\\
/./ /’ x
3D /

 

/ Mental I, .
I\ Rotation /' Enjoyment

Skill _, x \
VV\\V\“/I// ‘1 ‘i‘ \ //

 

 

 

 

 

 

Hypotheses
H1: If gender matters, boys will like video games more and perform better on

video game than girls.

Figure 2. Hypothesis 1

 

2D & 3D games
Boys Girls

 

Low & High MRT High Low

 

H2: If cognitive skill matters, individual differences in cognitive skill (3D
rotation ability) should predict game scores and enjoyment by showing that
high MRT people perform better in both 2D and 3D game. just by a wider

margin in 3D game than 2D game.

Figure I. Hypothesis 2

Performance
& Enjoyment

H3: If cognitive skill and gender interact, performance and enjoyment of the

games should be similar to the following figure.

Figure 4. Hypothesis 3

Performance &
Enjoyment

 

 

 

 

20 game

3D game

 

 

 

 

2D game

3D game

 

 

" ' ' High MRT
—an MRT

 

 

 

- - -HighMRTMale
- I -HighMRTFanale
LowMRTMalc
+LowMRTchale

 

 

 

 

Method

Sample description

Participants were recruited from communication classes at a large, midwestem
university. A total of 146 participants were involved in the experiment. Of the
participants, there were 74 males (50.7%) and 72 females (49.3%). The ages of
participants ranged from 17 to 24, with a mean of 19.76. Their majors were
communication-related (29.3%), business-related (30.4), engineering-science (9.9%),
literature-education (13.1%), and others (17.3%). For their participation, each student

was given course credit.

Experimental design

The experiment was done in a computer lab, which has 20 desktop computers.
To prevent participants from watching other’s play, only 10 subjects were run at any
given time. Participants were placed such that they could not look at each others‘
screen while playing.

The experiment took place in three parts. First, participants were randomly
assigned to play one of two video games used this experiment for 15 minutes,
including game play instructions from the researchers. Second, participants answered
questions about previous game experiences, game liking. frustration, and difficulty, as
well as demographic information. Finally, the Vandenberg and Kuse (1978) Mental
Rotation Test (MRT) was administered for seven minutes. To prevent the possibility
that game play improves mental rotation skill, participants were randomly assigned to

do the MRT either before the game play or after game play.

l3

Measurement
Independent variables

Gender: Gender was measured by self-report. which reflected the gender that
participants most identified with.

MRT (Mental Rotation Test): Mental —rotation performance was measured by
MRT-A version that is composed of the figures of provided by Shepard and Metzler
(1978), and is essentially an Autocad-redrawn version of Vandenberg and Kuse
(1978) original test by Peters (1996). The MRT is a test for assessing respondents’
spatial perception skills, particularly 3D mental rotation. In other words, it is for
measuring the respondents’ ability to mentally picture and manipulate spatial contents.
The test consists of 24 items that requires respondents to find two identical figures
with a target figure out of four total choices that are shown from different angles
(Figure 3). Credit for each item is given only when both correct figures are chosen. so

the highest score a participant can receive is 24, with no partial points.

Figure 5. A sample of mental rotation test

% E. E is

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Treatment (manipulated games)

For the experiment, two different video games were used; a 3D game and a 2D
game. The 3D game was Chex-Quest, a cereal promotion which was developed by
Digital Café distributed through the purchase of specially marked packages of Chex
cereal. Chex—Quest is a ‘mod’ of Ultimate Doom, one of the landmark PC-based
video games of the 19903, lauded for its innovative 3D environments and character
graphics. Doom also introduced the first-person shooter, which placed the gamer
inside the main character’s perspective. As a ‘mod’, the Chex game behaves the same
as Doom, but has different characters, backgrounds, and goals. It is a first-person
shooter and the makers intended it to be a non-violent game, replacing the violent
components and stories associated with Doom with non-violent ones. Although it has
not been officially rated by Entertainment Software Rating Board (ESRB), it is clear
that Chex-Quest would not receive the same ‘M’ rating (Mature; suitable for ages 17
or older) of original Doom. In the game, the gamer assumes the role of a Chex cereal
guardian that is sent to the C hex home-world to teleport evil Flemoid monsters back
to their home dimension, using a full arsenal of spoons. gloves, and teleport guns.
Along the way, players can eat healthy snacks to revitalize their health.

