 

.‘ii

" I. f 9"

,1"! L.)
K G) 4 J 7 o J i
5‘ ‘- I .' “'

LIBRARY
Michigan State
University

This is to certify that the
thesis entitled

KEEPING ABREAST OF HYPERSEXUALITY:
A VIDEO GAME CHARACTER CONTENT ANALYSIS

presented by

EDWARD PAUL DOWNS

has been accepted towards fulﬁllment
of the requirements for the

MA degree in Communication

 

Major Professor's Signature

06¢ 17 2007

Date

MSU is an Afﬁrmative ActiorVEqual Opportunity Institution

 

 

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

 

MAR 0 7 799

 

QSO‘ZQZ

 

 

 

 

 

 

 

 

 

 

 

 

 

2/05 cEiRmJndd-QIS

 

KEEPING ABREAST OF HYPERSEXUALIT Y:
A VIDEO GAME CHARACTER CONTENT ANALYSIS

By

Edward Paul Downs

A THESIS

Submitted to
Michigan State University
in partial fulﬁllment of the requirements
for the degree of

MASTER OF ARTS
Department of Communication

2004

ABSTRACT

KEEPING ABREAST OF HYPERSEXUALITY:
A VIDEO GAME CHARACTER CONTENT ANALYSIS

By

Edward Paul Downs

This study examined hypersexuality in video game characters. Sixty of the top

selling video games across three of the most popular consoles (Microsoft Xbox®,

Sony PlayStation®2, and Nintendo GameCubem) were content analyzed. A total of

489 characters with an identiﬁable sex were used for analyses. Female characters
(n=70) were underrepresented in comparison to their male counterparts (n=4l9). In
comparison to male characters, females were signiﬁcantly more likely to be partially
nude, featured with an unrealistic body image and shown wearing sexually revealing
clothing and inappropriate attire. Implications for these ﬁndings are discussed using

Social Cognitive Theory as a theoretical anchor.

ACKNOWLEDGMENTS

This study was made possible by a grant from the Annenberg School of
Communication at the University of Southern California.

I would like to thank Dr. Charles Atkin and Dr. Bradley Greenberg for their
helpful comments and suggestions. I would also like to thank Dr. StacySmith for her
guidance, her inspiration and for her ability to extract some of my best work. It has
been an honour and a pleasure. Thank you, also, to Kate, Elaine, and Marc. Without
their insight and their tireless coding, this thesis would still be a work in progress.

To my family and friends who supported me through this journey, thank you,
for always being there. Slainte! And to Rachel, for her unwavering strength and
support: thank you, for tending to all the bumps and bruises that I suffered through

grad school. I don’t know how I would have made it without you.

iii

TABLE OF CONTENTS

LIST OF TABLES .............................................................................. vi
CHAPTER 1: BACKGROUND ............................................................ 1
Attention .................................................................................. 4
Retention .................................................................................. 4
Production ............................................................................... 5
Motivation ............................................................................... 5
Video Game Content ............................................................................ 6
The Present Study ................................................................................ 8
CHAPTER 2: METHOD .................................................................... 10
Sample ............................................................ h ....................... 10
Unit of Analysis ........................................................................ 10
Demographic variables ............................................................... 11
Attributes of hypersexuality .......................................................... 12
Reinforcements for sexuality ......................................................... 14
Coder Training and Reliability ...................................................... 14
CHAPTER 3: RESULTS .................................................................... 15
Analysis Plan ........................................................................... 15
Descriptive Statistics .................................................................. 16
Hypothesis One ........................................................................ 17
Hypothesis Two ........................................................................ 17
Research Question One ............................................................... 19
Research Question Two ............................................................... 20
CHAPTER 4: DISCUSSION .............................................................. 21

iv

LIST OF TABLES

Tablel: Top Twenty Video Games by Game Platform Included in Analyses ......... 27
Table 2: Frequency of Characters by ESRB Rating, Platform, and Sex ................. 30

Table 3: Percentage of Characters Showing Hypersexuality Indicators by ESRB
Rating and Sex ................................................................................... 31

Table 4: Frequency and Percentage of Characters Showing Hypersexuality Indicators
by ESRB Rating ................................................................................. 32

Table 5: Percentage of Characters Showing Hypersexuality Indicators by
Platform .......................................................................................... 33

Table 6: Percentage of Characters Showing Hypersexuality Indicators by Platform
and Sex ............................................................................................ 34

vi

Chapter 1: Background

Recently, video games have been put under the media microscope. Empirical
studies focusing on violence dominate the research literature (for review, see Dill &
Dill, 1998). However, as violent characters seemingly became standard fare in many
videogames (Provenzo, 1991), sexual ones also have found a market in this medium.
Lara Croft, from Tomb Raider and Helena, from the Dead or Alive series are
examples of unrealistically proportioned female characters that ﬁght and play in skin-
tight, revealing outﬁts. Advocate group Children Now referred to these voluptuous,
disproportionately sized characters as “hypersexualized” (Children Now, 2001, pg.
15).

Such sexuality is starting to concern some parents and legislators. The Protect
Children ﬁom Video Game Sex and Violence Act of 2003 (Baca, 2003) proposes to
penalize those who sell video games containing nudity, sexual content, and sexual
violence. As Kari Peters, a parent from Sammamish, WA comments, “Occasionally, I
don’t mind adult themes, as they bring up wonderful discussions if the child is
mature, but the blatant sex and ﬁlthy language doesn't show the best parts of
adulthood, I feel” (Peters, 2004, p. E2).

Concerns from parents and others led to the development of the Entertainment
Software Review Board (ESRB) in 1994. The ESRB rates video games for the
appropriateness of different types of content for varying age groups. Sexual content is
one feature that the ESRB regularly monitors. They focus on assessing mature humor,
nudity, partial nudity, sexual themes, sexual violence, strong sexual content, and

suggestive themes. The majority of these categories are character focused, revolving

around different levels of nudity and references to sex and sexuality. Clearly,
sexuality seems to be a staple of video game content.

The use of sexual content may be a strategic attempt to lure the most avid
game player, an adult male (Entertainment Software Association, 2004). Yet video
games are growing increasingly popular with younger audiences who are just starting
to develop ideas about sexuality. Indeed, the results from a national study funded by
the Kaiser Family Foundation found that 16% of 4- to 6-year olds play video games
daily. Typically, children spend an hour per day playing, with girls devoting an
average of 53 minutes and boys 68 minutes (Kaiser Family Foundation, 2002).
Another nationwide survey by the Kaiser Family Foundation (1999) found that 8- to
18-year olds play video games 27 minutes a day, with boys spending considerably
more time gaming than girls (41 minutes vs. 12 minutes, respectively). Thus, children
may likely be exposed to many sexualized portrayals very early in their socio-
emotional development.

