University of Southern California Dissertations and Theses

Role enactment in interactive media: a role-play perspective

(USC Thesis Other)

Role enactment in interactive media: a role-play perspective

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

ROLE ENACTMENT IN INTERACTIVE MEDIA: A ROLE-PLAY
PERSPECTIVE

by

Younbo Jung

A Dissertation Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(COMMUNICATION)

August 2007

Copyright 2007 Younbo Jung

ii
Dedication

To My Family

iii
Acknowledgements
I would like to thank the fine faculty of the Annenberg School for
Communication and my dissertation committee members at the University of
Southern California for assisting me through my endeavor for a Ph.D. Especially, I
am grateful to my advisor and committee chairperson, Dr. Margaret McLaughlin, for
her guidance, encouragement, and mentorship for the past four years. Dr.
McLaughlin surely played multiple roles in my life as a mentor, friend, mother, and
supporter whenever I needed her. I would also like to extend my gratitude to Dr.
Peter Vorderer as he has always treated me as a colleague and supported me with his
big smiles. I am thankful for Dr. Kwan Min Lee for his invaluable lessons and
inspiration for me. I really appreciate his personal training for academic survival. I
am grateful for Dr. Albert Skip Rizzo for his comments and advice, as well as his
passion for virtual reality, which significantly influenced my area of interests.
Having greatly benefited from my own mentors, I will pay forward what I owe them
to my future students.
Most of all, I would like to thank my family members. Without their love,
trust, and encouragement, I could have not completed this work. I am sincerely
grateful to my wife, Jooyeun, for taking care of three children, including myself and
the second one, Leah, born in June 2007. I thank my son, Ian, for pushing me out of
the bed in every evening for more studies and for entertaining me. I thank my mother
for being the biggest fan of mine all the time. Her belief in me made me keep

iv
pursuing my goals. I am grateful for my sister, brother-in-law, and nephew for their
love and support. Especially, I thank my sister for her playing the role of a son to our
parents during my absence, in addition to her own role of a perfect daughter. I thank
my parents-in-law for their love, care, and trust. I really appreciate their kindness as
they have always treated me like a son. I am also grateful for my sister-in-law, and
her husband for their thoughtful consideration. I certainly owe many fashionable
outfits to my sister-in-law. I thank my brother-in-law, his wife, and nephew for their
support and care as well.
Finally, I would like to thank my father who wanted to see this moment
more than anyone else but passed away two years ago. My father has always been
my hero and role model. I wish I finished my degree earlier. Nevertheless, I thank
God for everything.

v
Table of Contents
Dedication

ii
Acknowledgements

iii
List of Tables

vii
List of Figures

viii
Abstract

ix
Introduction

1
Chapter 1: Role play and self-concept management in interpersonal settings
Self-concept in social contexts
Role Salience and Self-concept Construction
Role Play and Role Salience
Conclusions

4
4
7
16
18
Chapter 2: Origins of role play: Pretend play, metarepresentations, and theories
of mind
Development of Pretend Play
How to Play Pretense
The Origin of Pretend Play and Other By-Products
Conclusions

19
19
23
30
41
Chapter 3: Revisiting entertainment theories from a role-play perspective
Feeling Within or Feeling With
Feeling Overall
Role Play Model
Entertainment Theories, Revisited

42
43
57
64
72
Chapter 4: Role enactment based on gender stereotypes in interactive media:
Effects of role play on the management of the self-concept and physical distance
in virtual reality environments
Gender Identity and Gender-Stereotyping Behaviors
Method
Results
Discussion
Experiment 2
General Discussion and Conclusions

75
78
80
84
89
92
102

vi
Conclusion

106
References

108
Appendix A: A Screen Shot of the Virtual Bar (the Male-Role Condition with
the First-Person View)

118
Appendix B: A Screen Shot of the Virtual Bar (the Female-Role Condition with
the First-Person View)

119
Appendix C: A Screen Shot of the Virtual Bar (the Male-Role Condition with
the Third-Person View)

120
Appendix D: A Screen Shot of the Virtual Bar (the Female-Role Condition with
the Third-Person View)

121
Appendix E: Male-Role Script (Participants Played the Role of Albert)

122
Appendix F: Female-Role Script (Participants Played the Role of Ashley)

125
Appendix G: Measures

128

vii
List of Tables
Table 1. ANCOVA results in Experiment 1

86
Table 2. ANCOVA results in Experiment 2

98

viii
List of Figures
Figure 1. Two Step Model: The Relationship between Role play and Self-
concept

8
Figure 2. An example of multiple layers in pretend play

27
Figure 3. Layering Architecture

28
Figure 4. Multi-layering model of role play in the context of interactive
videogames

40
Figure 5. Cross sectional view of the entertainment experience

44
Figure 6. Identification in the multi-layering model of role play

49
Figure 7. Block view of the entertainment experience

58
Figure 8. Role-play model in the non-interactive entertainment experience

67
Figure 9. Role play model for interactive media

70
Figure 10. Identification as a product of gender role and biological sex

87
Figure 11. Main effect of gender role on centrality: Experiment 1

88
Figure 12. Main effect of gender role on centrality: Experiment 2

97
Figure 13. Interaction effect on physical distance

99

ix
Abstract
Role play prevails in our daily activities from working in a company to playing
a videogame. In this regards, conceptualizing the role-play process helps us
understand how people interact with other people in both face-to-face and
computer-mediated environments, as well as how such social interactions could
affect people’s self-concept and corresponding behaviors in particular ways.
The purpose of my dissertation is to propose a new theoretical model of role
play to explicate our virtual experience in media, to discuss theoretical
implications of the model, and to provide empirical evidence for the utility of
the role play perspective in understanding the effects of role enactment in
interactive media. In the first three chapters I discuss a new theoretical
framework for the multi-layering model of role enactment. Particularly, I review
the literature on social psychology and communication to address the
prevalence of role play in our daily activities and the relationship between role
play and changes in the self-concept and behaviors in Chapter 1. In Chapter 2, I
conceptualize a multi-layering model of role enactment and explain how we
engage in role play in videogames based on Clark’s multi-layering model (Clark,
1996) and Leslie’s metarepresentations (Leslie, 1987). In Chapter 3,
I elaborate more on the multi-layering model of role play to enhance our
understandings of entertainment experience based on the review of important aspects
from the seven entertainment theories. In the final chapter, I report on the results of

x
two experiments in order to provide preliminary empirical evidence to support the
utility of the role-play perspective. Not only do the results confirm the usefulness of
a new role-play perspective to understand our psychological and behavioral reactions
from the role enactment in interactive media, but also they suggest that altering
contextual features of interactive media could influence the degree to which people
conform to role play. Theoretical implications as well as practical strategies for the
design of interfaces for interactive media (e.g., virtual reality systems and
videogames) are discussed.

1
Introduction
Role play prevails in our daily activities from working in a company to
playing a videogame. In fact, we are given a role identity from the very moment
when we are born: we become children of our parents. As a son or daughter, a
student, and a member of a society, we play multiple role identities through which
we come to realize ourselves as both unique and harmonized individuals (see social
identity theory [Tajfel & Turner, 1986]). Thus, conceptualizing the role-play process
helps us understand how people interact with other people, as well as how such
social interactions could affect people’s self-concept and corresponding behaviors in
particular ways. In addition, understanding role-enactment process in face-to-face
environments from the role play perspective would enhance our knowledge of virtual
experience in interactive media (see Reeves & Nass, 1996 for the “media equation”).
The purpose of my dissertation is to propose a new theoretical model of role
play to explicate our virtual experience in media, to discuss theoretical implications
of the model, and to provide empirical evidence for the utility of the role play
perspective in understanding the effects of role enactment in interactive media.
Specifically, in Chapter 1 I review the literature on social psychology and
communication to address the prevalence of role play in our daily activities and the
relationship between role play and changes in the self-concept and behaviors. Based
on that discussion, I propose a two-step model of role play that serves as a basis for

2
model building in the next two chapters and preliminary model testing in the final
chapter.
In Chapter 2, I review the literature on cognitive, developmental, and
evolutionary psychology to discuss the evolutionary nature of role play, by
comparing it with children’s pretend play and the ability to understand other people’s
minds (i.e., theory of mind). The focus here is more on why questions of media
research rather than traditional how questions of effect studies: why do we enjoy
playing multiple roles in interactive media? Based on Clark’s multi-layering model
(Clark, 1996) and Leslie’s metarepresentations (Leslie, 1987), I conceptualize a
multi-layering model of role enactment and explain how we engage in role play in
videogames.
In Chapter 3, I review theories that explain our entertainment experience,
based on the role play model discussed in previous chapters. Seven entertainment
theories of identification, sympathy, empathy, para-social interaction, flow,
transportation, and presence are assigned to three categories, based on how each
theory explains the entertainment experience: feeling within, feeling for, and feeling
overall. Based on the review of important aspects from the seven theories, I elaborate
more on the multi-layering model of role play to enhance our understandings of
entertainment experience. Four assumptions and four dimensions of the role play
model are identified, as well as individual predispositions that may affect the
enjoyment of media. Finally, potential limitations of current entertainment theories

3
are revisited from the perspective of a new framework (e.g., the balance of role
expectation and role performance).
In the final chapter, I report on the results of an empirical study that consists
of two experiments. Experiment 1 (n = 48) investigates the effects of gender-role
play on gender-role identity salience, gender self-concept, and gender-stereotyping
behaviors in interpersonal distance in virtual reality environments. The results
support the utility of the role-play perspective and interpersonal theories to
understand a new relationship of human-computer interaction in a new context of
computer-mediated communication. Experiment 2 (n = 84) is conducted to confirm
the findings of Experiment 1 with larger samples, as well as to examine the effects of
different points of view on gender-role salience, gender self-concept, and physical
distance management in interactive media. Not only do the results confirm all the
findings of Experiment 1, but also they suggest that altering contextual features of
interactive media could influence the degree to which people conform to role play.
Theoretical implications as well as practical strategies for the design of interfaces for
interactive media (e.g., virtual reality systems and videogames) are discussed.
In conclusion, I summarize all chapters briefly and assert the utility of the
role-play perspective and multi-layering model in understanding role enactment in
interactive media.