The 30 C hex game was transformed into a 2D side-scrolling version by
converting maps, images, sounds, movements, and game logics. All of the graphics
and sounds in the 3D game were extracted and used to create the 2D game. The game
maps in 2D games were converted from the 3D games. The number of comers (left-
and-right in 3D and up-and-down in 2D), types of corners (3-ways and 2-ways), the
number and shape of monsters and items are nearly equivalent in 3D and 2D game

versions. Table 14 has a descriptive count of the components of each game.

IS

Figure 6. Screenshots of 20 and 3D games used for the experiment

Inuiimmiummmuimmiituuum'liuuumml

Iii

   

3D CHEX-QUES T 2D C HEX-QUEST

Dependent variables

Game performance: Game scores of both games were measured. The original
3D Chex-Quest has its own scoring system including percent of kills (number of
killed monsters out of all monsters), percent of items (number of items earned out of
all items appeared), percent of secret rooms found (number of secret room found out
of all secret rooms), and one measure of overall time. For the comparison with the 2D
version, only time to complete a level was used. Time was used as the lone indicator
of game performance because the cognitive skill of interest for this study was mental
rotation. For example, percent of monsters killed would be a measure of targeting,
which is a different cognitive skill.

Game enjoyment: Participants were asked to rate the game just played on game
enjoyment. Four questions from Sherry’s game enjoyment scale (Sherry et al, 2006)
were used to rate the game. The questions were “I enjoyed playing this game”, “,I

,9 6‘

would play this game longer if I had the opportunity , I would recommend this game

to a friend”, and “This game was fun”. The reliability of this scale was high enough to
use all four items (a =.939). The mean of four items was used for further analysis.
Other data collected: Age, major, and average hours (per week) of playing video
game were asked at the end of the experiment. These factors were not major variables
concerned in this study, but included for finding possible confounding influence on

the result.

Results
Gender and video game performance

Table I.Performance mean difference in males and females

 

 

Low & High MRT
Male Female
M = 428.36 M = 595.72
2D & 3D games SD = 232.20 SD = 285.77
N = 74 N = 72

 

Hypothesis I predicted that boys will perform better than girls in video game
play. An independent samples t-test was used to compare performance mean (seconds
to complete a level) of male participants and female participants. Results supported
the hypothesis 1, showing gender difference was significant, t(144) = -3.888, p < .001,
112 = .095. Males, who spent 428.36 seconds, completed the games faster than females,

who spent 595.72 seconds. That is, males performed better than girls in video game

play.

Gender and video game enjoyment

Table 2. Enjoyment Mean difference in males and females

 

 

Low & High MRT
Male Female
M = 2.74 M = 3.28
2D & 3D
SD= 1.45 SD= 1.57
games
N = 74 N = 72

 

Part of hypothesis 1 predicted that boys like video game more than girls. For
testing the enjoyment difference, an independent samples t-test was conducted to
compare the mean of enjoyment. The results did not support hypothesis 1, but showed
a significant difference in an opposite direction, t (144) = -2.115, p = .033, n2 = .030.
Males’ enjoyment (M = 2.74, SD = 1.45) was lower than females’ enjoyment (M =

3.28, SD = 1.57). However. the effect size was small.