What impact does seeing such extreme portrayals of sexuality in video games
have on youth? To date, there is little research in this area. There is a growing body
of literature on the inﬂuence of television on sexual socialization and stereotyping.
Studies have shown that children begin to develop attitudes toward body types
between the ages of six and seven years, and as children get older, these attitudes
become stronger (for review, see Spitzer, Henderson, & Zivian, 1999). Differences
between female body sizes in the media and actual body sizes have been linked to
eating disorders (Field, Cheung & Wolf, 1999), low self-esteem (Polce—Lynch,

Myers, Kliewer & Kilmartin, 2001), body dissatisfaction (Harrison & Cantor, 1997),

feelings of objectiﬁcation (F rederickson & Roberts, 1997), and to the attitude that
women’s bodies are projects that need to be worked on (Brumberg, 1997, of.
Mumen, Smolnak, Mills & Good, 2003). With effects such as these it should come as
no surprise that one study noticed an inverse relationship between playing
video/computer games and girls’ self-esteem scores (Funk & Buchman, 1996).

Females are not the only ones affected by unrealistic body images in the
media, however. Media portrayals of muscle-bound men may also set standards for
male gamers. Such depictions may contribute to body dissatisfaction among some
men, as well as such risky behaviors as steroid use (Labre, 2002). Sexualized
portrayals of women in the media have been linked to objectiﬁcation, leading some
males to view women and their bodies as possessions (F rederickson & Roberts,
1997). Adolescent males learning to objectify women may lead to negative social,
legal, and relational damage. Exposure to sexually explicit material such as
pornography has been associated with poor evaluations of sexual partners and
increased sexual aggression (J ansrna, Linz, Mulac & Imn'ch, 1997).

Since hypersexualized video game characters may encourage or reinforce
these negative effects, it is necessary to understand how this may occur. Social
Cognitive Theory (SCT, Bandura, 2002) provides a theoretical framework to
understand how children and adolescents can learn attitudes and beliefs from
exposure to video game characters. Bandura’s (2002) SCT indicates that we can
intentionally and unintentionally learn ﬁ'om the models that we encounter in our

environment. Bandura (2002) proposes that the processes of attention, retention,

production, and motivation are functions that govern our ability to learn through
observation.

Attention. Attentional processes are those that determine which information
and observed actions are extracted from a modeled, learning environment.
Attractiveness is one variable that affects attentional processing. If the game
character is perceived as attractive, literature suggests that observers are more likely
to ﬁnd them socially acceptable and inﬂuential (Bandura, 2002). A recent quote from
a game reviewer suggests: “There’s no arguing that sex sells. And with the typical
gamer a young adult male, you can't really blame the industry for catering to their
target audience” (Morris, 2004, n.p.). Overtly sexual, provocatively dressed, or
partially naked male or female characters may be very attractive, attention-getting
attributes of video game play.

Retention. Getting a gamer’s attention is not enough to invoke learning.
Without the ability to symbolically code and retain information, observational
learning would not be possible. Retention processes involve restructuring modeled
events into rules for memory (Bandura, 2002). In video games, a player has the
ability to witness how characters present themselves and interact with other
characters over and over again. Recalling that 8- to 18 year-old males typically play
video games for 41 minutes per day (Kaiser Family Foundation, 2002), repeated
exposure through extended play should make it very easy for a video game player to
create, rehearse, and strengthen symbolic representations of events. In fact, gender
schema theory suggests (Calvert & Huston, 1987) and research reveals that exposure

to stereotyped, sexual interactions on television can provoke schema that categorize

women as sexual, submissive, and less intelligent than their male counterparts
(Thompson & Pleck, 1986). Repeated exposure to attractive, scantly clad female
characters may be problematic for young male gamers’ sexual development.

Production. Production processes “guide the construction and execution of
behavior patterns” (Bandura, 2002, p. 129) which can then be compared to the
symbolic representations stored in memory for congruity. In using a symbolic
representation of a modeled act, a person can adapt it for use in a variety of different
situations. Not many children or adolescent players are likely to be as attractive,
muscular, or sexy as the characters they control and play in video games. However,
even though the physical attributes may not be accessible to the average player,
verbal and attitudinal attributes may be replicated more easily. Gamers could learn
attitudes and beliefs about sex roles and gender stereotyping and apply them to the
real world situations they ﬁnd themselves in every day. For example, if a video game
player notices that males in video games are active, problem-solving heroes and
women are incapable, and always needing help from men, .sex-based stereotypes may
manifest themselves in the home, school, or other social circles.

Motivation. Motivational processes revolve around a person’s ability to create,
change and modify their behavior within a social structure. In the context of SCT, if
attitudes are shared by a peer group or reinforced, the schemata are likely to be
strengthened. Studies generally support this rationale in the violence domain.
Children and adults who are exposed to models that are rewarded or not punished for
their violent actions are more likely to be aggressive immediately after viewing

(Bandura, 1965; Lando & Donnerstein, 197 8; Paik & Comstock, 1994). Video games

certainly contain reward mechanisms. Increases in scores, extra lives, or access to
passwords or codes can serve as agents of reinforcement. In the game BMXm
players that successfully accomplish their missions are rewarded with codes to view
real video clips of dancers performing a strip-tease.

To summarize, video game players through repeated exposure to attractive,
hypersexualized characters may begin to develop scripts about gender stereotypes.
When hypersexuality is used as a reward, players may assume that this type of
activity is reinforced and sanctioned by society. Because of this, it is important to
examine sexuality in video games.

Video Game Content

Several studies have examined stereotypical portrayals in video game content.
In one of the earliest studies, Dietz (1998) examined gender stereotyping in 33 of the
most popular Nintendo and Sega Genesis video games. The results revealed that
almost half (41%) of the games were devoid of female characters. When human
women were portrayed, 21% of the time they were depicted as “darnsels in distress”
or as “visions of beauty” (Dietz, 1998, p. 434). Irnportantly, 28% of the games coded
with human characters showed women as sex objects.

A few years later, Children Now (2001) assessed 70 games for violence,
gender, and racial diversity. The ten top selling games across 6 types of platforms
(PCs, Sega Drearncast, Sony PlayStation, PlayStationZ, Nintendo 64, Game Boy
Advance, and Game Boy Color) were unitized at both the game and character levels.
The ﬁndings revealed that of the 874 characters coded, 73% were male and 12% were

female. Video game players had more opportunities to select playing a non-human

character such as a robot or anthropomorphized being such as a hedgehog (i.e., Sonic
the Hedgehog) than a female character. This analysis also found that women were
oﬁen hypersexualized. Nearly 20% of the female characters in the sample were
shown with unrealistic body sizes. In terms of clothing, 21% of female characters
bared their breasts, 13% brandished their buttocks, and 20% modeled their midriffs.
Finally, female characters were twice as likely as males to be coded as wearing
seductive attire (Children Now, 2001).

Similar to Children Now’s research (2001), Beasley and Standley (2002)
examined clothing as an indicator of gender stereotyping in games rated E (i.e., for
everyone), T (i.e., for teen audiences, 13 and older), and M (i.e., for mature
audiences, 17 and older). Forty-seven Nintendo 64 and Sony PlayStation games were
assessed. Consistent with Children Now’s ﬁndings (2001), their results revealed that
out of almost 600 characters, less than 14% were women. They also demonstrated
that female characters wore clothing that exposed more skin than did male characters.
Out of all female portrayals, nearly 41% were coded as “voluptuous.” Also
noteworthy was of those 41 % of voluptuous women, 31% were featured in video
games with an ESRB rating of “E” for everyone (Beasley & Standley, 2002).