4
Chapter 1: Role play and Self-Concept Management in Interpersonal
Settings
Various disciplines including mathematics, biology, chemistry, philosophy,
epistemology, sociology, and psychology have illustrated the concept of the self
historically (e.g., Adams & Marshall, 1996; see Baumeister, 1987 for a detailed
historical review of identity). Even within the field of psychology, the self could be
understood and explained independently based on the literature such as
psychoanalysis, cognitive psychology, and social psychology (Westen, 1992).
Among the various notions of self in different fields of scholarship, I will mainly
focus on the literature from social psychology to define the self-concept and identity
because the purpose of the current chapter is to examine the relationship between the
self-concept and role play. That is, the aim of the current chapter is to explicate how
individuals internalize the self-concept from playing multiple roles (e.g., through
interpersonal interactions in social contexts) and how the salient self-concept
influences their corresponding behaviors and the “true” selves or vice versa.
Self-Concept in Social Contexts
Westen (1992) defines the self-concept as “the prototypic, generalized
representation of self that most people verbalize when asked to do so” (p. 7).
Similarly, Baumeister and Muraven (1996) define identity as a composite definition
of the self, including “social roles, reputation, a structure of values and priorities, and
a conception of one’s potentiality” (p. 406). As shown by the above definitions, the

5
terms self-concept and identity are often used interchangeably (e.g. Adams &
Marshall, 1996; Baumeister, 1987; Baumeister & Muraven, 1996; Goossens &
Phinney, 1996), and will be so used in the current chapter.
One commonality in the above definitions is the importance of social
interaction in constructing and confirming the self-concept .Individuals strive to
construct the desired self-concept by strategically managing the impressions they
make on others, which may not be fully appreciated until other people perceive the
constructed self-concept as intended. In other words, self-representations and the
self-concept are often embedded in relationship schemas so that it is essential to get
validation of one’s identities from other people for realization of the self (Baumeister,
1992; see Goffman, 1959 for impression management). In fact, the expectations of
falsely evaluated self-concept result in a psychological burden such as a state of
unpleasantly heightened sensitivity to others’ evaluations to the self. Bosson,
Prewitt-Freilino, and Taylor (2005) found self-conscious discomfort in individuals
when they expected to be falsely accused of being a deviant based on their enacted
role-violating behaviors. The discomfort with the role violation was eased when
individuals used a public disclaimer about their true identities against the enacted
role-violating behaviors. To sum up, the study successfully demonstrated the
importance of the validation of ones’ desired self-concept in social contexts.

6
The self-concept in social contexts becomes clearer when we look at the
following definition. Adams and Marshall (1996) define identity as a “social-
psychological construct that reflects social influences through imitation and
identification processes and active self-construction in the creation of what is
important to the self and to others” (p. 433). According to them, social interactions
provide two functions of socialization: an individual and a social function. The
individual function of socialization is related to the self as a unique and individuated
person. On the other hand, the social function is related to the self as belonging to
and caring about significant others. Thus, social interactions may facilitate not only
the perception of the self as a unique individual but also the perception of the self in
a social context (see also Tajfel & Turner, 1986). In a similar vein, self-
categorization theory posits that individuals categorize themselves with regards to
social contexts and activate a specific self-categorization or identity more salient
depending on the cognitively represented self through social interactions (Palomares,
2004). Although the self-concept could be internally conceptualized in isolation from
others, the self-concept is often internalized from interpersonal events in that the self
also evolves as an interpersonal tool for developing interpersonal relationships
(Baumeister, 1992).
Taken together, I define the self-concept as a representation of the self that
could be (re)constructed through the internalization of social interactions with
regards to the uniqueness of an individual, as well as the harmonized self in a social

7
context. Internalization here refers to the private and intrapsychic process of
(re)conceptualizing one’s self-concept in accordance with his or her recent social
behaviors (see Tice, 1992).
Finally, it is worth noting that my definition of the self-concept is very
different from the perspective of the social identity and deindividuation (SIDE)
theory (Lea & Spears, 1992). The SIDE model posits that de-individuation referring
to the lack of unique attributes of an individual may enhance social identity of in-
groupness. In the SIDE model, personal and social identity are assumed to have a
negative relationship in that de-accentuating one identity may enhance the salience
of the other identity: by depersonalizing the self-concept, “I becomes we” (Brewer,
1991; p. 476). The major difference is the internalization process of social
interactions vis-à-vis the co-existence of the unique and harmonized self-concept in
my definition.
Role Salience and Self-concept Construction
A brief review of the literature in the previous section has demonstrated that
the self-concept is fluid and can be (re)constructed through social interactions.
Having said that, I propose that (1) role play is one of the most natural, efficient, and
implicit ways of making specific role identities highly salient, imposing new
expectations, beliefs, and desires embedded within the salient role identities on
individuals; and thus (2) influencing their self-concept construction (see Figure 1). In
this section, I review theories about role-identity salience, symbolic interactionism,

8
and relevant empirical studies in order to explain the second part of the proposed
hypothesis.

Figure 1. Two Step Model: The Relationship between Role play and Self-concept

According to symbolic interactionism, social interactions based on structured
role relationships influence the self-concept. The self is viewed as both social
structure and personality that is considered to develop through social interaction
(Callero 1985). In other words, individuals come to realize the self-concept through
social interactions with others by attaching to themselves symbols the meanings of
which grow out of the interaction (Stryker & Serpe, 1982). It is not surprising to see
that symbolic interactionism also emphasizes the self-concept in social contexts.
Based on symbolic interactionism, identity theory posits that individuals have
many identities, probably as many identities as the number of “distinct sets of
structured relationships in which they are involved” (Stryker & Serpe, 1982; p. 206).
Indeed, people are overwhelmed by their many identities in modern society. For
instance, I am a student, husband, father, son, research assistant, teaching assistant,
and Korean, to name a few. These identities are pertinent to social roles that are a set
of expectations about behaviors for a position in the structure of social relationships
Role Play
(Role Performance)
Role Identity
Salience
Self-concept
Construction
(1) (2)

9
(Shaw & Costanzo, 1982). Thus, they are often described as role identities to
emphasize an intimate relationship between the role and identity (Stryker & Serpe,
1982).
Interestingly, the overwhelming number of identities resides within an
individual in a hierarchically organized way without causing much of conflict. That
is, the salience of a specific role at the moment determines the hierarchical order of
various role identities so that the salient role identity along with embedded
expectations, beliefs and desires are brought into play in a given situation of social
interactions. The role salience, hierarchical structuring of role identities, is very
flexible, temporal, and, most importantly, context-dependent so that it is hard to
recognize the importance of role identities in general without considering social
contexts.
What then would be the consequences of role salience? According to Callero
(1985), the salient role influences self-concept construction in the following process:
(1) role salience leads to individuals’ selective perception of social objects (e.g., a
father may perceive toy stores easily among other stores; and a student libraries); (2)
the selective perception implies behaviors through which role identities are realized
and validated (e.g., buying toys may validate the role identity of a father); and (3)
finally, the validated role identities by a combination of the social structure and
individual characteristics such as self-esteem or satisfaction constitute the
representation of the self (e.g. the self-concept as a good father), which is consistent

10
with the definition of the self-concept in the current chapter (see the definition in the
previous section).
For example, my role identity as a scholar becomes highly salient when I am
writing this chapter. As a result, I have selective perception of things to do and
perform my behaviors according to the salient role identity and selective perception,
respectively. Because of the highly salient role identity at this moment, my other
identities such as a father are not activated, staying on the bottom of the hierarchical
organization of my identities. Then, these whole changes are likely to influence my
self-concept reflecting the salient role identity at least temporarily (e.g., feeling good
as a productive scholar or feeling bad as a father). If the period of role salience gets
longer, then the reconstructed self-concept would be continuously operative.
In fact, studies have demonstrated a positive relationship between role
salience and self-concept (re)construction. For example, Callero (1985) found a
positive relationship between the salience of role identity and self-concept as a blood
donor. In particular, Callero surveyed 658 correspondents from a sample of blood
donors through a mail questionnaire, including measures for the role salience among
seven role identities and self-concept as a blood donor. He also collected behavioral
data by looking at the respondents’ donation record for approximately six months
after completing the questionnaire. The results indicated that individuals with highly
salient blood-donor role identity had a self-concept as regular blood donors and
donated blood more often. In spite of positive results, the study has some limitations.

11
First of all, Callero (1985) treated the hierarchical structure of role identities as if it
were stable. He predicted behavioral consequences from the survey questionnaire
completed a maximum six month earlier. Given the flexible and context-dependent
nature of role salience as discussed above, it may be hard to claim a direct
relationship between role salience (i.e., a blood donator) and behavioral outcomes
(i.e., blood donation). It is possible that the self-concept as a regular blood donor has
been established prior to the survey, which may result in the behavioral outcome of
donating more blood. Secondly, Callero measured role salience via a mail survey
rather than inducing a certain role identity to become highly salient. As a result, the
causal direction of the relationship is unclear. Nonetheless, the study has successfully
demonstrated a positive relationship between role salience, the self-concept, and
behavioral consequences.
Similarly, a set of empirical studies by Tice (1992) showed that the self-
concept could change through the internalization of public behavior. Tice (1992)
states that induced behaviors in a social context may make people conform to a
certain aspect of the self implied by their overt behaviors. For the manipulation in
her study, participants were asked to portray themselves as either emotionally stable
or emotionally responsive, and in the following study either as extroverted or
introverted. The next manipulation was the social context (the public vs. private
condition). In the public condition, participants were told that an experimenter would
interview them from behind the one-way mirror in the laboratory. In the private

12
condition, participants were told that they would not be observed during the
experiment. After answering questions about the portrayed personality, participants
were asked to complete another survey for their “true” personalities in order to
compare the true personalities with the portrayed personality. The results showed
that inducing role salience in the social context (i.e., the public condition) influenced
individuals’ internalization and self-concept construction more significantly than
doing so in the private condition. Based on these findings, Tice concluded that
portraying a role in a public way creates greater internalization of the behavior than
portraying a role without others present to observe. It is also interesting that such
changes in the self-concept were not so completely temporary in her study because
the changes endured in a subsequent situation with a new interaction partner
consistently. In her second study participants were asked to move to the next room
where each of them brought in a chair to interact with another experimenter after
their finishing the initial experiment in a one-way mirror laboratory. Tice used
interpersonal distance as a behavioral measure for extroversion The results
confirmed the earlier findings that participants who portrayed an extroverted
personality seated themselves closer to the experimenter even in a subsequent
situation. Thus, enduring self-concept changes might be possible depending on the
crucial role of interpersonal or social contexts. The results imply that newly
constructed self-concept through social interactions may remain even after the end of
the initial social interaction. To sum up, the results of the study attest to the

13
importance of role salience through interpersonal contexts for producing self-concept
changes.
More recently, Collier and Callero (2005) found that establishing
commitment to the role of recycler among high school students created cognitive
schemata as a new role-based view of self, developed during their six-week field
experiment. Their role-identity salience process consisted of four elements: (1)
modeling high school students’ behavior after voluntary college students (i.e., role-
modeling); (2) providing information about meanings of recycling and recycler; (3)
letting students engage in multiple acts related to recycling; (4) creating self-
awareness consistent with those of recycler. The salient role through these
acquisition processes in social contexts developed students’ self-identification with
the role of recycler and associated cognitive schemata (i.e., the self-concept)
accordingly. From these findings, Collier and Callero argued that roles served as
resources for thinking and creating social positions. Playing a role in social contexts
may render a specific role identity highly salient, which, in turn, may affect
corresponding self-concept and behaviors without the individual's conscious
awareness (see the next section for detailed discussions about the relationship
between role play and role-identity salience).
Another very interesting area of research that investigates the relationship
between role salience and self-concept construction is gender. Studies about sex and
gender have started to regard gender as one of the social identities, instead of treating