Cognitive skill and video game performance

In this study, the MRT scores collected showed a significant difference between
males and females, t(144) = 3.913,p < .001, n2 = .096. Males (M: 9.62, SD = 5.72,
Range = 0-23) had significantly better mental rotation skill than females (M = 6.44,
SD = 3.88, Range = 0-16). The data collected showed the mean score of 8.05. The
distribution of score was slightly skewed to the right, due to a few participants who
have very high score or more male subjects than female subjects. Therefore, the high
and low distinction would used in analysis was based on median, which was 8.00.

When a distribution is skewed, a median is more appropriate estimate than a mean.

Those who have the score lower than 8 were categorized as low MRT group while
those have higher than 8 were classified as high MRT group. The mean score of MRT
in low MRT group and high MRT group were 12.01 (SD = 4.09, N = 74) and 4.03
(SD = 1.92, N = 73), respectively.

When measuring mental rotation ability in an experiment session, the possibility
that video game play improves mental rotation ability was a concern. 3D game play
may affect the result of MRT, especially when MRT was done afier the game play.
Therefore, subjects were randomly assigned order of MRT in the experiment session.
The result showed that there was no order effect of MRT, t(117) = 1.445, p = .151, n2
= .017. Those who did MRT before the game play (M = 8.75,SD = 5.430) did not
statistically differ from those who did MRT after the game play (M = 7.36, SD =
4.848). When divided by gender, the difference was still having no significance. For
males, those who had MRT-game order (M = 9.68, SD = 5.95) did not differ from
those who had game-MRT order (M = 10.18, 6.62), t(52) = -.227, p = .783, n2 = .000.
Females who had MRT-game order (M = 7.69, SD = 4.61) was not different from
females who had game-MRT order (M = 5.88, SD = 2.80) either, t(62) = 1.90, p

= .062. n2 = .051.

19

Table 3. Performance mean comparison by dimensionality and cognitive skill

 

 

 

 

Boys and Girls
2D game 3D game Total

M = 447.93 M = 694.43 M = 549.23
Low MRT SD = 264.10 SD = 220.42 SD = 274.12

N = 43 N = 30 N = 73

M=371.26 M=615.l2 M=476.72
High MRT SD = 224.55 SD = 258.73 SD = 267.48

N = 42 N = 32 N = 74

M=410.05 M=653.50 M=512.73

Total SD = 246.93 SD = 242.31 SD = 272.33
N = 85 N = 62 N = 147

 

Hypothesis 2 predicted that those who have high cognitive skill perform better in

playing 2D and 3D games by a wider margin in 3D game than in 2D game. For

testing hypothesis 2, a 2 * 2 factorial ANOVA was used to compare level of mental

rotation skill, dimensionality of the game, and the interaction between the two. The

analysis revealed a significant main effect for dimensionality of the game, F (l, 143)

= 36.313,p < .001, n2 = .203. This means that 3D game play (M = 653.50, SD =

242.31) took more time to complete a level than 2D game play (M = 410.05, SD =

246.93). However, the analysis did show a significant main effect for mental rotation

ability, F(1, 143) = 3.674,p = .057, n2 = .025. High MRT people (M = 476.72, SD =

267.48), on average, performed better than low MRT people (M = 549.23, SD =

274.12), but the difference was not significant. There was no significant interaction

effect for MRT and dimensionality, F ( 1 143) = .001 , p =.974, n2 = .000. This did not

support a wider margin in 3D game than in 2D game. Therefore, hypothesis 2 was not

supported in terms of performance.

Figure 7. The role ofmental rotation ability in 2D and 3 D game performance

 

 

 

 

700'— o MRT
--- High
LO'VV
q
600-— \
._\
g \\ ' ~
8500— ‘
. \
U) \\ O
400“ ‘\\
"o
300—1
I 1
3D ZD
DIMENSION

21

Cognitive skill and video game enjoyment

Table 4. Enjoyment mean comparison by dimensionality and cognitive skill

 

 

 

 

 