Most recently, Haninger and Thompson (2004) assessed 81 “T” rated games
for Nintendo 64, Sega Dreamcast, Sony PlayStation, and Sony PlayStationZ consoles.
The results revealed that 27% of their 81 game sample depicted sexual themes and
profanity. They also noted that their sample was signiﬁcantly more likely to depict

women as compared to men as partially nude and/or engaged in sexual behaviors.

To summarize, previous content analyses demonstrate that women are far less
likely to appear in video games than are male characters. When women are present,
they are oﬁen depicted as helpless and in need of saving. Women are also more likely
than men to be portrayed as sex objects, with unrealistic body proportions and
considerable amounts of exposed skin.

Although the previous content analyses have provided insight as to what is
being portrayed in video games, they are not without limitation. The primary
problem across all of the previous research is that it is simply outdated. The samples
feature games played on antiquated platforms such as Nintendo 64, Sega Drearncast,
and Sony PlayStation. Faster graphic engines and more realistic looking characters
have evolved, which calls for more recent research on games played on consoles such
as Microsoft Xbox, Nintendo GameCube, and Sony PlayStation 2.

Perhaps the most information about hypersexual characters comes from
Beasley and Standley’s study (2002). However, these researchers coded attributes of
all characters, rather than those that are primary or secondary to game play. As these
scholars point out, “in future studies, characters should be coded as to their
importance in the video game” (Beasley & Standley, 2002, p. 290). It is important to

look at primary and secondary characters, as they are the ones that probably receive

-_ -mH, ...— -

 

the most attention from those playing the games.
The Present Study

The purpose of this study is to analyze primary and secondary characters in
the most popular console games for Nintendo GameCube, Sony Playstation 2, and

Microsoft Xbox. Previous media research has demonstrated pronounced gender

differences in portrayals (Beasley & Standley, 2002; Children Now, 2001; Dietz,
1998). Women are not only presented less frequently than men, but when they are
featured they are more likely to be shown in a hypersexualized (i.e., seductive
clothing, nudity, unrealistic body proportions) light. Therefore, the ﬁrst two
hypotheses are advanced:

H1: Female video game characters will appear less frequently than will male
characters.

H2: Female video game characters will appear more hypersexualized than
male characters.

I was also interested in how ERSB rating may inﬂuence the presentation of
hypersexual portrayals. According to the ESRB rating system, E games (i.e., for
everyone) should not contain any of the variables associated with hypersexuality,
including partial or full nudity, references to sex, mature humor or sexual content. T
games (i.e., for teens) should contain only suggestive themes and M games (i.e., for
mature players) may contain any or all of the variables associated with
hypersexuality, excluding graphic depictions of sex. Yet previous research suggests
that ESRB does not always rate objectionable content consistently (Haninger &
Thompson, 2004). For example, Thompson and Haninger (2001) found that 14 E-
games featuring violence in their sample did not receive a content label from the
ESRB. To assess the dependability of the ESRB’s rating system with regards to
hypersexuality, the ﬁrst research question asks:

RQ1: Does game rating inﬂuence hypersexualized portrayals in video games?

In addition to ESRB rating, another feature that may inﬂuence hypersexuality

is the console. Web site reviews about game consoles do not identify any one console

as providing more racy content than another, although all three manufacturers in this
study (Nintendo, Sony, and Microsoft) provide game titles containing sexual content
(Loftus, 2004; Parents Television Council, 2004). Given this lack of information
empirically or in the popular press, the second research question asks:

RQZ: Does console inﬂuence hypersexualized portrayals in video games?

Method

Sample

A total of 60 top-rated games in Q2 of 2003, for the Nintendo GameCube
(n=20), Sony PlayStation 2 (n=20), and Microsoft Xbox (n=20) were analyzed (a
complete list of games can be found in Table 1). Within each platform, the sample
included the top-selling games based on available sales ﬁgures from the NPD Group
for 2003. The sample of games included 22 E-rated games, 23 T-rated games, and 15
M-rated games. Three undergraduates with considerable gaming experience played
each game for 20-minutes. Game play was captured on VHS tapes via a
videocassette recorder. Coders then coded the tapes containing these 20-minute
segments for various character attributes.
Unit of Analysis

There are two units of analysis: game segment and character. Game segment
consists of the 20-minute pre-recorded sample of each video game. The variables
coded at the game level are ESRB rating (e.g., E, T, M) and the brand of console
(e. g., Microsoft Xbox, Nintendo GameCube, and Sony PlayStation 2).

Characters are deﬁned as human, animal, robotic, or anthropomorphized

entities that are involved in game play. Only attributes of primary and secondary

10

characters will be assessed. Primary characters are those that a video game player
actively manipulates and controls onscreen. In almost all cases, the primary character
is the protagonist or the person around which the plot and successful completion or
failure of the game’s objective centers. Secondary characters are those that are
immediately tied to or related to the primary character (i.e., they aid or deter the
primary character from fulﬁlling their quest). Based on the review of literature and
social cognitive theory outlined above, three types of character measures will be
analyzed: demographic, hypersexual qualities, and reinforcements for sexuality. The
following variable deﬁnitions were developed for the present study (unless otherwise
referenced).

Demographic variables. Five different variables were coded for character
demographics.l The ﬁrst demographic variable is species. Species are comprised of
related organisms or populations potentially capable of interbreeding. Species will be
coded as human, animal, robot, supematural creature, anthropomorphized animal,
anthropomorphized supematural creature, or cannot tell. The second demographic
quality is biological sex. It will be coded as male, female, no sex, or cannot tell. The
third demographic quality is ethnicity. Ethnicity refers to large groups of people that
are categorized together, sharing one of more of the following qualities such as racial,
tribal, religious, linguistic, or cultural origins. Ethnicity will be coded as Caucasian,
Aﬁ'ican, Hispanic, Asian/Paciﬁc Islander, Native American, other, not applicable, or
cannot tell. The fourth variable is age. Since the actual ages of video game characters
are rarely provided, character ages in approximate years will be assessed. The age

groups were broken down such that character cues and associated age stereotypes (ex.

11

grey hair and use of a staff or cane to walk would be associated with an elderly
character) should fall neatly into mutually exclusive categories. Other studies have
used a roughly similar age break down and have achieved high levels of reliability
(Smith et. al. 1998 and Wilson et a1, 1997). Character ages are divided into child (0-
12 years), teen (13-19), adult (20—64), elderly (65+), not applicable, or cannot tell.

The ﬁfth variable is resolution. Resolution refers to the image quality (i.e.,
sharpness, clarity) of the onscreen graphics and is coded low (i.e., pixilated graphics
with irregular boundaries, choppy motion sequencing, little or no shadowing),
medium (i.e., fairly seamless motion sequencing, some shadowing), and high (i.e., 3-
D animation, character shadowing, ﬂuid motion).