14
gender as one of the given traits at birth (Cameron, 2004; Cameron & Lalonde, 2001;
Deaux, 1984; and Palomares, 2004). In fact, the simple biological distinction of male
and female cannot fully explain the complexities of gender-related self-concept in
modern society. Therefore, it is argued that gender-related identity derives not only
from biological sex but also from “attitudes and beliefs regarding sex roles and the
nature of gender relations” (Cameron & LaLonde, 2001, p. 61). For instance,
Cameron & LaLonde (2001) conducted an empirical study with 262 participants to
develop a measure of multidimensional gender-derived social identification and to
examine its relationship with gender-related ideology. The results showed three
dimensions of identity: in-group ties; cognitive centrality; and in-group affect. The
results also indicated a significant difference in gender-related self-concept within
women, depending on how individuals categorize themselves as traditional, non-
traditional, and feminist. It implies that gender-related self-concept may not be
exhibited in the same way by all group members (i.e., men or women). In other
words, the degrees of gender-related identity could vary as a social identity,
depending on how people identify with a specific gender in a given situation.
To reflect the idea of gender-derived social identity, Palomares (2004)
examined the relationship between gender salience and the use of gender-related
language. He hypothesized that individuals tend to behave in the salient gender
(identity) prototypical ways through communication in which they enact the gender
identity when the gender role becomes salient. In order to manipulate gender-identity

15
salience, participants were asked to think about themselves in terms of their gender
group in the high gender-salience condition, whereas participants were asked to think
about themselves in terms of being a student (another in-group identity) in the low
gender-salience condition. Then, participants in the two different conditions read two
parallel paragraphs tailored for each condition. The results showed that only gender
schematics, in both men and women, whose gender was salient used typical gender-
related languages. Based on the findings, Palomares (2004) suggests that gender
could be explained as the result of culture and that gender salience depends on
people’s “perceptions of the differences between men and women and the socially
meaningful group differences” (p. 580).
A potential limitation of the study is that role salience was induced by
participants’ portrayal of perceptions. If role salience was induced by social
interactions (e.g., playing a role with other people), the results might have been
influenced more significantly by the salient role. In addition, the study only
investigated the relationship between role salience and behavioral outcomes,
measured by the use of gender-related languages. Thus, we do not know whether or
not the salient behaviors might lead to reconstructing the self-concept. Although we
can infer the relationship from the previous discussion about internalizing behaviors
for validating and constructing the self-concept, future study needs to examine how
gender-role salience affects self-concept construction or gender-stereotyping.

16
Role Play and Role Salience
To summarize, studies have indicated a positive relationship between role
salience and self-concept construction in blood donation (Callero, 1985), portrayed
traits in public circumstances (Tice, 1992), recycler-role identity (Collier & Callero,
2005), and gender-related language use (Palomares, 2004). Although these studies
did not explain the inducing procedures of role salience as role play, all the
participants were asked to play certain roles either psychologically or physically, if
not both. In this regard, the first part of the proposed model, the relationship between
role play and role salience, could be intuitively understood, which completes the two
step model of the relationship between role play and the self-concept. Nevertheless,
the relationship becomes manifest when we look at the definitions of role and role
play.
The role is defined as a set of expectations for a position in a social structure
(Shaw & Constanzo, 1982). When a role becomes salient, the expectations
embedded within the role, including beliefs, desires, and behaviors are activated or
go to the top in the hierarchical structure of role identities. Having said that, role play
or role performance could be understood as an activity in which individuals try to
satisfy the imposed expectations through salient role behaviors in social contexts
(Sogunro, 2004). Similarly, Johnson and Johnson (1997) define role play as “a way
in which you can experience concretely the type of interaction under examination” (p.
59). They argue that role-play activities may lead individuals to change behaviors

17
and attitudes through experiencing social interactions. In addition, role play is also
known to encourage the acquisition of knowledge, skills, and attitudes in teaching
and learning environments by allowing direct experience similar to real life
situations (Sogunro, 2004). Thus, playing a role refers to experiencing a specific type
of simulated social interaction through which a relevant role may become salient,
followed by imposed expectations, behaviors, and (re)constructed self-concept
accordingly.
So far, I have discussed that role play is one of the most natural, efficient, and
implicit ways of making a certain role identity highly salient in social contexts
through which the embedded behaviors, beliefs, and desires may also become salient,
and thus may result in changes in one’s self-concept. Given the hypothesis, we can
assume that the degree to which individuals appreciate role play is likely to affect the
degree of role-identity salience and self concept construction respectively. If the
experience of role playing is pleasant, enjoyable, or favorable such as playing
massively multiplayer online role-playing games (MMORPGs), the impact of role
play on constructing the self-concept may be more significant. Indeed, the positive
relationship between positive self-esteem and successful performance of salient role
identities has been identified (Callero, 1985). The simple rules, clear role
expectations, and immediate feedback on role performance in MMORPGs may
stimulate users to reconstruct their self-concept accordingly at least temporarily
during their play, regardless of their true selves.

18
Conclusions
In this chapter I reviewed the literature to discuss relationships among role
play, identity salience, changes in the self-concept and corresponding behaviors.
Based on the discussion, I proposed a two-step model of role play: playing a role
may make a certain role-identity highly salient at a given moment, which, in turn,
may influence the self-concept and behavior correspondingly. I believe that the role-
play perspective can provide an explanation not only to understand our daily
activities, but also to comprehend under-appreciated activities in interactive media
such as videogame play. For example, becoming aggressive after playing violent
videogames (see Anderson & Bushman, 2001) or learning after playing educational
videogames (see Goodman, Bradley, Paras, Williamson, Bizzochi, 2006) could be
explained based on the role-play perspective, in addition to widely cited theories
such as priming and social learning theory. In Chapter 3, I will discuss theoretical
implications of the role-play perspective in greater detail. In Chapter 4, I will report
on results from an empirical study that examines on the proposed role-play
perspective.

19
Chapter 2: Origins of Role Play: Pretend Play, Metarepresentations and
Theories of Mind
In Chapter 1, I discussed what would happen to our self-concept and
behaviors when we play a certain role, which is related to typical how questions of
media research (e.g., the effects of role play on role-identity salience, the self-
concept and behaviors). In this chapter I address understudied why questions of
media research (e.g., fundamental reasons why we enjoy playing roles and
interactive media; and why most people easily exhibit the remarkable ability of role
play in daily activities). To do so, I first review the literature on young children’s
pretend play and discuss its relationship with the ability to understand someone
else’s mind (i.e., perspective taking) and role play. Then, I propose a new lens of
multi-layers of role-enactment model through which we may enhance our
understanding of the enjoyable experience of playing interactive media, mainly
videogames.
Development of Pretend Play
Pretend play (pretense) refers to children’s make-believe transformations
(Kavanaugh & Haris, 1994), behavior in a simulative or “as if” mode (Garvey, 1977),
the projecting of a supposed situation onto an actual one (Lillard, 1993), or an
extreme form of assimilation (Piaget, 1962). The importance of pretend play has
been identified with regard to the abilities for transformational thinking, social rituals,
reduction in egocentricity, perspective taking, and cooperative social problem

20
solving (see Fein, 1981, for a detailed review). Nonetheless, the complex and
remarkable abilities in pretend play are often overlooked in daily life because
pretend play is commonly and naturally observed among children over the age of
two (Jackowitz & Watson, 1980; Kavanaugh & Harris, 1994; Leslie, 1987).
Pretend play seems to first appear at about 12 or 13 months of age and
rapidly develops so that two year olds can understand not only pretend play by a
partner but also the consequences of such pretend play (e.g., children pretend that a
teddy bear gets wet after pretending to be poured water from an empty cup).
According to Fein (1981), the early pretend play at about 12 months is mainly self-
referenced (e.g., the child feeds self). The shift in pretend play occurs at about 15 to
21 months so that pretend play becomes other-referenced (e.g., the child plays with a
doll) and becomes more complex (e.g., sociodramatic play).
In fact, developmental changes in pretend play have been investigated in
many studies. For instance, Jackowitz and Watson (1980) propose a developmental
sequence of five steps of object transformations to explain children’s developmental
capability of joining pretend play (c.f., Leslie [1987] proposes three forms of
pretense: object substitution, attribution of pretend properties, and imaginary objects).
Jackowitz and Watson (1980) define the child’s level of transformation as the ability
to distance symbolic objects from their referents in terms of form and function,
which is similar to the decoupling argument by Leslie (1987). In particular, they
examined the relationship between age and pretend play in five levels of

21
transformation: (1) similar form and similar function (e.g. toy phone); (2) similar
form and dissimilar function or dissimilar form and similar function (e.g., toy banana
as a phone or walkie-talkie as a phone respectively); (3) dissimilar form and
ambiguous function (e.g. wooden block as a phone); (4) dissimilar form and
dissimilar function (e.g., toy car as a phone); and (5) no form and no function (e.g.,
pretending without object). In the study, they compared two groups of children, with
an average age of 16 and 24 months respectively, with a total of 48 normal children
joining pretend play in a playroom environment. The results indicated that older
children showed more object substitutions than younger children in general.
Specifically, 38% of the children in the older age group passed the last step of
transformation, whereas none of the children in the younger age group passed the
last step.
Similarly, Lewis and Ramsay (2004) examined the relationship between self-
recognition and pretend play with 66 children. They argued that mental state
knowledge of self (e.g., I know that my play is not real or I pretend) is necessary to
distinguish the pretend world from the real world; thus they believe that pretend play
is an early manifestation of theory of mind. In their study, the experimenter
surreptitiously applied unscented rouge on children’s noses. And self-recognition
was measured by the behavior of touching one’s own nose after looking at oneself in
front of a mirror. Then, the children were observed in three different scenarios of
pretend play. The results indicated that children with the ability of self-recognition

22
showed more pretend play than children without self-recognition. Based on the
findings, Lewis and Ramsay stated that pretend play involves mental state
knowledge of self that emerges among children around two and a half years old.
Although mental state knowledge of self is not as sophisticated as
metarepresentations in theory of mind or false-belief tasks (c.f., I know you do not
know I know), the study supports the metarepresentational abilities of children in the
second year, which is in contrast to an argument positing no relationship between
pretend play and any mental state knowledge (see Currie 1998; Leslie, 1987; Lillard,
1993; Nichols & Stich, 2000; and also the ‘how to play pretense’ section in the
current chapter for detailed discussions about metarepresentation in pretend play).
Metarepresentational abilities in pretend play, understanding a partner’s
pretend play, have been examined in another study. Kavanaugh and Harris (1994)
conducted a series of empirical studies in order to examine the relationship between
age and the ability to engage in joint pretend play. In their first experiment, two
groups of children with the average age of 21 and 29 months respectively were asked
to point out six pictures that depicted the relevant transformation after the
experimenter’s enacting the pretense. The results showed that the older group
performed significantly better than the younger group in terms of selecting the
pictures matching the pretend outcomes. In a subsequent experiment, they included
one more option of a picture depicting the irrelevant transformation in order to
eliminate an alternative explanation for the findings. That is, children might choose