Boys and Girls
2D game 3D game Total
M=3.01 M=3.35 " M=3.15
Low MRT SD = 1.58 SD =1.72 SD = 1.63
N = 43 N = 30 N = 73
M = -.82 M = 3.07 ' M = 2.92
High MRT SD = 1.37 SD = 1.65 SD = 1.49
N = 42 N = 32 N = 74
M = 2.92 M = 3.20 M = 3.04
Total SD = 1.49 so = 1.68 SD = 1.56
N = 85 N = 62 N = 147

 

For testing enjoyment part of hypothesis 2, another 2 x 2 factorial ANOVA was
used to compare mental rotation skill, dimensionality of the game, and the interaction
between the two, when enjoyment was set as a dependent variable. The results did not
reveal a significant main effect for MRT (F [1, 143] = .820. p = .367. n2 = .006) or a
main effect for dimensionality of the game (F [1, I43] == 1.125, p = .270, n2 = .008).
There was no significant interaction effect for MRT and dimensionality on enjoyment,
F(1, 143) = .025. p = .874. n2 = .000. Therefore, hypothesis 3 was not confirmed in

terms of enjoyment.

Figure 8. The role ofmental rotation ability in 2D and 3D game enjoyment

Enjoyment

 

co
0
l

300-4

2.90—

2.80—

 

 

 

(o
O

DIMENSION

Gender and cognitive skill interaction for performance

MRT
--- High
LOW

Table 5. Performance comparison by gender, dimensionality, and cognitive skill

 

 

 

 

Boys Girls
2D 3D Sub total 2D 3D Sub total

M = 422.75 M = 558.36 M = 486.03 M = 462.85 M = 815.73 M = 588.88
Low

SD = 286.34 SD = 190.57 SD = 251.84 SD = 254.45 SD = 179.22 SD = 285.15
MRT

N=16 N=14 N=30 N=27 N=15 N=42

M = 333.88 M = 461.63 M = 389.05 M = 426.24 M = 839.46 M = 605.30
High

SD = 209.08 SD = 197.68 SD = 211.79 SD = 241.33 SD = 149.99 SD = 291.93
MRT

N=25 N=19 N=44 N=17 N=13 N=30

M = 368.56 M = 502.67 M = 428.36 M = 448.70 M = 826.75 M ‘-= 595.72
Total SD = 242.70 SD = 197.17 SD = 232.30 SD = 247.73 SD = 163.70 SD = 285.77

N=41 N=33 N=70 N=44 N=28 N=72

 

 

 

A 2 (Dimension) * 2 (Cognitive skill) * 2 (Gender) ANOVA was performed to
find the three—way interaction for gender, cognitive skill, and gender. However,
equality of cell variance should have been done before interpreting F values because
two cells (2D-high MRT boys and 2D low MRT girls) had relatively more samples
than other cells. Levene’s test of equality of error variance supported that it is
reasonable to assume equal variance, F (7,138) = 1.823, p = .106. (There was a
significant main effect for sex, F (1, 138) = 25.702, p < .001, n2 = .157 such that male
subjects (M = 428.36, SD = 232.30) performed better than female subjects (M =
595.72, SD = 285.77) as found in hypothesis 1. In addition, a main effect for
dimensionality was significant, (F (l, 138) = 46.223, p < .001, n2 = .251). This means
that subjects performed better in 2D game play (M = 410.50, SD = 246.93) than 3D
game play (M = 651.43, SD = 243.77), as found in hypothesis 2. A main effect of
MRT (F [1, 138] = 1.718, p = .192, n2 = .012) was not significant. A two-way
interaction effect was found. Interactions for sex and dimensionality (F [1, 138] =
11.024, p = .001, n2 = .074). Other two-way interactions, MRT and dimensionality (F
[1,138]=.120,p = .729, n2 = .001) and sex and MRT (F [1, 138] = 1.301,p = .256.
n2 = .009) did not show statistical significance. A three-way interaction of sex,
dimensionality, and MRT was not significant, F (1, 138) = .203, p = .653, n2 = .001.