Attributes of hypersexuality. Hypersexuality is deﬁned as excessive concern
with, or indulgence in, a sexualized appearance or activity. Eight different variables
assess aspects of hypersexuality.2 The ﬁrst is sexually revealing clothing, which
refers to any garment that is worn in order to enhance, exaggerate, call attention to, or
accentuate the curves or angles of any part of the body (ﬁ‘om the neck to above the
knee) and which by design, would arouse interest of physical intimacy ﬁ'om others.
The presence of sexually revealing clothing will be coded as yes, no, not applicable,
or cannot tell. The second variable is nudity. Nudity refers to the amount of exposed
skin a character shows. The levels are none (i.e., body is covered in such a way that
there is no visible skin ﬁ'om the knees to the neck line about the shoulders as by an
unaltered crew-neck, t-shirt, or tank top), partial (i.e., exposed midriffs for women,
shirtless males, exposed cleavage, thighs, and/or buttocks), full (i.e., human

character’s body does not appear to be covered by any clothing, save shoes or a hat,

12

or if a covering of any sort as by a towel or such is transparent in nature), not
applicable, or cannot tell.

The third variable is character body proportion.3 Character body proportion is
deﬁned as the harmonious relation of parts to each other or to the whole as it applies
to a human body. This variable will be gauged by the propensity of characters to
resemble an average male or female human. Therefore, by deﬁnition, any woman or
man with an overly exaggerated chest-size, extremely long arms or legs, and/or a
disproportionately small waist would be coded as having an unrealistic image. The
codes for this variable will be realistic, unrealistic, not applicable, or cannot tell.

The fourth variable is sex talk. Consistent with previous research (Kunkel &
Cope, 1999), sex talk refers to any verbal reference or dialogue regarding sex or
sexual issues. For each character, sex talk will be coded as present or absent. The
ﬁfth variable is sexual behavior and is coded independent from sex talk. Based on
the Kaiser Family Foundation’s (2003) deﬁnition, sexual behavior is deﬁned as a
character’s actions that imply a sense of likely sexual intimacy. Sexual behavior will
be coded as yes or no.

The sixth variable is appropriateness of attire. Appropriateness of attire is
deﬁned as those garments worn by a character that are suitable and functional in
consideration of the task at hand. For example, if a woman beside a pool wears a
swimsuit, this would be considered appropriate. If the same woman wears a swimsuit
while competing in a dirt bike race, then the swimsuit would be considered
inappropriate. This variable will be coded as appropriate, inappropriate, not

applicable, and cannot tell.

l3

The seventh variable is breast size. For female characters only, breast size
will be coded as ﬂat, average, voluptuous (Beasley & Standley, 2002) or carmot tell.
The eighth variable is waist size. This code will examine the waist-line of characters
in proportion to the rest of the body. The codes for this variable will be
disproportionately small, average, disproportionately large, or cannot tell.

Reinforcements for sexuality. Consistent with a social cognitive approach, the
last two variables tap reinforcements for sexuality. Verbal reward for sexuality is
deﬁned as any verbal praise (e. g., “Hey, good looking”,” “Nice legs”) that is given to a
character as a result of their own sexuality. The codes for this variable will be present
or absent. In addition, nonverbal reward for sexuality will be coded. This variable is
deﬁned as any nonverbal offering (e.g., suggestive licking of lips, winking, blowing a
kiss) that is given to a character as a result of their own sexuality. This variable is coded
present or absent.

Coder Training and Reliability

A total of 3 undergraduate coders from a large west-coast university coded all
of the 20-minute, pre-recorded video game segments. The coders received weeks of
extensive classroom and lab training prior to coding the sampled games. Some
training sessions were conducted in person with coders and trainers in the same room,
and other sessions were conducted via phone with PowerPoint presentations when
both authors could not be physically present. Scott’s Pi (it) with the Potter and
Levine-Donnerstein (1999) correction formula was used to calculate coder reliability.
Coders did not begin assessing game segments until they reached a reliability of .80

or better for every code. The following are the Scott’s Pi estimates for each variable:

14

character type (1.0), point of view (1.0), resolution (1.0), sex (.99), species (1.0),
ethnicity (.88), age (.95), sexually revealing clothing (.87), nudity (.90), body
proportion (.82), sex talk (.99), sexual behavior (1.0), appropriate attire (.90), breast
size (.98), waist size (.88), verbal sex as a reward for character (1.0), and nonverbal
sex as a reward for the character (1.0).
Results

Analysis Plan

Prior to the execution of the analyses, hypersexuality had to be deﬁned. From
the most liberal standpoint, any one of the eight variables outlined above (i.e.,
sexually revealing clothing, appropriate attire, nudity, breast size, waist size, body
proportion, sex talk, sex behavior) could constitute a hypersexual portrayal. Because
of this, each of these attributes will be examined in isolation to test the relevant
hypotheses and answer the research questions. To this end, chi—square analyses were
executed on each hypersexual variable blocking by gender, rating, or console. Two
criteria were used to determine the signiﬁcance of any ﬁnding. First, only chi-
squares that were statistically signiﬁcant at the p < .05 level were reported. Second,
only those percentages that differ by 10% or more were delineated. In this way, the
ﬁndings will reﬂect both statistical and practical differences in the patterns of
hypersexuality in video games. The sample size for some analyses may vary, as
items coded as “cannot tell” or “not applicable” will be excluded. Below, descriptive
statistics are provided across the entire sample followed by speciﬁc tests of the

hypotheses and research questions.

15

Descriptive Statistics

Of the top 20 games played on each platform (Microsoft Xbox, Nintendo
GameCube, and Sony PlayStation 2) coders distinguished 576 different primary and
secondary characters for analysis. 4 Since we are interested in portrayals of sexuality,
only those characters that could be identiﬁed as male or female remained in the
sample. As such, a total of 87 characters had to be excluded from analysis. Of the
remaining 489 characters, 40% of characters appeared in E-rated games (n=196),
26% appeared in T-rated games (n=128), and 34% appeared in M-rated games
(n=165). Most of the characters were classiﬁed as secondary (76%, n= 371) and just
under a quarter (24%, n=118) were primary. Almost all primary characters were
played in the third person (99%) perspective and almost all characters (99%, n=486)
were viewed in high resolution.

In terms of ethnicity, half of the 489 characters were Caucasian (50%, n=246),
followed by Aﬁ'ican (21%, n=101), Asian/Paciﬁc Islander (7%, n=36), and Hispanic
(3%, n=14). A total of 19% (n=92) of the characters had an unidentiﬁable ethnicity or
were coded as not applicable.

Looking at ethnicity by sex, 465 characters had available codes for both
variables. These characters break down into the following groups: Out of 246
Caucasian characters, (13%, n=32) were female. Out of the 101 African characters,
(9%, n=9) were female. For the 14 Hispanic characters, (14%, n=2) were female. For
the 36 Asian/Paciﬁc Islanders, (25%, n=9) were female. The other three categories,
other, cannot tell and mixed, combined for 68 instances with (22%, n=15) being

female.