23
any altered pictures without understanding the pretend enactment. Nonetheless, the
results confirmed the findings of the first experiment. The older group rarely chose
the irrelevant transformation and never failed to make a response, whereas the
younger group performed no better than chance and showed inconsistency in their
responses. However, Kavanaugh and Harris do not rule out the possibility that
children under 2 years understand pretend transformation in spite of their poor
performance in the experiments, due to methodological limitations (i.e., choice of the
picture depicting the pretend outcome). In fact, a recent study (Onishi & Baillargeon,
2005) demonstrated that 15 month old infants were able to predict an actor’s
behavior in both the true-belief and false-belief tests by using a nonverbal violation-
of-expectation method (e.g., they measured the infants’ looking times).
To summarize, although there is still a debate about the relationship between
pretend play and mental state knowledge (i.e., theory of mind), a developmental
transition in pretend play begins to emerge in children between ages two and three
(Fein, 1981). Besides, the positive relationship between mental state knowledge of
self and pretense has been reported among young children.
How to Play Pretense
A brief review of the literature in the previous section has demonstrated the
complex and notable abilities in pretend play among young children. What is more
amazing is that children show these abilities so naturally and intuitively at the very
beginning of childhood by joining pretend play. How do young children play

24
pretense even when they are not aware of these underlying abilities? Researchers
from various fields such as psycholinguistics, developmental psychology, and
cognitive psychology have proposed theoretical models in order to understand how
children deliberately distort reality by engaging in pretend play and resume their
non-pretend activities without being confused between the pretend world and the real
world (c.f., “representational abuse” in Leslie’s term). In the following sections, I
explain how children engage in and disengage from pretend play based on
metarepresentations (Leslie, 1987) and the multi-layering model (Clark, 1996).
Metarepresentations
Four categories of metarepresentations include “mental representations of
mental representations, mental representations of public representations, public
representations of mental representations, and public representations of public
representations” (Sperber, 2000, p. 3). Among the four categories, Leslie’s
metarepresentational theory of pretense focuses on the ability for mental
representations of mental representations (e.g., I think you believe that the ability to
play roles is important) to explain “properties of the internal mental representations
that underlie the external symbolic activity of pretending” (Leslie, 1987, p. 414).
According to Leslie, the primary representation reflects aspects of the real world
accurately and literally. Although pretend representations are very much related to
primary representations by the former referring to the later, pretend representations
are opaque in the sense that they are not transparent representations of the real world

25
but representations of representations, vis-à-vis metarepresentations. In other words,
metarepresentations are cognitively quarantined by decoupling the primary
representations in the real world from their normal input-output relations (c.f.,
psychological frames by Bateson, 1972). Decoupled expressions do not directly refer
to objects in the real world but are partially anchored to primary representations. In
this regard, only coherent information such as inference rules (e.g., pretending to
wash a doll by holding up an empty bucket) from children’s minds in the real world
is filtered into the implied selves in the pretend world semi-automatically, which, in
turn, enables elaborating pretense.
The most important and somewhat controversial component (e.g., Currie,
1998; Nichols & Stich, 2000) of Leslie’s theory is that metarepresentations are
marked in a special way in order to prevent representational abuse in pretend play.
That is, metarepresentations in pretend play entail an agent to decouple a different
semantic code from normal semantics in primary representations so that children can
coordinate two representations simultaneously without representational abuse of
actual semantic relations (see also Lillard, 1993). Pretend play in young children
involves the understanding of mental states so that the children can understand the
external representations as symbolic substitutions and the representations of other’s
mental representations. In Leslie’s theory, young children know that they pretend
during pretend play. How children engage in pretend play becomes more manifest
when the process is explained with the multi-layering model by Clark (1996).

26
Layers of joint actions
Clark (1996) proposes layers of joint actions in which the domains of joint
activities divide into multiple layers. In the first layer, children join pretend play in
reality. In the second layer, children initiate imagining actions as a joint pretense like
“a theatrical stage created on top of the first layer” (Clark, 1996, p. 355). Clark states
that further layers can be added as pretend play becomes more complex and that his
layering model can explain not only pretend play among children but also explain
how people communicate with virtual partners through computer mediated
communication (Clark, 1999).
In Clark’s model, two worlds (the real world & pretend world) are domains
of action. Each domain is characterized by “its participants, their roles, the place, the
time, the relevant features of the situation, the possible actions, and so on” (Clark,
1996; p. 355). For example, during pretend play, Bo's behavior is construed as an
action by Bo in Domain 1 (the first layer), and in Domain 2 (the second layer), as an
action by an implied Bo (a character that Bo is pretending to be). Domain 3 (the third
layer) can be added in which Bo’s behavior is construed as an action by the implied
Bo’s action (e.g., Bo is pretending to be a car racer [Domain 1] and describing his
previous racing experience [Domain 3: pretend experience within the pretend play]
of the implied Bo, a car racer [Domain 2], to a play partner; see Figure 2).

27

Figure 2. An example of multiple layers in pretend play

The two most important properties of his layering model are; (1) simultaneity
of multiple domains; and (2) asymmetric relation between joint actions in multiple
domains. All of them are related to how these layers are overlapped. First, multiple
domains are salient at the same time in pretend play. This notion is similar to the
argument that self-recognition is necessary to distinguish the pretend world from the
real world (see Lewis and Ramsay, 2004; and the previous section). The salient
multiple domains ironically make children enjoy pretend play without abusing
representations or simple reality confusion. That is, Bo’s beliefs and desires as well
as understanding of the world in reality can be applied to an implied Bo’s beliefs,
desires, and understanding if necessary, which helps Bo enjoy pretend play. It is
possible because higher layers are always embedded within lower layers so that the
real world is always the first layer of background in pretend play. It is important to
note that higher layers are not completely overlapped over lower layers so that there
Domain 3:
Pretend World
Domain 1:
Real World
Actual Play Partner
An Implied Experience by
an Implied Partner

Domain 2:
Pretend World
An Implied Play Partner:
A fellow car racer
Implied experience by
an Implied Bo

Actual Bo
An Implied Bo:
A car racer

28
is room for rules and beliefs in pretense to be bent or creatively interpreted without
violating these in the real world (see Figure 3). In other words, pretense is not just a
signified version of the real world but contains imagination (see Harris, 2000). This
architecture is slightly different from Clark’s layering model, but contains all the
properties of his layering and explains the uniqueness of pretend play as well.

Figure 3. Layering Architecture

The second property of asymmetric relation defines rules in multiple layers.
Both access and interruptions to multiple layers are asymmetrical. That is, actual Bo
in the first layer can interrupt activities of an implied Bo in the second layer to return
to the real world. But an implied Bo in the second layer cannot suspend the activities
of actual Bo in the real world, except for those who are mentally ill (The example of
the reverse interruption has been shown in a movie, the Matrix). Thus, this property
prevents children from having reality confusion.
Layer 1 (reality):
beliefs, desires,
knowledge,
rules, etc.
Layer 2 (pretense):
beliefs, desires,
knowledge, rules,
etc.

29
In addition to the above properties, I would include one more property of
event specificity. Each layer should be identified with regards to a specific event. For
instance, one cannot identify the number of layers involved in pretend play in
general because the number of layers changes continuously as events change in the
pretend play.
To sum up, Clark uses an analogy of layers to explain how people use
language, the concept of which is also useful to understand how people play pretense
and appreciate complex activities in pretend play without reality confusion. His
layering model helps understand the way young children enjoy pretend play
embedded within the background layer of their real world, and later helps develop a
new layering model to understand how people engage in role play in interactive
media, which is basically the same as the above described model.
In conclusion, researchers seem to agree as to the major features of pretend
play: (1) during pretend play, both the real world and pretend world are salient, with
the real world serving as a background layer by storing beliefs, desires, and
knowledge that can be inferred to the pretend world upon necessity; (2) there are
special cognitive mechanisms that help children to distinguish pretend
representations from primary representations and to be creative in pretend play at the
same time.

30
The Origin of Pretend Play and Other By-Products
In this section, I will take some time to discuss the origin of pretend play and
its relationship with the ability to understand other people’s minds, and finally with
the ability to play roles, because it is important to link pretend play, theory of mind,
and role play in order to explicate why people enjoy playing videogames and why
role play prevails in our daily activities.
Pretend Play and Theory of Mind
Theory of mind refers to the ability to understand other people’s beliefs and
desires (c.f., folk psychology), and thus to predict the behavior of others (Baron-
Cohen, 1997; Leslie, 1987). One commonality between pretend play and theory of
mind is the deployment of metarepresentations as Leslie (1987) looks at pretense as
the beginning of a capacity to understand the mental state of self and others. In other
words, pretense is an early manifestation of theory of mind. Similarly, Lillard (1993)
acknowledges many of the same skills between pretense and mental state
understanding (i.e., metarepresentations and theory of mind). In fact, Dias & Harris
(1990) conducted an empirical study to investigate theory of mind in early childhood.
Specifically, they examined deductive reasoning among children at four to six years
of age. The results indicate that children reason better about counterfactual
syllogisms in the pretend situation than in the real situation, suggesting that children
may decouple representations from referents in the pretend world (see Lillard, 1993
for an alternative explanation of acting as-if).

31
Another important piece of evidence for relating pretense to theory of mind is
the false-belief task with autistic children. The purpose of the false-belief test is to
see if children can understand that someone might hold a different (false) belief. In
the Sally and Anne test (see Baron-Cohen, 1997), Sally puts a marble in one place
and goes out. And Anne comes in and changes the location of the marble and goes
out. Then, children are asked where Sally would look for her marble when she enters
again. Normal children with the age of three tend to fail the false-belief test, whereas
most normal children by the age of four seem to succeed (Baron-Cohen, 1997; Leslie,
1987; Lillard, 1993). The false-belief test is often used to diagnose autism because
three cardinal symptoms in autism are severe impairment in social development, in
communication development, and in pretend play (Baron-Cohen, 1997). In fact,
children with autism hardly pass the false-belief test, which suggests a genuine
inability to understand other people’s different beliefs among autistics (Baron-Cohen,
1997; Pinker, 1997). In Baron-Cohen’s term, autistic children are mind-blind
because their ability to attribute minds to others is damaged. Thus, autistic children
do not seem to have both theory of mind (a more advanced form of representational
mental states) and pretense (a primitive form of metarepresentations), which suggests
the link between pretense and theory of mind (Leslie, 1987).
On the other hand, some researchers argue that there is no substantive
evidence that young children understand metarepresentations or mental states of
others in pretend play (e.g., Currie, 1998; Kavanaugh & Harris, 1994; Lillard, 1993).