It was not consistent with what original hypothesis 3 predicted. Gender,
dimensionality, and MRT did not consistently influence game play. Males generally
performed better than females, but the difference was more notable in 3D game play

than in 2D game play. Females who have high mental rotation skill (M = 839.46, SD

24

= 163.70) did not show better performance in 3D game play than males who have low

mental rotation skill (M = 558.36. SD = 190.57).

Gender and cognitive skill interaction for enjoyment
To test hypothesis 3 on enjoyment, another 2 (Dimension) * 2 (Cognitive skill) *

2 (Gender) ANOVA was performed. The equal variance across groups was supported

by Levene’s variance equality test, F (7.138) = 1.318, p = .246.

Table 6. Enjoyment comparison by gender. dimensionality, and cognitive skill

 

 

 

 

Boys Girls
2D 3D Subtotal 2D 3D Subtotal
M=2.13 M=3.44 M=2.74 M =3.54 M=3.01 M=3.35
Low
SD = .992 SD = 1.55 SD = 1.43 SD = 1.63 SD = 1.67 SD = 1.65
MRT
N=16 N=14 N=30 N=27 N=15 N=42
M=2.48 M=3.10 M=2.75 M=3.32 M=3.01 M=3.19
High
SD=1.4O SD=1.56 SD=1.49 SD=1.19 SD=1.84 SD=1.49
MRT
N=25 N=19 N=44 N=17 N=13 N=30
M =2.34 M=3.25 M =2.74 M=3.46 M=3.01 M=3.28
Total SD = 1 26 SD = 1.54 SD = 1.45 SD = 1.47 SD = 1.72 SD = 1.57
N=41 N=33 N=74 N=44 N=28 N=72

 

 

There were no main effects for gender (F [1, 138] = 2.915, p = .090, n2 = .021),
MRT score (F [1, 138] = .041,p = .841, 112 = .000), or dimension (F [1, 138] = 1.180,

p = .279, n2 = .008). A two-way interaction effects was found: dimension * gender (F

[1, 138] = 7.368, p = .007, n2 = .051). Males generally enjoyed 3D game more than
females while females enjoyed 2D game more than males. Other two—way
interactions, dimension * MRT (F [1, 138] = .211, p = .647, n2 = .002) and gender *
MRT (F [1, 138] = .052, p = .820, n2 = .000), were not significant. Three way
interaction effect of dimension * gender * MRT was also not significant, F (1, 138)
= .809, p = .370, n2 = .006. Therefore, the hypothesis 3, which is that high MRT

females enjoy 3D game more than low MRT males, was not supported.

Conclusion

This study examined the contribution of cognitive skill and gender in game
performance and game enjoyment. Specifically, this study hypothesized and tested
that cognitive ability is a stronger predictor of game performance and liking than
gender, or at least, individual differences in cognitive processing may confound the
relationship we see between sex and genre preference (Sherry, 2004).

By comparing differences of performance and enjoyment in playing 3D game
and 2D game, this study set three hypotheses. First, the expectation that if gender
matters in video game play, males perform better and enjoy more video game play
than females was partially confirmed. With regard to performance, males, on average,
showed better performance than females. However, males’ enjoyment was less than
females’, although the effect size was small. Second hypothesis was that high MRT
people will perform better in both 2D and 3D games while low MRT people perform
better only in 2D game play. Therefore, the margin between high MRT people and

low MRT people in 3D game was expected wider in 3D game than 2D game. The

result, however, was not consistent with this hypothesis. There was no evidence
supporting that the margin between high MRT people and low MRT people in 3D
game is wider in 3D game than 2D game. Statistically non-significant interaction was
found between mental rotation ability and dimensionality on enjoyment. Third
hypothesis was to test the three-way interaction of cognitive skill, gender, and
dimensionality. No significant three-way interaction effect for mental rotation skill,
dimensionality, and gender was found. Only the two-way interaction of gender and
dimensionality was found both in performance and enjoyment.