16

Almost all of the characters were adults (94%, elderly = 2%, children =l%,
unidentiﬁable = 3%). Most characters in the sample were human (88%), followed by
anthropomorphized characters (10%), supernatural creatures (1.6%), robots (.2%),
and animals (.2%).

Hypothesis One

It was expected in Hypothesis 1 that female characters would appear less
frequently than would male characters. A one-way chi-square yielded a signiﬁcant
effect for character sex, X2 (1, N = 489) = 249.08, p < .01. As shown in Table 2, a
substantially higher proportion of male characters (86%, n=419) than female
characters (14%, n=70) were featured in the games. A similar pattern held when
examining sex of primary characters, X2 (l, N = 118) = 68.60, p < .01. Substantially
more males (88%) were primary characters than were females (12%). A one-way
chi-square also revealed a signiﬁcant effect for secondary characters, X2 (1, N = 371)
= 180.80, p < .01. Secondary characters were more likely to be males (85%) than
females (15%). Thus, the data are highly consistent with Hypothesis 1.

Hypothesis Two

Hypothesis 2 predicted that female characters would be hypersexualized more
often than would male characters. To test this hypothesis, each hypersexuality
variable relevant to both sexes (sexually revealing clothing, appropriate attire, nudity,
body image, waist size, sex talk, sex behavior) was assessed by biological sex. All
primary and secondary characters coded as male or female were included in the
analyses. In terms of sexually revealing clothing, a signiﬁcant chi-square was

observed by sex, X2 (l, N = 469) = 39.70, p<.001, (I) = .29. A higher proportion of

17

female characters (41%) than male characters (11%) were portrayed in sexually
revealing clothing (see Table 3).

The impact of character sex on nudity also was explored. Prior to executing
the analysis, the nudity variable was collapsed into two levels: no nudity (none) vs.
some nudity (partial, full). A chi-square analysis revealed a statistically signiﬁcant
effect, X2 (1, N = 468) = 99.88, p<.001, (D = .46. When compared to male characters
(4%), a substantially higher proportion of female characters (43%) were depicted
nude (partial or full). See Table 3.

Next, body type was examined. This variable was collapsed into two
categories: realistic vs. unrealistic body image. A chi-square revealed a signiﬁcant
difference in body type by sex, X2 ( 1, N = 464) = 67.15, p<.001, (I) = .38. Female
characters were more likely to be shown with an unrealistic body type (25%) than
were male characters (2%).

The waist size of characters also was examined. Due to an interest in
uncharacteristically small waists, this variable was collapsed into two categories:
small waist vs. not small waist (average, large). A chi-square analysis revealed a
signiﬁcant difference by sex, X2(1, N = 461) = 140.88, p<.001, (I) = .55. When
compared to males (1%), female characters were substantially more likely to be
displayed with small waists (40%).

The appropriateness of the character’s attire also was examined. A signiﬁcant
chi-square was observed for this variable by sex, X2 (1, N = 472) = 36.94, p<.001, (I)

= .28. As noted in Table 3, a higher proportion of female characters (16%) than male

18

characters (2%) were shown wearing clothing that would not be appropriate for
completing the task at hand.

The last two hypersexuality variables, talk about sex and sexual behavior,
occurred infrequently in the data set. Across the entire sample of characters, only 1
instance of sex talk and 5 instances of sexual behavior were observed. Due to this
low observed n, chi-square analyses could not be executed. The verbal and nonverbal
reinforcements delivered to characters for their sexuality were also assessed.
However, no instances of these variables were observed across all of the characters in
the sample.

Overall, differences were observed across ﬁve of the seven variables
measuring attributes of hypersexuality that applied to both males and females. This
pattern of results seems to be consistent with Hypothesis 2.

Research Question One

Research question one asked if game rating inﬂuenced hypersexualized
portrayals in video games. These results are displayed in Table 4. Across the
hypersexuality variables, (sexually revealing clothing, appropriate attire, nudity, body
image, waist size) only one analysis including males and females in the sample met
the criteria for statistical (p<.05) and practical signiﬁcance (10% difference). A
signiﬁcant chi-square was observed for nudity by ESRB rating, X2 (2, N = 468) =
19.52, p<.001, V* = .20. Characters were more likely to appear nude in games rated T
(15%) or M (14%) than were those in games rated E (2%).

Looking at the distribution of hypersexuality variables by ESRB within sex, a

slightly different picture emerges. Because the analyses would have many cells with

19

sexually revealing clothing in Nintendo GameCube games (31%) than in Microsoft Xbox
(12%) or Sony PlayStation 2 games (10%).

The distribution of hypersexuality variables by console within sex presents a
slightly different set of results. Again, because a majority of the cells contain less than
ﬁve observations, the pattern of results is described. For women (see Table 6), Microsoft
Xbox has the highest frequency of unrealistic proportions, sexually revealing clothing,
partial nudity and voluptuous breasts. Also noteworthy is that Nintendo GameCube' is
most likely to show men wearing sexually revealing clothing.

DisCus 'on

The purpose of this study was to examine ypersexuality in video game

  
 
    
 
 

  
 
  
 

. Overall, the results show marked differe ces in the way males and females
are presen in video games both in terms of sheer am unt of appearance in games as
well as the y in which their sexuality is represented. In hort, females occur far less
frequently in v eo games and are more likely to be shbwn ' a hypersexual light than are
males.
Hypothesis received empirical support for the predictio that female characters
would appear less fr uently than would male characters (14%, n= , 86%, n=419,
Standley’s (2002) ﬁndings\ Of the characters in their sample, 14% were emale and 72%
were male. The results are a o congruent with the Children Now (2001) s dy, which-

found that 73% of video game aracters are male and 12% are female. Whil these

numbers are not too terribly diff t from gender representation on television ( erbner,

20

Discussion

The purpose of this study was to examine hypersexuality in video game
characters. Overall, the results show marked differences in the way males and
females are presented in video games both in terms of sheer amount of appearance in
games as well as the way in which their sexuality is represented. In short, females
occur far less frequently in video games and are more likely to be shown in a
hypersexual light than are males.

Hypothesis 1 received empirical support for the prediction that female
characters would appear less ﬁequently than would male characters (14%, n=70,
86%, n=419, respectively). The proportion of females to males observed here is
similar to Beasley and Standley’s (2002) ﬁndings. Of the characters in their sample,
14% were female and 72% were male. The results are also congruent with the
Children Now (2001) study, which found that 73% of video game characters are male
and 12% are female. While these numbers are not too terribly different from gender
representation on television (Gerbner, 1997), they pale in comparison to the US.
Census Bureau’s 2001 report that women make up 50.9% of the US. population.