32
For instance, Kavanaugh and Harris (1994) tested 12 children with autism to
investigate their ability to understand the pretend actions of a play partner. The
results show that the children with autism perform better than chance in selecting a
picture that depicts the pretend outcome. However, potential limitations in their
experimental design might have lead to rather contrasting results. They tested some
level of metarepresentational abilities of children with autism, which is close to
mental state knowledge of self (e.g., I know that you are pretending). It is different
from mental state knowledge of others, which is false belief (e.g., I know that you do
not know that I know). More importantly, the children with autism were not asked to
predict different belief (false belief) in another person, but asked to predict the
pretend outcome enacted by a partner’s pretend play. In order to redesign the test as
a false-belief test, a confederate needs to sit next to the children with autism. The
confederate’s eyes should be covered during the transformation or pretend enactment.
Then, the children with autism should be asked to guess which picture the
confederate would select. In fact, Kavanaugh and Harris did not fully reject the
relationship between pretend play and theory of mind. In stead, they argue that
autism may limit some of the components in pretend play such as producing
symbolic substitutions by themselves. When children with autism receive
appropriate support from a play partner, they may understand the pretend partner,
which is the aim of therapeutic interventions for children with autism. Indeed, all the
participants were attending special education units at the time of the experiment.

33
One of the alternative explanations against metarepresentations in pretend
play is behavioral understanding of the action component in pretend play (Lillard,
1993; Nichols & Stich, 2000). For example, symbolization is the representing of an
object by something other than itself. To deploy metarepresentations in pretend play,
Bo must represent himself representing a block as a cookie for representational
component of pretense. On the other hand, Bo simply needs to know that he is acting
on a block as if it were a cookie for the hypothetical substitution (see Lillard, 1993
for detail discussion on acting as-if). An alternative argument of behavioral
understanding seems to be plausible. However, psychological understanding is likely
to be nested within behavioral understanding as we cannot separate the subjective
mental world from the objective physical world (c.f., criticism on Cartesian dualism;
see Dreyfus, 1992). Although the behavioral component of pretend play seems to
emerge earlier than the mental component of pretend play, it may not be very
meaningful to separate the two components. Without the action component, pretend
play becomes a pure hallucination. Without the mental component, pretend play
becomes a sensorimotor practice. Besides, empirical evidence to support this
explanation is not robust either. The nature of experimental design might be too
complicated for children to understand. Thus, behavioral understanding still remains
as an alternative explanation.
To summarize, in spite of the on-going debate, researchers agree that (1) by
the age of four to five, normal children show the ability to understand mental states

34
of others (metarepresentation), thus understand theory of mind; (2) there are many
commonalities in both pretend play and theory of mind; and (3) there is some
evidence that shows mental states of self in children around the age, of two although
some researchers believe that these studies are not robust enough. Taken together, I
believe that pretend play is an early manifestation of theory of mind because (1) the
two worlds (i.e., the real world and pretend world) should be salient simultaneously,
which implies children’s awareness of their pretense during pretend play in spite of
the fact that they cannot describe it; (2) there are many indirect indications (both
theoretical and empirical) that provide support for the relationship (e.g., Baron-
Cohen, 1997, Leslie, 1987; Lewis & Ramsay, 2004; Lillard, 1993); (3) and yet there
is little evidence that can fully falsify the link between pretend play and theory of
mind. This perspective also complements the innate idea of theory of mind and
pretend play that will be discussed in the following section (see Perner & Ruffman,
2005 for the evolutionary basis arguments for theory of mind).
Evolutionary Nature of Pretend Play
Evolutionary psychology posits that the mind is a set of functionally distinct
domain-specific modules adapted by natural selection in order to solve the kinds of
problems our ancestors faced throughout our history (Pinker, 1997; see also
Cosmides and Tooby, 1994; Lee, 2004b for more detailed discussion about domain-
specificity). According to Lee and Jung (2005), domain-specific modules refer to
localized brain components that govern specific human responses to special

35
situations. For instance, when a specific situation occurs, the corresponding module
may become salient and take charge of our entire cognitive process and behavioral
responses temporarily (see Lee & Jung, 2005 for the cognitive hijacking hypothesis).
However, innate modules are incomplete when we are born. These innate modules
become complete and organized into higher level systems during development via
various forms of experience, including pretend play (Tooby and Cosmides, 1994).
Due to the methodological limitations, only a few modules have been
discovered such as the amygdala for emotional responses (Goleman 1995), a
language module for speech perception and production (Carroll 1999), or the theory
of mind module for social interaction (Baron-Cohen 1996; Buss, 2004). The
evolutionary nature of theory of mind has been discussed in various studies (e.g.
Baron-Cohen, 1996, Buss, 2004; Costall & Lendar, 2004; Leslie, 1987; Lillard, 1993;
Pinker, 1997; 2002). In brief, theory of mind—the ability to understand other
people—has known to provide significant advantages to solve the everyday survival
and reproduction problems that our ancestors faced throughout generations. Theory
of mind implies a component of cooperativeness or reciprocity through perspective
taking, social competence, social rituals and interaction (Lillard, 1993). Indeed,
reciprocal altruism that provides benefits to nonrelatives with expectation for future
benefits from them has been identified as the strongest weapon for human survival.
For instance, tit for tat was the winning strategy in iterated prisoner’s dilemma
games among many other strategies (Axelrod & Hamilton, 1981). The main features

36
of tit for tat are (1) reciprocate first to your partners; (2) retaliate only after the other
has defected; and (3) forgive the previously defecting partners as soon as they start to
cooperate. In order to reciprocate first for mutual benefits, people need to understand
or represent other people’s beliefs and desires, which is exactly in line with
metarepresentations and theory of mind.
Having said that, pretend play may be an innate module for social
interactions as well because pretense is hypothesized to be an early manifestation of
theory of mind. In fact, pretend play, especially joint pretend play, is a cross-
culturally universal and species-typical phenomenon that emerges at the beginning of
childhood. Although Steen and Owns (2001) state that animal chase play resembles
the elementary form of pretense, animal chase play is structurally different from
chase play in children. Animal chase play does not involve any object transformation,
the lowest level of pretend play (see Jackowitz & Watson, 1980). On the other hand,
chase play in children develops into complex forms involving object transformation
and narratives. Similarly, the deception activities among great apes and their ability
to recognize themselves in mirrors are believed not to be the evidence for animal
theory of mind but to be the results of complex reinforcement schedules and
kinesthetic self, respectively (Ohler & Nieding, in press).
The rapid emergence of metarepresentations in pretend play also implies a
major developmental discontinuity, which indirectly suggests the specific innate
module for our commonsense theory of mind and social interactions accordingly

37
(Leslie, 1987). Steen and Owen describe pretense as an evolution’s pedagogy to train
domain-specific skills such as predator-evasion and fighting. In addition, I propose
that pretend play is an evolutionarily adapted pedagogy to train skills for social
interactions such as reciprocity. In fact, joint pretend play is all about reciprocity.
When Bo sees his play partner’s pretend enactment, Bo has to reciprocate it by
corresponding pretend responses (e.g., Bo pretends to drink tea from an empty cup in
a pretend tea party invited by his play partner). Otherwise, the joint pretend play
cannot be performed. Therefore, children and adults intuitively regard pretend play
as an intrinsically rewarding activity because of its strong advantages for survival,
thus making people feel entertained, excited, and satisfied during the pedagogy of
pretend play.
Pretend Play and Beyond: Video Game Experience
A brief review of the literature and discussion in previous sections indicates
the innate abilities of understanding mental state knowledge (metarepresentations) in
early pretend play that pedagogically develop further to the abilities to understand
other people (i.e., theory of mind). Based on what have been discussed previously, I
would like to add a final piece to the puzzle of pretend play, that is, its relationship
with role play and our enjoyment of interactive media, mainly focusing on
videogame experience.
When people play videogames they play the role of a character in the
videogames by controlling the character regardless of the videogame genre. Even in

38
a first-person shooting game, a user plays the role of the character, to kill as many
enemies as possible. Construing videogame play as role play becomes clearer when
we look at the definition of role play. Role play is defined as “behavior in which
people simulate the identity of another person” (Fein, 1981; p. 1101). Thus, when
people play videogames, they simulate the role identity of a videogame character.
For example, when Bo plays a fighting game with his friend, Bo represents the role
of an implied Bo so that the implied Bo fights against an implied friend on the screen
who is controlled by his friend in reality. This is exactly how pretend play works (see
Figure 2). There is more striking evidence for the resemblance of pretend play and
videogame play. For instance, the ability to sustain role play increases between 4 and
5 years, which coincides with the emerging period of theory of mind (Fein, 1981).
Rubin (1976) shows a positive relationship between pretend play and performance
on role-taking tasks (see also Lillard, 1993). Similarly, the positive relationships
between pretend play and transformational thinking, social rituals, reduction in
egocentricity, perspective taking, and cooperative social problem solving has been
identified (see Fein 1981 for detailed reviews).
Taken together, I propose that videogame play is role play, an advanced form
of pretend play. The major difference is that video game play, enjoyment of
entertainment and interactive media in general, is a by-product of the innate
pedagogical module for social interactions (i.e., pretend play). It is a by-product of
adaptation rather than a direct product because videogame play does not yield direct

39
advantages to our survival. Videogame play seems to have direct advantages because
of its almost identical features to pretend play. Therefore, the ability to enjoy playing
roles in videogames should be understood with regards to the abilities to play and
enjoy pretense.
In a similar vein, Media Equation (Reeves & Nass, 1996) can be understood
with regard to pretend play. The term “media equation” simply means "media equal
real life” (Reeves & Nass, 1996, p. 5)." In detail, it implies that an "individual's
interactions with computers, television, and new media are fundamentally social and
natural, just like any interaction in real life” (p. 5). Thus, human-computer
interaction is like people engaging in pretend play with computer agents. People
know that it is a computer agent, but behave as if it were a social actor by decoupling
semantics of the functions from the form of the computer. Dual identities of the
computer or computer agent (a social actor and machine) are salient at the same time
and yet do not cause representational abuse because of our abilities of mental state
knowledge as seen in young children’s pretense. One of the arguments against Media
Equation and Computers Are Social Actors (CASA) paradigm is that people may
think of other people represented through computers when they apply social
stereotypes to computers. However, studies have demonstrated that participants
showed social responses to computers even when they clearly acknowledged that
there were no hidden programmers or co-players behind the computer (Lee, Jung, &
Nass, 2007; Nass, Moon, & Green, 1997), which supports the pretend play argument:

40
people precisely know that they are computer agents, but play pretense as if
computer agents were social actors.
Multi Layers of Conjoint Action in Videogame Play
All in all, Figure 4 explains the process of videogame play based on a new
multi-layering model for role play. This is very similar to how children engage in
joint pretend play as discussed. An actual player understands the representations of
an implied agent decoupled from an actual computer by creating a higher layer of an
implied self. Because the interaction occurs in a higher level of the decoupled layers,
identification, sympathy, or empathy could be established between an implied person
A and an implied computer agent B. It is worth note that identification does not
occur between an actual player A and an implied agent B or between an actual player
A and an actual computer B.