In summary, main findings of this study were categorized largely into three. First,
gender is a significant factor affecting video game perfOrmance. Male gamers are
more likely to play better than girls. Second, gender and dimensionality of video
games are related. Males performed better than females, but the difference was more
notable in 3D game play than in 2D game play. In terms of enjoyment, males enjoy
3D game more while females enjoy 2D game more. Third, game performance and
enjoyment could not go together. In the study, males performed better in 2D game
than females but females enjoyed the 2D game more than males. The overall
correlation between enjoyment and performance was also non-significant, r(100) = -

.062, p = .452.

Discussion
This study was an early study in video game research to try to find the
interactions of human cognitive skill, gender, and dimensionality of video games by

using the same game in different dimensions. Though this study purported to find the

27

superior influence of cognitive skill on video game play, the result was not consistent
with the expectation. There would be several explanations for this finding. One of
possible explanations for it could be from difficulty imbalance between 2D game and
3D game. For example, the difficulty of the 3D game may have been too high for the
participants. Another explanation would be that 2D game play also requires mental
rotation skill, according to the result showing wider gap between low MRT players
and high MRT players in 2D game play.

This study raised several questions to be solved in further research in order to
improve knowledge of video game play.

Methodologically, this study has several errors. Although random assignment
was done for the experiment, there were unbalances between male subjects and
female subjects and between high MRT people and low MRT people. Due to the
limited availability of participants, data did not fit to the equal proportions of gender
and game type, which was originally designed. Measurement of mental rotation skill
should have been done before the experiment session that includes the actual game
play, for proportionate assignment to an experimental condition. Therefore, to ensure
equal or at least similar number of males and females participants who have similar
MRT scores in each cell, a pretest of MRT should have been done. However, this
experiment arranged to measure mental rotation skill and game play in one given
session. Therefore, this study did not have equal distribution of high and low MRT
people across gender, and it made a possible threat to the validity of the study. In
addition to gender imbalance in MRT, the distinction of high and low MRT groups

was also problematic. MRT is one of standardized tests, which assume normal

distribution of scores. The division by median (or mean) could confound the result
seriously, especially in between group comparison tests as done in this study. This
study divided two groups based on median of 8. The number of participants who have
7 and 8 in the test was 9 and 21, which was about 20% of all participants. Therefore,
it was possible that the distinction did not reflect high or low cognitive skill
appropriately.

On experiment design level, there are also many possible confounds to make this
study vulnerable. First, different game plays are sensitive to innate technical
differences, such as different size, perspective (1St person vs. 3"d person perspective),
structures of game map, and controller layout. Moreover, these technical differences
are likely to be related to how much realism the game has. Realism of the game is
frequently proposed as a critical component for video game play. An increasing
degree of realism in computer games would seem to aggravate the gender gap
(Hartmann & Klimmt, 2006). 2D game is missing some components that 3D game
has, like some ceiling backgrounds and stereo sounds. Due to the loss of one
dimension in 2D game, the variety of component of the 3D games should be much
higher than that of 2D games. That is, 2D game is more likely to be less realistic than
3D game because of 3rd person perspective, relatively smaller size of game characters,
and more simple controllers. Therefore, realism in video games is more likely to
influence performance and enjoyment of the games. Second, there are some
individual differences that are likely to covariate with video game play, although
those can be canceled out by perfect randomization. For example, vision ability can

be a critical variable for video game play. In this study, age and major did not show

29

any association with other variables. Importantly, previous game experience could be
a confounding variable. Game hours (per week) was related to game performance,
r(145) = .252, p = .008. Especially, game hour was associated with 3D game
performance (r[60] = .396, p = .003) more than 2D game performance (r[83] = .127,
p = .352). This bias can be bettered by a pretest.