Not only are women underrepresented, but when they are shown it is often in
a hypersexual context, which supports hypothesis 2. Signiﬁcant ﬁndings were found
for 5 of the 7 hypersexuality variables that were applicable to both males and females,
including sexually revealing clothing, nudity, body image, waist size, and
appropriateness of attire. To recap, 41% of all female characters were shown wearing
sexually revealing clothing compared to 11% of males. Further, 28% of all females,

compared to 2% of all males, were depicted with inappropriate clothing for the task at

21

hand, ostensibly to help accent the feminine form. One quarter of the women in this
study (25%) were shown with an unrealistic body image. This is consistent with
Children Now’s ﬁndings which revealed that nearly 20% of female characters are
shown with an unrealistic body image. Nudity had the greatest disparity between
females and males. Women (43%) were ten times more likely to be shown nude
(partial or full) than males (4%).

As long as game developers design games for a male audience,
hypersexualized female characters will continue to hold their attention for hours of
game play. Although this may seem harmless, theory suggests otherwise. The
repetition of seeing women as hypersexualized beings can reinforce the belief that
women are to be regarded as sex objects. The disproportionately small waists and
voluptuous breasts for female characters may contribute to body dissatisfaction for
female gamers and unrealistic expectations of women for impressionable male
gamers. This has two major implications.

First, this medium may be reinforcing to some that women are supposed to
have unrealistic body proportions. When these proportions are unattainable, it can
lead to low self-esteem and body dissatisfaction for female gamers. Taking that a
step ﬁrrther, low self-esteem and body dissatisfaction may prompt some to adopt
unhealthy eating habits trying to attain an idealized image. Second, it also sends a
message to male gamers of what women should look like. Since very few women can
achieve video game character proportions naturally, males may have unrealistic
expectations of the feminine body image. If males are disappointed that women

cannot achieve these proportions, this may hinder relational development. Because

22

these assertions are speculative at best, future research is needed to assess the impact
of repeated exposure to such portrayals on male and female gamers.

Research question I asked if game rating inﬂuenced the nature of
hypersexualized portrayals in video games. Across sex, the ﬁndings indicate that
there is a signiﬁcant difference between E, T, and M rated games on nudity. At ﬁrst
glance, it appears that the ESRB ratings do a fairly good job of ﬁltering out this one
hypersexuality variable. However, sexually revealing clothing, disproportionately
small waists, unrealistic body proportions and appropriate attire did not differ across
game ratings. This indicates that variables of hypersexuality as deﬁned by this study
are prevalent throughout E, T, and M rated video games instead of just T or M as may
be expected. If parents choose to screen video games by ESRB rating it is important
that they understand that the portrayals of video game characters wearing sexually
revealing clothing with small waist sizes and unrealistic body proportions can still
have an impact on young garners. Repeated exposure to video game characters with
unrealistic proportions in combination with similar “thin ideal” reinforcement from
other media may explain why as many as 10 in 100 young women in the US suffer
from eating disorders (AACAP, 2004).

When ESRB is broken down by sex, women are represented the least in E
rated games (n=11). It is also important to note that the highest proportion of
unrealistic body images and small waist sizes for women appear in E-rated games.
Thus, if video games were screened for young gamers by ESRB rating, women would
be virtually nonexistent. When they were shown, impressionable video game players

would be exposed to women with unrealistic body shapes.

23

Research question 2 asked if brand of console inﬂuenced the amount and the
type of hypersexualized portrayals in video games. The ﬁndings indicate that there
was a signiﬁcant difference for sexually revealing clothing but not for nudity,
appropriateness of attire, unrealistic body image, or small waist size. Characters
wearing sexually revealing clothing are more prevalent in games for the Nintendo
GameCube. On paper, it appears that Sony PlayStation 2 contains less hypersexuality
than the Nintendo GameCube or the Microsoft Xbox. However, these results need to
be interpreted with caution. First, the variables that compose hypersexuality still
appear in Sony PlayStation 2 games. Second, when one hypersexuality variable
appears, it is likely that another one or multiple others will also appear. This
phenomenon was just as likely to occur on the Sony PlayStation 2 as on the Microsoft
Xbox console. It is apparent then, that the best way to screen for hypersexuality
would be to look at ESRB rating and not simply the brand of game console.

When hypersexuality variables were examined across sex by console,
differences appeared. Three out of the seven variables applicable to both sexes (i.e.,
unrealistic proportions, sexually revealing clothing, nudity) appeared in the greatest
frequency for women versus men on the Microsoft Xbox console. Although it would
seem that parents would want to avoid the Microsoft Xbox, in favor of the Sony
Playstation 2 or Nintendo GameCube, this may not be the case. Even though the
Microsoft Xbox has the highest frequency for three of the seven hypersexuality
variables tested, at the time of this writing, it was also the only game console that

included a parental control.

24

Overall, this content analysis was designed to examine hypersexuality in
video game characters. Since hypersexuality as a variable has received considerably
less attention than violence in terms of its effects, this study serves as a platform from
which to further examine hypersexuality in video games. F ollow—up studies are
recommended to test the validity of the variables used to determine hypersexuality in
video game characters.

One limitation of this study is its focus exclusively on game play. The pre—
recorded video games segments that were coded did not include the introduction
section. Nor did they include the grand ﬁnales where sexual behaviors or sex talk
may occur. Future studies should be conducted to see if game introductions or game
ﬁnales contain hypersexuality codes such as sexual talk or sexual behavior. Since the
ﬁnales are a form of reward for the player, future studies should examine whether or
not sex talk and sexual behaviors are present. If they are, it is important to know
which types of behaviors are depicted.

In sum, the ﬁndings show that there are pronounced differences in the amount
of males and females in video games as well as the way in which they are depicted.
Females are more likely in games than males to be presented in a hypersexual light.
Such portrayals, from a social cognitive approach, may have negative effects on

young gamers’ socio-emotional development.

25

FOOTNOTES

1 Character role was originally coded as a demographic variable. However, it
was dropped from the study as it did not meet the predetermined minimum level of

.80 for inter coder reliability.

2 Physical attractiveness was originally coded as a hypersexuality variable.
However, it was dropped from the study as it did not meet the predetermined

minimum level of .80 for inter coder reliability.

3 The variable character body proportion examines all aspects of the body
including feet, legs, waist, hips, breasts, head size, etc. Breast size and waist size are
not mutually exclusive of this variable but do contribute to overall body
proportioning. The decision was made to code breast size and waist size individually
as previous studies have placed importance on the import of looking at them in

isolation.

4Since many video games contain multiples of certain characters that are

indistinguishable from one another, only single images were coded and frequency
was noted. For example, in the Incredible Hulk video game, many Army characters
were observed that were indistinguishable from one another. When this occurred, the
coder completed the codes for the character and then tallied the frequency for which
that character occurred in the video game. A total of 56 characters in the sample
(11%) featured replicates. Of those characters that featured replicates, a total of 38%
featured 1-5 replicates, 30% featured 6-15 replicates, 23% featured 16-42 replicates

and 7% featured 74-361 replicates.