Figure 4. Multi-layering model of role play in the context of interactive videogames

A player could represent or understand intentions of an implied computer
agent B through its representations (e.g., an implied agent B fires a gun shot toward
An Implied Person A

Actual Person A Actual Computer B
An Implied Agent B

Interactions
Inference Rules: Beliefs,
desires, knowledge, etc.
Decoupling

41
an implied person A). A truth that an implied computer agent B cannot possess any
intention or mental states does mot matter in this specific videogame situation due to
the decoupling. The semantics of the computer agent is replaced by the semantics of
the other human being in the way that children often do in their pretend play.
Conclusions
In this chapter I proposed a multi-layering model to understand how people
engage in role play in the context of videogames and why playing videogames might
be enjoyable and entertaining, based on the literature of pretend play, theory of mind,
evolutionary psychology, and multi-layering model in language use. Because of the
similarities, the major characteristics of pretend play could be applied to the
understanding of our experience in interactive media, as a function of role play.
Finally, future studies need to look at differences between pretend play and
videogame play. We seldom observe reality confusion in children’s pretend play.
However, reality confusion from playing videogames has been often reported. By
focusing on the differences, we may be able to provide ways to reduce the potential
harmful effects of playing videogames such as becoming aggressive after playing
violent videogames. For example, in Chapter 4, I will report on results from an
empirical study that examines the effects of altering contextual factors (i.e., points of
view) on players’ self-concept and behaviors in interactive media.

42
Chapter 3: Revisiting Entertainment Theories from a Role-Play
Perspective
In Chapter 2, I discussed the evolutionary nature of role play by comparing it
with children’s pretend play, and conceptualized a multi-layering model of role play
to explain how people play multiple roles by creating perceptual layers. In this
chapter, I elaborate on theoretical implications of the model with a particular focus
on videogame play.
It may be worthwhile to return to the multi-layering model of role play
discussed in the previous chapter. In order to understand the entertainment
experience, it is important to understand what is happening to us during our
experience. For instance, when Ian reads a book, Ian opens up a book and starts to
read. This is what is happening at the background level of any entertainment
experience where actual Ian experiences actual objects (e.g., people go to theaters
and watch motion pictures, turn on and play videogames, or go to concerts and listen
to music). I describe this background level as the real world. On the other hand, as
Ian starts to read a book, he portrays the texts in his minds at the perceptual level
where an implied Ian experiences interactions of implied protagonists and
antagonists. I describe this perceptual level as the entertainment world. These two
worlds become salient simultaneously when we experience the entertainment product
(e.g., actual Ian imagines that an implied Ian watches an implied pirate hiding
implied gold when actual Ian reads a sentence from a book in reality). The way we

43
coordinate the two worlds (e.g. making one world become more salient than the
other at a given moment) is similar to how children play pretense (see Chapter 2 for
detailed discussions about pretend play). Because of shared properties such as beliefs,
desires, or emotions between the two salient worlds, we are, ironically, able to enjoy
the entertainment product and to distinguish entertainment experience from actual
experience in reality at the same time (see Landy, 1993 for the dramatic paradox).
Thus, a theory for entertainment experience needs to explain how the two worlds
influence each other and consequently affect our overall experience of the
entertainment product. However, most current theories that explain entertainment
experience focus on our virtual experience in one dimension, either in the
entertainment world or the real world. In the following sections, I review seven
theories that explain our entertainment experience and discuss how to reinterpret
these theories based on the role-play perspective.
Feeling Within or Feeling With
The point-of-view in the entertainment product is important in determining
the audience’s position with regard to either spectators or identification (Oatley,
1999). The use of the third-person view may influence the audience to take the
spectator position (i.e., feeling with the protagonist), whereas the use of the first-
person view may favor identification (i.e., feeling within the protagonist). This
positioning of the audience during the entertainment experience is a continuous
spectrum running from complete observation to complete identification. And the

44
audience has the ability to maintain the optimal aesthetic distance on the spectrum
unconsciously in order to enjoy the entertainment product as much as possible
(Oately, 1999). That is, a person can be a spectator at one moment, and then can
identify with a specific character at the next moment freely during the person’s
entertainment experience, if not take both positions simultaneously. Nonetheless, a
person is likely to take a specific position in the aesthetic distance for maximum
enjoyment of the entertainment product at the given moment when we look at the
cross sectional slice of the whole entertainment experience. In this section, I review
four theories—identification, sympathy, empathy, and parasocial interaction—that
explain our entertainment experience at a specific moment with regard to the
different points of view (see Figure 5).

Figure 5. Cross sectional view of the entertainment experience
Theories for a Specific Experience at a
Cross Sectional Moment of the Event
• Identification
• Sympathy
• Empathy
• Para-Social Interaction
Entertainment
Experience

45
Identification
According to Oately (1994), a theory of identification is based on Aristotle’s
concept of mimesis that he interprets as simulation rather than imitation or
representation. Based on the concept of mimesis, Oately (1994; 1999) defines the
entertainment experience as a simulation of human actions that run on the minds of
the audience (see also Freud (1985 [1904]) for the argument that identification is the
assimilation of the representations of other people into one’s personality). That is,
when a person experiences the entertainment product (e.g., the simulation), the
person understands the representations of a fictional character (e.g., an implied
actor/actress) within the existing beliefs, desires, knowledge, and emotions of herself
by taking on the beliefs, desires, or goals (e.g., personas) of the implied actor/actress.
During the successful simulation, the person feels the emotions of the implied
actor/actress, the process of which is defined as identification and later described as
meeting of minds (Oately, 1994; 1999). This process is almost identical to the multi-
layering model of role play. It is important to note that these emotions are not exactly
the same as the emotions of the implied actor/actress because these emotions are the
audience’s own. Thus, the most important modality in the process of the simulation
is that the entertainment product is experienced within the minds of the audience in
the real world, not vice versa. Indeed, the appreciation of the entertainment product
takes place in our minds, not on the screens. Nonetheless, this important prerequisite
for the entertainment experience is often neglected in some entertainment theories or

46
when the concepts are operationalized in empirical research. I will continuously
discuss this issue throughout this chapter.
In particular, Oatley (1994) proposes a taxonomic tree of emotions, including
emotions of identification that occurs during the entertainment experience. He uses
an analogy of a semi-permeable membrane to distinguish the entertainment world
from the real world. A semi-permeable membrane works in a way to filter some of
characteristics of the real world into the entertainment world of necessity during the
entertainment experience (see Goffman, 1961). To put it differently, people semi-
automatically filter out incompatible information from the real world in order to
optimize their enjoyment of the entertainment product (c.f., a coherence theory of
truth in Oatley’s term [1999]). Thus, the internal filtering process (i.e., semi-
permeable membrane) helps the audience understand the representations of the
implied actor/actress in the entertainment product (c.f., cognitive psychologists use
different analogies such as the Updater Box in the Possible World Box [Nichols &
Stich, 2000] or decoupling [Leslie, 1987]). According to Oatley (1994; 1999),
identification is one of the emotions that occurs inside the permeable membrane. To
explain this within the multi-layering model of role play, the audience identifies with
a protagonist or with any implied actor/actress through whom the audience becomes
the implied actor/actress rather than merely sympathizing with the persona (see
Figure 2).

47
On the other hand, Cohen (in press) defines identification as “an imaginative
process that is evoked as a response to characters presented in mediated texts.” More
specifically, he describes the process of identification as follows: a person feels a
strong affinity toward an implied actor/actress (e.g., the protagonist) so that the
person develops empathetic feelings for the experiences, desires, and goals of the
protagonist. Thus, Cohen’s notion of identification is slightly different from that of
Oatley’s. It is a broader definition in which the distinction among identification,
sympathy, and empathy becomes blurry. Although a broad definition is necessary,
we may want to have a narrow definition of identification in order to understand a
specific entertainment experience at a given moment. In addition, feeling affinity
may be unnecessary for identification since we sometimes identify with the
antagonists without liking them (see Ang, 1985).
Cohen also emphasizes that identification occurs through the filters of the
audience’s own understanding and experience. However, his explanation for the way
the filter works is not compatible with the process of identification by Oatley and
even with his own. For instance, Cohen (in press) states that when a person identifies
with the protagonist in the entertainment product, the person simply suspends her
identities in the real world temporarily. When a person suspends herself in the real
world at any given moment during the person’s experience of the entertainment
product, then the person also suspends all of his/her beliefs, desires, knowledge, and
emotions embedded within the person’s identities in the real world that are

48
fundamental components to appreciate simulations in the entertainment world. In
other words, a person cannot understand or enjoy the entertainment product without
inferring the coherent information in the real world just as one cannot suspend or
leave the real world even for one second. How then can a person identify with an
implied actor/actress during the person’s entertainment experience? In stead of the
real world’s being suspended, the entertainment world becomes more salient. It is
important to note that both of the worlds become salient simultaneously, but one
becomes more salient than the other depending on the degree of identification
1
(see
also the dual salience hypothesis in the previous chapter). The entertainment world
becomes more salient when a person can understand and act on the beliefs, desires,
or goals of the protagonist by inferring coherent information within the person’s
mind in the real world. By coherent, I mean that incompatible information
potentially disrupting the optimal entertainment experience is filtered out through the
cognitive mechanism or permeable membrane (in Oatley’s term) as needed (see
Oatley, 1999). Through the filtering process (both in and out), a person can be
engrossed enough to enjoy the creative entertainment beyond the person’s own real
world (see Figure 6).

1
The followings are reinterpretations of pretend play with regards to our experience in interactive
media. First, a player puts representations of the implied character from the interactive media into the
player’s minds in the real world. In this way, the player can make both the entertainment and real
world become salient, which is very important for the enjoyment of interactive media. For example,
children know that they are engaging in pretend play so that they can enjoy pretend play without any
reality confusion during and after the play. In a similar vein, a player needs to know killing people is
not real in violent videogames in order to enjoy playing them. But violent videogames should be
realistic enough to provide exciting and thrilling experience. Thus, both entertainment world and real
world should be salient at the same time where the real world serves as a background layer.

49

Figure 6. Identification in the multi-layering model of role play

Finally, identification has been criticized for its limitations in explaining our
experience of motion pictures or theatrical performances because of, I propose, the
predominant third person view or God’s eye view in such entertainment products.
For instance, Zillmann (1994) points out children’s taking spectator positions when
they interact with the characters in the puppet theater, Kasperletheater. Gliech (1997)
also shows that a majority of German adults identifies with the opposite sex
characters in a television program. Similarly, Press (1990) indicates that working
class women identify with upper class characters on a television show. Ang (1985)
also reports that many people identify with the antagonist (J. R. Ewing) in spite of
their disliking the character on the television program Dallas. If people become the
implied actor/actress through identification, we can never explain such findings
Aesthetic distance management:
From identification to sympathy
An Implied Actor B

Interaction:
Identification
Decoupling
Association
An Implied Person A

Actual Person A Actual Actor B
Filtering in/out Coherent
Information

50
(these limitations of identification will be revisited when I introduce a new
theoretical model later in the current paper).
Nonetheless, the notion of identification seems to regain its momentum along
with our rapidly increasing experience with new interactive media such as
videogames. Videogame players seem to identify with the characters (implied
players) because players have an illusion of control over their characters and overall
game experience
2
. Thus, the interactivity and control of the interactive entertainment
product may favor the perception of the first-person view vis-à-vis identification. In
addition, reverse identification can occur in videogame play. For instance, let’s
imagine that you play a fighting videogame against your friend. When you win, you
say you have defeated your friend. You do not say that your character has defeated
your friend’s character, nor you say you have defeated your friend’s character. In this
scenario, not only do you identify with your own character in the videogame, but
also you identify your friend’s character with your friend. The later seems to be
reverse identification that we also see in traditional non-interactive media (e.g., in
parasocial interaction, the audience identifies the implied celebrities with the actual
celebrities).
To summarize, a theory of identification explains a specific entertainment
experience, feeling within, at a given moment. In spite of its potential for theoretical

2
I use the term, illusion, because videogame characters’ movements as well as the outcomes from the
movements are already programmed, are thus limited by the producers. In this regard, videogame
players only have an illusion of control that has already been created to work in that way.