In addition, there are several suggestions for further studies in video game. First,
video game plays need more than one cognitive skill at a time. For example, 3D first
person shooting game, like Rainbow six, require 3D mental rotation skill to find right
paths, shooting skill to point an enemy correctly, and memory ability to remember the
paths a player passed to find a right way. There is no clue for how much contribution
one of those skills has over the others and what the most important skill is for the
game. For better studies that focus on the role of cognitive skills in video game play,
measuring various cognitive abilities and matching them to the performance and
enjoyment in a specific video game is desirable, regardless of the stereotypical genre
classification of video games, like shooting, puzzle, and action. That is, video game
play is a multilevel communicative phenomenon (Lucas & Sherry, 2004), so cannot
be explained by single factor. Second, there are few studies that specify what good
performance in video game is. The scoring system programmed in most of available
video games is one of the convenient indexes for showing game performance. A large
number of video games provide the scoring system based on the number of enemy
killed or the time elapsed to complete a mission. However, it is also possible for some
players that how creatively or how effectively they complete a mission is a crucial

index for good performance. Unless the game is simple in terms of the variety of

game components, most video games are products of numerous game components.
The more complex a game is. the more indexes the game has for measuring the
performance.

Though we are quite familiar with video game uses, more profound parts of
video games are still veiled. What researchers have to do is not describing the video

game play. but finding what video game play is.

31

APPENDIX

Table 7. Statistics ofobjects in 2D and 3 D (.‘hex-Quest

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Level Object Name Name Version
type In the original Doom 1n the Chex 3D ZD
Trooper Commonus 27 27
Monsters Shotgun Guy Bipedicus
Imp Bipedicus with Armor 1 1
Items Shell, Bullet Zorcher rechargers 6 6
1 Health potion Water, Fruits 9 9
2 ways comers (L-R) 2 ways comers (L-R) 1 1 13
Stairs (U-D) Stairs (U-D) 2
Map 3 ways 3 ways 1 1
Acid Ponds Acid Ponds 0 0
Teleport Teleport O 0
Trooper Commonus 59 59
MONSICI'S Shotgun Guy Bipedicus 3 3
Imp Bipedicus with Armor 9 9
Items Shell, Bullet Zorcher rechargers 16 16
Health potion Water, Fruits 20 20
2 2 ways 2 ways 12 16
Stairs Stairs 4
3 ways 3 ways 3 4
Map 4 ways 4 ways 1 0
Acid Ponds Acid Ponds 2 2
Teleport Teleport 2 2
Trooper Commonus 56 56
MOHSICFS Shotgun Guy Bipedicus 3 3
Imp Bipedicus with Armor 10 10
Items Shell, Bullet Zorcher rechargers 19 19
Health potion Water, Fruits 10 10
3 2 ways 2 ways 1 1 17
Stairs Stairs 6
3 ways 3 ways 4 5
Map 4 ways 4 ways 1 0
Acid Ponds Acid Ponds 2 2
Teleport Teleport 1 1

 

BIBLIOGRAPHY

AAUW Educational Foundation Research. (2000). T ech-Savvy: Educating girls in the
new computer age. Washington, DC: AAUW.

Anderson, C. A. (2004). An update on the effects of playing violent video games.
Journal of Adolescence, 27, 113-122.

Blumberg, F. C., & Sokol, L. M. (2004). Boys' and girls' use of cognitive strategy
when learning to play video games. The Journal of General Psychology, 131,
151-158.

Bosco, A., Longoni, A. M., & Vecchi, T. (2004). Gender effects in spatial orientation:
Cognitive profiles and mental strategies. Applied Cognitive Psychology, 18,
519-532.

Brown, R. M., Hall, L. R., Holtzer, R., Brown, S. L., & Brown, N. L. (1997). Gender

and video game performance. Sex roles, 36, 793-812.

Caplan, P. J., MacPherson, G. M., & Tobin, P. (1985). Do sex-related differences in
spatial abilities exists? - A multilevel critique with new data. American

Psychologist, 40, 786-799.