26

Table 1
T op Twenty Video Games by Game Platform Included in Analyses

Nintendo GameCube
1. Godzilla: Destroy All Monsters Melee
Animal Crossing
Super Mario Sunshine
Luigi 's Mansion
Sonic DX Adventure
NBA Street Vol 2
Enter the Matrix
Def Jam Vendetta

l°9°N9~Prtki~N

Super Smash Bros Melee
10. Social Mega Collection
11. Sonic Adventure 2
12. James Bond, 007: Nightﬁre
I 3. Star F ox Adventures
14. Lord of the Rings: Two Towers
15. Mario Party 4
16. Metroid Prime
1 7. Resident Evil: Zero
18. Legend of Zelda: Wind Walker
19. Wario World
20. The Sims
Sony PlayStation 2
1. Grand Theft Auto III
2. Midnight Club 11
3. Hack: Infection
4. Grand Theft Auto: Vice City
5. Dragonball-Z: Budokan

27

6. Def Jam Vendetta

7. Enter the Matrix

8. NBA Street Vol. 2

9. NBA Live 2003

10. Tiger Woods PGA 2003

11. Tom Clancy's Splinter Cell

12. Xenosaga Episode J

13. T enchu 3: Wrath of Heaven

14. ATV Oﬂ‘road Fury 2

15. MVP Baseball 2003

16. SOCOM: US Navy seals

1 7. Dynasty Warriors 4

18. Tom Clancy’s Ghost Recon

19. Yu Gi Oh!

20. The Getaway
Microsoft Xbox

1. World Series Baseball 2K3
MVP Baseball 2K3
Midnight Club 11
Project Gotham Racing
Tom Clancy's Ghost Recon
NBA Street Vol. 2
The Sims
Wolfenstein: Tides of War
Dead or Alive Extreme Beach Volleyball

PPONQSArkSHN

10. Enter the Matrix

1 1. Tiger Woods PGA Tour 2003
12. Tao F eng: Fist of the Lotus
13. Max Payne

1 4. Halo

28

15. Unreal Championship

16. Brute Force

1 7. Tom Clancy's Splinter Cell
18. X2: Wolverine's Revenge
19. ATV 2: Quad Power Racing
20. Hulk

29

Table 2

Frequency of Characters by ESRB Rating, Platform, and Sex

 

 

 

Microsoft Sony Nintendo Total
Xbox PlayStation GameCube
2

Everyone

Men n=44 n=72 n=69 185

Women n=1 n= =8 11
Teen

Men n=29 n=3 5 n=3 7 l 01

Women n=10 n=1 1 =6 27
Mature

Men n=72 n=5 1 n=10 133

Women n=12 n=19 n=1 32
Total 168 190 131 489

 

30

Table 3

Percentage of Characters Showing Hypersexuality Indicators by ESRB Rating and

Sex

 

 

 

 

 

Everyone Teen Mature Total
Women 7
Sexually Revealing Clothing 33% 30% 53% 41% r
n=3 n=8 n=17 n=28/ 70
Partially or Totally Nude 22% 41 % 50% 43%
n= n=1 1 n=16 n=29
Unrealistic body image 44% 8% 34% 25%
n=4 =2 n=1 1 n=17
Small waist 56% 39% 38% 40%
n=5 n=10 n=12 n=27
Breast size (voluptuous) 1 1% 12% 43% 26%
n=1 n=3 n=13 n=17
Inappropriate attire 9% 19% 17% 16%
n=l n=5 n= n=1 1
Men 4 i Q
Sexually Revealing Clothing 18% 6% 6% 11% c '
n=32 n=6 n=7 n=45 /s"r 'C
Partially or Totally Nude 1% 8% 5% 4%
n=2 =8 n=6 n=16
Unrealistic body image 1% 6% 0% 2%
=1 n=6 n= n=7
Small waist 0% 3% 1% 1%
=0 n=3 =1 n=4
Breast size (voluptuous) na na na na
Inappropriate attire 0% 6% 0% 2%
n=0 n=6 n=0 n=6

 

Note: All percentages are rounded in the table and not carried out to the ﬁrst decimal

point.

31

Table 4

Frequency and Percentage of Characters Showing Hypersexuality Indicators by

ESRB Rating

 

 

Everyone Teen Mature Total

Sexually Revealing Clothing n=35 n=14 n=24 n=73
(19%) (11%) (15%) (16%)

Partially or Totally Nude n=4 n=19 n=22 n=45
(2%) (15%) (14%) (10%)

Unrealistic body image n=5 n=8 n=11 n=24
(3%) (7%) (7%) (5%)

Small waist n=5 n=13 n=13 n=3l
(3%) (l 1%) (8%) (7%)

Breast size (voluptuous) n=1 =3 n=13 n=17
(1 1%) (12%) (43%) (26%)

Inappropriate attire n=l n=11 n=5 n=17
(.5%) (9%) (3%) (4%)

 

32

Table 5

Percentage of Characters Showing Hypersexuality Indicators by Platform

 

 

Microsoft Sony Nintendo Total
Xbox PlayStation 2 GameCube

Sexually revealing clothing n=19 n=18 n=36 n=73
(12%) (10%) (31%) (16%)

Partially or Totally nude =22 n=16 n=7 n=45
(13%) (9%) (6%) (10%)

Unrealistic body image n=14 n=6 n=4 n=24
(9%) (3%) (4%) (5%)

Small waist n=16 n=1 1 n=4 n=3]
(10%) (6%) (4%) (7%)

Breast size (voluptuous) n=13 n=3 n=1 n=17
(59%) (10%) (8%) (26%)

Inappropriate attire n=9 n=3 n=5 n=17
(6%) (2%) (4%) (4%)

 

33

Table 6

Percentage of Characters Showing Hypersexuality Indicators by Platform and Sex

 

Microsoft Sony Nintendo Total
Xbox PlayStation 2 GameCube

 

 

 

 

Women
Sexually Revealing Clothing 61% 34% 23% 41%
n=14 n=1 1 n= n=28
Partially or Totally Nude 70% 28% 31% 43%
n=16 n=9 n=4 n=29
Unrealistic body image 55% 13% 8% 25%
n=12 n=4 n=1 n=17
Small waist 59% 31% 31% 40%
n=13 n=10 n=4 n=27
Breast size (voluptuous) 59% 10% 8% 26%
n=13 n=3 n=1 n=17
Inappropriate attire 30% 6% 20% 16%
n=6 n=2 n= n=1 1
Men
Sexually Revealing Clothing 4% 5% 32% 11%
n= n= n=33 n=45
Partially or Totally Nude 4% 5% 3% 4%
n=6 n= n= n=16
Unrealistic body image 1% 1% 3% 2%
n= n=2 n= n=
Small waist 2% 1% 0% 1%
n= =1 n=0 n=4
Breast size (voluptuous) na na na na
Inappropriate attire 2% 1% 2% 2%
n= =1 =2 n=6

 

34

BIBLIOGRAPHY

American Academy of Child & Adolescent Psychiatry (AACAP), (2004) Teenagers
with Eating Disorders. Retrieved on August 17, 2004 from
http://www.aacap.org/publications/factsfam/eatinghtm

Baca, J. (2003). Fact Sheet on HR. 669: Protect Children From Video Game Sex &
Violence Act of 2003. United States House of Representatives (n.d.). Retrieved
on June 13, 2004 from
http://www.house.gov/bﬂhotissues/video_factsheet.htrn

 

Bandura, A. (1965). Behavioral modiﬁcation through modeling practices. In L.
Krasner & I. Ullrnan (Eds), Research in behavior modiﬁcation (pp. 310-340).
New York: Holt, Rinehart & Winston.