51
explication, many empirical studies do not seriously explore it (Cohen, in press),
which may result in the current limitations of identification theory.
Sympathy
Sympathy refers to feeling with (Oately, 1994). The process of sympathy is
exactly the same as that of identification. The only difference is the perceptual point-
of-view. That is, in sympathy, the audience maintains aesthetic distance as spectators
and feels with the implied actor/actress in the entertainment product. A person
becomes sympathetic to one character, then another, to extend the person’s
emotional understanding of people in the entertainment product (Oately, 1994).
According to Tan (1994), the audience’s emotional response is mainly one of
sympathy with the implied actor/actress because the traditional entertainment
products such as books or motion pictures naturally put the audience in the spectator
position as a witness of a scene. Given the process of maintaining the optimal
aesthetic distance, distinguishing sympathy from identification may not provide
much meaningful understanding of the overall entertainment experience (e.g., a
person may experience a great deal of identification and sympathy by changing the
aesthetic distance continuously to maximize the person’s enjoyment during the
entertainment experience). However, it is important to have distinct definitions in
order to understand the audience’s experience of a specific scene at a given moment
during the entertainment experience (i.e., a cross sectional understanding).

52
Empathy
Empathy often refers to feeling with or feeling for another individual,
mediated by the ability of people to put themselves into others’ emotional
experiences (see Zillmann, 1994; in press for various definitions). The concept of
empathy is more inclusive, compared to that of identification and sympathy as
defined by Zillmann (1994): empathy is “any experience that is a response (1) to
information about circumstances presumed to cause acute emotions in another
individual and/or (2) to the bodily, facial, paralinguistic, and linguistic expression of
emotional experiences by another individual and/or (3) to another individual’s
actions that are presumed to be precipitated by acute emotional experiences, this
response being (4) associated with an appreciable increase in excitation and (5)
construed by respondents as feeling with or feeling for another individual” (p. 40).
Thus, empathy is emotional responses to any emotional representations of the
implied actor/actress in the entertainment product that may result in a great deal of
affinity. In this sense, empathy is similar to sympathy or parasocial interaction
(Vorderer, Klimmt, Ritterfeld, 2004) although a specific position of aesthetic
distance is not considered in the theories of empathy.
There are three different theories that explain how we experience empathy
(Zillmann, 1994; in press). First, motor-mimicry posits that our innate dispositional
mimicking behaviors foster empathic affect (e.g., mimicking a variety of facial
expressions of the protagonist; c.f., see Nichols & Stich [2000] for behavioral

53
understanding of pretense). Although this innate reflexive empathy is plausible, it
alone cannot fully explain our entertainment experience since we do not copy every
single facial expression of the characters during the experience of empathy.
Secondly, affective disposition theory posits that affective dispositions
toward the characters in entertainment products are formed based on moral judgment
of the outcomes. That is, when the outcomes reflect dichotomous moral judgment
(e.g., positive when the protagonist receives a fortune, or negative when the
antagonist receives a fortune), the audience feels strong empathic pleasure or distress,
respectively. In other words, the audience develops its own expectations about the
entertainment product. In order to feel empathy, the portrayal of the entertainment
product should match the expectation of the audience.
Thirdly, the cognitive model posits that people put themselves into another’s
place not to identify with the protagonist but to understand the emotions of the
protagonists by taking perspectives or roles. This type of empathy has clinical
applications such as psychotherapy (Fein, 1992). Although the cognitive empathy
requires constant cognitive efforts to take roles, people may apply cognitive
maneuvers quasi-automatically in order to engage in perspective taking frequently
and yet effortlessly as discussed in Chapter 2 (see also Feshbach, 1978). This notion
of empathy is very similar to that of identification because taking roles means taking
on beliefs, desires, emotions, and goals embedded within the roles of the implied
actor/actress.

54
To sum up, empathy explains more inclusive experience of the audience in
that the audience feels with the emotional representations of the implied actor/actress
in the entertainment product. As such, there are three different approaches to explain
how empathy may work: reflexive, affective, and cognitive models. However,
empathy may not be the best construct to explain a specific entertainment experience
at a given moment (i.e., a cross sectional experience) because of the inclusive nature
of its theories. In addition, empathy theories do not consider the audience’s relative
position in aesthetic distance during the entertainment experience, which may be
related to potential limitations of the theories. For instance, counter-empathy occurs
when the audience feels happy for the pain of the implied actor/actress. This
discordant affective experience cannot be construed as feeling with or feeling for,
and thus cannot be considered as an empathic reaction (Zillmann, in press). However,
if the audience takes a complete spectator position at the very moment (c.f., similar
to the zoom-out function in a camera), the audience could infer the emotional
responses of another implied actor/actress who may not be seen at the particular
scene. Thus, the audience may identify with, sympathize with, or show empathic
responses to the emotional representations of another implied actor/actress (e.g.,
implied person C) although the emotional responses of the audience seem to aim
directly at the pain of the implied actor/actress (e.g., implied person B). Another
potential limitation is related to individual difference in moral judgment. In affective
dispositional empathy, it is assumed that people like for the protagonist to succeed

55
and the protagonist to suffer. This assumption may not be true for someone who
simply does not want to see anyone to suffer even if it is the antagonist. For example,
a person may not want to see any punishment of the antagonist not because the
person feels for or feels with the antagonist but because the person cannot take any
more violence or misfortune in the entertainment product. In this case, the person
may disengage from the entertainment product even though the person likes the
protagonist and dislikes the antagonist because the person’s moral judgment, more
specifically the person’s expectations for the entertainment product is different from
the basic assumption. Having said that, the relationship between the outcome of the
entertainment product and the moral judgment of the audience in the empathy theory
is a very important component in explaining entertainment experience but needs to
be elaborated further.
Parasocial Interaction
Parasocial interaction refers to the audience’s ability to relate to the implied
actor/actress in the entertainment product (Vorderer, Klimmt, Ritterfeld, 2004).
Horton and Wohl (1956) introduced the concept of parasocial interaction to explain
how people interact with the implied actor/actress in TV as if the implied
actor/actress were real (c.f., see Reeves & Nass, 1996 for Media Equation). The
interaction is labeled as parasocial because the implied self interacts with the implied
actor/actress in the entertainment world without actual interactions in the real world.
The result of this interaction is described as a parasocial relationship in which the

56
audience is believed to develop a sense of intimacy—a rapport—with the implied
actor/actress or even with the animated characters (Basil, 1996; Klimmt, &
Hartmann, in press). Thus, the point-of-view in parasocial interaction is similar to
that of sympathy and empathy in that the audience takes a spectator position in the
aesthetic distance. In particular, developing an affinity towards the implied
actor/actress is similar to empathy, but identifying the implied actor/actress with the
actual actor/actress is similar to the reverse identification.
In spite of the difference in theoretical definitions between parasocial
interaction and identification, studies often use the two terms interchangeably (e.g.,
Basil, 1996; Cohen, in press; Klimmt & Hartmann, in press). In fact, Cohen (in press)
states that there is no convincing evidence to distinguish identification from
parasocial interaction. However, the operationalization or measure of parasocial
interaction in empirical studies is based on the spectator position of the audience in
which the audience feels for or feels about the implied actor/actress (see Rubin,
Perse, & Powell, 1985 for the parasocial interaction scale).
Taken together, parasocial interaction also explains a specific entertainment
experience at a given moment, which is similar to empathy. When repeated, the cross
sectional experience of parasocial interaction may result in the parasocial
relationship in that the audience is believed to develop a rapport with the actual
actor/actress in the real world.

57
Feeling Overall
A brief review of the literature about the entertainment experience in the
previous section has demonstrated that (1) the audience puts the representations of
the implied character from the entertainment world into their own minds in the real
world; (2) only coherent information from the audience’s minds in the real world is
filtered into the implied self in the entertainment world semi-automatically; (3) the
audience can have a specific entertainment experience in accordance with the
aesthetic distance, ranging from identification to parasocial interaction, at a cross
sectional moment of the audience’s entire experience; and (4) the congruence
between the outcomes and expectations of the entertainment product may facilitate
these experiences.
Identification, sympathy, empathy, and parasocial interaction are all event-
specific experiences that explain the audience’s psychological responses to the
representations of the implied actor/actress at a given moment. Thus, these theories
cannot be used to explain the overall entertainment experience. The theories that
better explain entertainment experience are flow, transportation, and presence (see
Figure 7). The three theories also help us understand the states of the audience during
the entertainment experience. I review these three theories in this section.