Chu, K., Heeter, C ., Egidio, R., & Mishra, P. (2004). Girls and games literature
review. Retrieved September 10, 2006, from

httfl/gacgioneersmsu.edu/girls and games 1it_review.htm

Eastin, M. S. (2006). Video game violence and the female game player: self— and
opponent gender effects on presence and aggressive thoughts. Human

Communication Research, 32, 351-372.

Funk, J. B., & Buchman, D. D. (1996). Children's perceptions of gender differences

in social approval for playing electronic games. Sex roles, 35, 219-231.

33

Funk, J. B., & Buchman, D. D. (1996). Playing violent video and computer games

and adolescent self-concept. Journal of Communication, 46, 19-32.

Gao, Y. (2004). Appeal of online computer games: A user perspective. The Electronic
Library, 22, 74-78.

Hartmann, T., & Klimmt, C. (2006). Gender and computer games: Exploring females'
dislike. Journal of Computer-Mediated Communication, 11. Retrieved
September 10, 2006, from

http://icmc.indiana.edu/vov1 1/issue4/11artmannhtml

Huh, S. (2004). Ritualized uses of interpersonal communication media: focusing on
communication direction and perceived media attributes. Master’s thesis,

Yonsei University, Seoul, Korea.

Jones, C. M., Braithwaite, V. A., & Healy, S. D. (2003). The evolution of sex

differences in spatial ability. Behavioral Neuroscience, 117, 403-411.

Kim, K, Park, J, & Kim, D. (2002). E-life style and motives to use online games. Irish
Marketing Review, .15, 71-77.

Lachlan, K. A., Smith, S. L., & Tamborini, R. (2005). Models for aggressive
behavior: The attitude of violent characters in popular video games.

Communication Studies, 56, 313-329.

Levine, S. C., Huttenlocher, J ., Taylor, A., & Langrock, A. (1999). Early sex
differences in spatial skill. Developmental Psychology, 35, 940-949.

Lucas, K., & Sherry, J. L. (2004). Sex differences in video game play: A

communication-based explanation. Communication Research, 31, 499-523.

Windahl, S. & McQuail, D. (1993). Communication Models: For the Study of Mass

Communication, 2nd ed. Pearson Education.

34

Quaiser-Pohl, C., Geiser, C., & Lehmann, W. (2006). The relationship between
computer-game preference, gender, and mental-rotation ability. Personality
and Individual Differences, 40, 609-619.

Rahman, Q., & Wilson, G. D. (2003). Large sexual-orientation—related differences in
performance on mental rotation and judgment of line orientation tasks.

Neuropsychology. 17, 25-31.

Sherry, J. L. (2004). Flow and media enjoyment. Communication Theory, 14, 328-
347.

Sherry, J. L., Rosaen, S., Bowman, N.D., & Huh, S. (2006). Cognitive skill predicts
game ability. Paper presented at the 91th annual conference of the

International Communication Association, Dresden, Germany.

Smith, S. L., Lachlan, K. A., & Tamborini, R. (2003). Popular video games:
Quantifying the presentation of violence and its context. Journal of

Broadcasting & Electronic Media, 4 7, 58-76.

Terlecki, M. S., & Newcombe, N. S. (2005). How important is the digital divide? The
relation of computer and videogame usage to gender differences in mental

rotation ability. Sex roles, 53, 433-441.

Voyer, D., Voyer, S., & Bryden, M. P. (1995). Magnitude of sex differences in spatial
abilities: A meta-analysis and consideration of critical variables.

Psychological Bulletin. 117, 250-270.

Waller, D. (2000). Individual difference in spatial learning from computer-simulated

environments. Journal of Experimental Psychology: Applied, 6. 307-321.

Wood, R. T. A., Griffiths, M. D., Chappell, D., & Davies. M. N. O. (2004). The
structural characteristics of video games: A psycho-structural analysis.

CyberPsychology & Behavior, 7, 1-10.

35

   

lljllljlljlijlliljl