Bandura, A. (2002) Social cognitive theory of mass communication. In Bryant &
Zillman, Media eﬂects: Advances in theory and research (pp. 121-153). '9
Mahwah, NJ: Lawrence Erlbaum Associates, Inc.

Beasley, B., & Standley, TC. (2002). Shirts vs. skins: Clothing as an indicator of
gender role stereotyping in video games. Mass Communication & Society,
5(3), 279-293.

Brurnberg, J. J. (1997). The body project: An intimate history of American girls. NY:
Random House.

Calvert, S.L., & Huston, A.C. (1987). Television and children’s gender schemata. In
L.S. Liben & M.L. Signorella (Eds), Children ’s gender schemata (pp. 75-88).
San Francisco: Jossey Bass. \/

Children Now (2001). Fair Play? Violence Gender and Race in Video Games. ,/
Oakland, CA: Children Now. ”I

Dietz, T.L. (1998). An examination of violence and gender role portrayals in video
games: Implications for gender socialization and aggressive behavior. Sex ¢
Roles, 38 (5/6), 425-442.

Dill, K.E., & Dill, J .C. (1998). Video game violence: A review of the empirical
literature. Aggression and Violent Behavior, 3(4), 407-428.

Entertainment Software Association. (2004). Top Ten Industry Facts. Retrieved on

June 15, 2004 from http://www.theesa.com/pressroomhtml. ‘/

Field, A.E., Cheung, L., & Wolf, AM. (1999). Exposure to mass media and weight
concerns among girls. Pediatrics, 103(3), 36.

35

Frederickson, B.L., & Roberts, TA. (1997). Objectiﬁcation theory: Toward
understanding women’s lived experiences and mental health risks. Psychology
of Women Quarterly, 21, 173-206.

Funk, J .B., & Buchman DD. (1996) Playing violent video and computer games and
adolescent self-concept. Journal of Communication, 46(2), 19-32.

Gerbner, G. (1997). Gender and age in prime-time television. In S. Kirschner & D.
A. Kirschner (Eds), Perspectives on psychology and the media (pp. 69-94).
Washington, DC: American Psychological Association.

Haninger, K., & Thompson, KM. (2004). Content and ratings of teen-rated video V
games. JAMA, 291(7), 856-865.

Harrison, K., & Cantor, J. (1997). The relationship between media consumption and
eating disorders. Journal of Communication, 47(1), 40-67.

Jansma, L., Linz, D., Mulac, A. & Irnrich, D. (1997). Men’s interactions with women
after viewing sexually explicit ﬁlms: Does degradation make a difference?
Communication Monographs, 64 (1), 1-24.

Kaiser Family Foundation (1999). Kids & media @ the new millennium: A
comprehensive national analysis of children ’s media use. Menlo park, CA:
Author.

Kaiser Family Foundation (2002). Children and video games. Menlo Park, CA: \/
Kaiser Family Foundation.

Kaiser Family Foundation (2003). A biennial report of the Kaiser Family
Foundation: Sex on TV 3. Menlo Park, CA: Kaiser Family Foundation.

Kunkel, D., & COpe, KM. (1999). Sexual messages on television: Comparing
ﬁndings from three studies. Journal of Sex Research, 36(3), 230-236.

Labre, M.P. ( 2002). Adolescent boys and the muscular male body ideal. Journal of
Adolescent Health, 30 , 233-242.

Lando, H. A., & Donnerstein, E. (1978). The effects of a model's success or failure on
subsequent aggressive behavior. Journal of Research in Personality, 12, 225-
234.

Loﬁus, T. (2004) Is that your joystick? Or are you playing “Leisure Suit Larry? ”
Retrieved October 30, 2004, from http://msnbc.msn.com/id/6072813/

Morris, C., (2002, September 17). Bikes, sex and volleyball: CNN Money. Retrieved
June 01, 2004, from http://money.cnn.com/2002/O9/ 16/commentary/ game

36

Mulac, A., J ansma, L.L., & Linz, D.G. (2003). Men’s behavior toward women after
viewing sexually-explicit ﬁlms: Degradation makes a difference.
Communication Monographs, 69(4), 311-328.

Mumen, S.K., Smolnak, L., Mills, J.A., & Good, L. (2003). Thin, sexy women and
strong, muscular men: Grade—school children’s responses to obj ectiﬁed
images of women and men. Sex Roles, 49, 9/10, 427-437.

Paik, H., & Comstock, G. (1994). The effects of television violence on antisocial
behavior: A meta-analysis. Communication Research, 21, 516-546.

Parents Television Council (2004). Video game violence, sex, and profanity content
reviews for parents. Retrieved October 30, 2004, from \/
http://www.parentstv.org/PTC/videogames/reviews/mainasp

Polce-Lynch, M., Myers, B.J., Kliewer, W., & Kilmartin, C. (2001). Adolescent self-
esteem and gender: Exploring relations to sexual harassment, body image,
media inﬂuence, and emotional expression. Journal of Youth and
Adolescence, 30(2).

Provenzo, E.F., Jr., (1991). Video kinds: Making sense of Nintendo. Cambridge, MA:
Harvard University Press.

Peters, K. (January 24, 2004). Parents weigh in on the media. The Seattle Times, 4th
Edition, E2.

Smith, S. L., Wilson, B. J ., Kunkel, D., Linz, D., Potter, W. J ., Colvin, C., &
Donnerstein, E. (1998). Violence in television programming overall:
University of California, Santa Barbara. National television violence study

(Vol. 3, pp. 5-220). Newbury Park, CA: Sage Publications.

Spitzer, B.L., Henderson, K.A., & Zivizn, M.T. (1999). Gender differences in
population versus media body sizes: A comparison over four decades. Sex
Roles, 40 (7/8), 545-565.

Thompson, E. H., & Pleck, J. H. (1986). The structure of male role norms. American ’9
Behavioral Scientist, 29, 531-543.

Unites States Department of Commerce, Bureau of the US. Census (2001). Male
Population Grew Faster Than Female Population Census 2000 Analysis
Shows. Retrieved on August 17, 2004 from http://wwwcensusgov/Press-
Release/www/releases/archives/population/000440.html

Vessey, J. A., & Lee, J. E. (2000) Violent video games affecting our children.
Pediatric Nursing, 26(6), 607-609.

37

Wilson, B. J ., Kunkel, D., Linz, D., Potter, W. J., Donnerstein, E., Smith, S.L.,
Blumenthal, E., & Berry, M. (1997). Violence in television programming
overall: University of California, Santa Barbara. National television violence
study (Vol. 2, pp. 3-204). Newbury Park, CA: Sage Publications.

38

           
 

     

RAH S

iiillllllilllil

3 666

M STE

llllillilljlllllll

93