58

Figure 7. Block view of the entertainment experience

Flow
Flow refers to an enjoyable entertainment experience in which the audience
loses self-consciousness by intensely focusing on the entertainment product (Sherry,
2004). Flow is often observed from artists when they continue to work on their
creative arts and seem to cut off from the real world (Csikszentmihalyi, 1993),
although it is likely that the work process of artists is bringing the creative arts into
the real world (e.g., the real world is a background layer). Thus, flow is understood
as an immersed state into the entertainment product so that the entertainment world
becomes extremely salient relative to the real world. Given the above descriptions of
the flow state, flow explains a block of entertainment experience, if not the whole,
rather than a cross sectional experience at a given moment.
Experience of
Entertainment
Theories for Overall Experience during the Event
• Flow
• Transportation
• Presence

59
In order to create the flow state, flow theory emphasizes, there must be a
balance between the entertainment content and the audience’s ability to understand
the content (Sherry, 2004). For instance, the audience feels bored if the entertainment
product is too easy (e.g., when an adult watches Sesame Street), whereas the
audience feels anxiety if the entertainment product is too difficult (e.g., when a
senior plays a videogame for the first time). This requisite balance for the flow state
is similar to the matching between the outcome of the entertainment product—
difficulty—and the moral judgment of the audience—skills—for empathy. It is also
related to the requirements of both of the entertainment world and the real world.
While skills are requisites in the real world, difficulties are requisites in the
entertainment world. Once again, both of the two worlds should be salient in order to
coordinate a balance between skills of the audience and difficulties of the
entertainment product.
Finally, flow is often used to understand our experience of videogames
because activities known to facilitate the flow state are similar to the characteristics
of videogames such as concrete goals, clear rules, clear instruction, or adjustable
difficulties (Csikszentmihalyi, 1993; and Sherry, 2004).
Transportation
Transportation refers to an enjoyable entertainment experience in that the
audience is taken into the entertainment world vis-à-vis “transportation into a
narrative world” (Green, Brock, & Kaufman, 2004; p. 312). Indeed, transformation is

60
conceptually very close to flow because both theories elucidate the extremely
immersed states during entertainment experience.
Green, Brock, & Kaufman (2004) state that the audience enjoys suspense or
risks in the entertainment product because transportation enables the audience to
experience endangered events without being exposed to real dangers (see also Lee &
Jung, 2005). The risks are unraveled at the end either within the story itself or by the
audience’s disengaging from the entertainment product. However, transportation
does not explain why some people enjoy Holocaust stories (i.e., dark stories without
resolution), whereas some do not. Nonetheless, the positive relationship between
enjoyment and transportation has been reported with respect to escapism
(Moskalenko & Heine, 2003), role-play opportunities (Markus & Nurius, 1986), and
satisfying needs-to-belong through identification, empathy, and parasocial
interaction with the implied actor/actress (Green & Brock, 2000).
The potential limitation of transportation theory is the way it describes the
entertainment experience. Researchers seem to posit transportation into the
entertainment world because the audience is immersed in the entertainment world
and experiences somehow a different self during the state of transportation (Green,
Brock, & Kaufman, 2004). However, as discussed, the audience can be engrossed in
a fantastic entertainment world through filtering out incompatible information and
filtering in coherent information from the real world. So, the state of transportation
should be described as transportation (of the entertainment world) into (the

61
audience’s minds in) the real world. It is simply impossible to transport the actual
self into the entertainment world even temporarily or to transport an implied self
separated from the actual self (see Dreyfus, 1992 for criticisms against the Cartesian
dualism that assumes the independence of the subjective mental world from the
objective physical world). Besides, some entertainment experience (e.g., self
presence) cannot be explained if we consider the entertainment world as a
background domain of our entertainment experience (see discussion in the next
section).
Presence
Probably the most simple and yet comprehensive definition of presence is
“the perceptual illusion of nonmediation” (Lombard, Reich, Grabe, Bracken, &
Ditton, 2000, p.77). The definition states that the audience would respond to the
implied actor/actress as if the medium or entertainment products were not there.
Thus, this notion of presence is consistent with the basic component of the multi-
layering model of role play in that the entertainment world is experienced within our
minds in the real world. Lee (2004a) further develops the concept of presence and
defines it as “a psychological state in which virtual (para-authentic or artificial)
objects are experienced as actual objects in either sensory or nonsensory ways” (p.
37). Lee’s definition reflects two important characteristics of presence: (1) presence
is a psychological state similar to flow and transportation; (2) the representations of
an implied actor/actress is experienced by the implied audience, and then the

62
experience of the implied audience is transported into the audience’s minds in the
real world as if the experience were real (c.f., the experience of the implied self is a
representation of representations [metarepresentations]; see Leslie, 1987 and Chapter
2). As shown by the above definitions, presence explains our overall entertainment
experience very well and accurately although there are various definitions of
presence some of which do not fit in the hypothesis of the real world being a
background level (e.g. “being there,” Sheridan, 1992).
In addition, typologies of presence posit different types of specific presence
that can be used to understand a cross sectional experience of the entertainment
product. Three types of presence are social presence (Biocca, 1997), physical
presence or spatial presence (Wirth, Hartmann, Bocking, Vorderer, Klimmt,
Schramm et al., in press), and self presence (Lee, 2004a). I discuss self presence
particularly in the current paper because the notion of self presence raises an
interesting issue with regards to where entertainment experience occurs (e.g.,
domains).
Self presence is defined as a psychological state in which virtually
constructed self/selves are experienced as the actual self in real life (Lee, 2004a). If it
is true that we put ourselves into the entertainment world during the entertainment
experience, then we face a dilemma of self presence. According to Baumeister &
Muraven (1996), self is based on “having a physical body, experiencing reflexive
consciousness, having interpersonal connections, and belonging to small groups, and

63
exercising the executive functions of decision-making and self-regulation” (p. 406).
All these components of the self can be simulated in the entertainment world, except
for having a physical body (see Zimmke, 2001 for more discussion about
embodiment). Therefore, it is impossible to recognize self in the entertainment world
if we are transported into the entertainment world by suspending ourselves in the real
world (see the discussion in the previous section of ‘transportation’). To solve the
dilemma of self presence, we should regard the actual selves in the real world as a
background and take the representations of the entertainment world into our minds
during the entertainment experience. The self in the real world, of course, has all the
components of the self and has the ability to understand and take on
metarepresentations from the entertainment world through which the audience
experiences the identification between an implied self and an implied actor/actress
vis-à-vis self presence.
In conclusion, theories that explain a block of entertainment experience are
fairly new concepts, compared to the four theories that explain a cross sectional
entertainment experience. The theories—flow, transportation, and presence—all
explain the immersed states of the audience during the entertainment experience. In
fact, these concepts are often measured with self-report questionnaires about
immersion and involvement during the entertainment experience.

64
Role Play Model
In the previous sections, I reviewed seven theories, each of which uniquely
explains the entertainment experience. In spite of their uniqueness and different
traditions, there are important common assumptions and components among the
theories. Some of these commonalities are briefly indicated during the review of the
theories. In this section, I develop the multi-layering model of role play further to
provide a better understanding of the entertainment experience by focusing on the
important components of the seven theories. More specifically, the conceptual work
by Oatley (1996), Zillmann (1994), and Csikszentmihalyi (1993) has served as a
building block in the model.
Entertainment Experience as Role Play
Vorderer (2001) identifies many common aspects between entertainment
experience and play experience, suggesting that entertainment is a form of play.
Since play is a generic concept that usually lacks precise and clear operational
definition (Fein, 1981), I use an easily identified and readily defined form of play—
role play—to explain the entertainment experience.
As discussed in Chapter 1, a role is defined as a set of expectations for a
position in a social structure (Shaw & Constanzo, 1982) and a set of rights and duties
embedded within a social status (Linton, 1936). Thus, a role implies beliefs, desires,
knowledge, emotions, and all the expectations embedded within the pertinent social
position. Having said that, role play can be defined as a way in which you can

65
experience concretely the representations of interactions under examination (Johnson
and Johnson, 1997). To apply the definition to the entertainment experience, role
play is a way in which the audience can experience the emotional and cognitive
representations of the implied actor/actress in the entertainment product. There are
two different types of role play in entertainment, one from the audience and the other
from the actor/actress. In order to distinguish one from another, I describe role play
from the audience as role taking, whereas I use the term, role performance, to
describe activities of the actor/actress. Similar to the concept of role play, role taking
refers to social intelligence in that people put themselves in the place of other people,
similar to theory of mind (Mead, 1934). That is, the ability of role taking is directly
related to how we come to enjoy the entertainment product.
Indeed, the notion of role taking from the audience has already been
discussed in the process of identification (e.g., Oately [1994; 1999] uses a term,
“take on.”) and cognitive empathy (Zillmann [1994; in press] uses a term,
“perspective taking.”) although the processes are not explicitly described as role
taking. Another application of role taking ability is in clinical psychology.
Psychotherapy deals with role problems in that a defective role is reexperienced
through identification, sympathy, or empathy during the therapy (Fein, 1992).
Therefore, dramatic role taking and role performance are at the heart of the
transformations of our understanding, feeling, and valuing, whether in drama,
therapy, or everyday life (Landy, 1993).

66
Model Development
There are four assumptions in the proposed model based on the discussion
and literature review in the previous sections: (1) the audience puts the
representations of the implied character from the entertainment world into the
audience’s minds in the real world; (2) only coherent information from the
audience’s minds in the real world is filtered into the implied self in the
entertainment world semi-automatically; (3) the audience can have a specific
entertainment experience with regard to aesthetic distance, ranging from
identification to parasocial interaction, at a cross sectional moment of the audience’s
entire experience; and (4) the congruence between the outcomes and expectations of
the entertainment product may facilitate these experiences. Accordingly, I propose
four dimensions of the model: (1) the actual audience; (2) the implied audience; (3)
the actual actor/actress; and (4) the implied actor/actress. The entertainment
experience depends on interactions among the four dimensions (see Figure 4). The
role play process within the individual is labeled intra role play, whereas between
individuals is labeled inter role play. The optimal role play is defined as a
psychological state in that the entertainment world becomes extremely salient,
compared to the real world, through the optimal matching process in inter role play.
During the optimal role-play state, the audience may experience maximum
enjoyment of the entertainment product.

67
First we shall consider the audience side. The actual audience (e.g., actual
person A in Figure 8) is required to have enough skills to understand the
entertainment product such as knowledge, emotions, beliefs, and so on (e.g.,
requisites for the flow state). In the role taking process, incompatible information to
the entertainment experience is filtered out semi-automatically, which is the filtering
skill of the implied audience.

Figure 8. Role-play model in the non-interactive entertainment experience

For example, some of the audience can easily take on roles of the
actor/actress and show emotional responses in soap operas, regardless of the role
performance of the actor/actress. The levels of skill on the audience side do not
Role Expectations of
Implied Person A
• Beliefs
• Desires
• Knowledge
• Emotions
• Understanding
• Experience
Actual Person A Actual Actor B
Role Performance of
Implied Actor B
• Beliefs
• Desires
• Goals
• Emotions
• Experience
Inter Role Play
• Optimal matching
• Aesthetic distance
Intra Role Play
• Filtering skills
• Role taking
Intra Role Play
• Filtering Skills
• Role Performance

68
usually prevent the audience from enjoying non-interactive entertainment products
because they are usually created in a way to minimize required skills from the
audience side. Indeed, it is obvious that producers want as many people as possible
to watch their motion pictures.
Secondly, the actual actor/actress is also required to have enough skills for
role performance such as training for acting, emotions, knowledge, and so on. Unlike
the audience case, the actual actor/actress plays a role of different persona. Thus, role
performance process requires not only filtering skills but also the ability to identify
with the implied actor/actress. That is, the actual actor/actress needs to fill in
important information deliberately in order to feel as complete identification as
possible with the implied actor/actress. These skills from the actor/actress side
influence the overall role performance along with external factors that also influence
the overall role performance such as storylines, the director’s perspective, and so on.
It is important to note that these intra role plays from both sides are requisites for the
inter role-play process.
The successful inter role play (i.e., optimal role play) occurs when role
expectations (i.e., role-taking) from the implied audience and role performances from
the implied actor/actress are optimally matched or balanced. For instance, the
audience may expect the role of the protagonist to be a super hero/heroine. When the
role performance of the implied actor/actress meets this expectation, the audience is
likely to enjoy the entertainment product and feel identification, sympathy, or

69
empathy with the implied actor/actress during the entertainment experience. Once
the optimal role-play state is established, the audience may freely change their
position with respect to aesthetic distance for maximum pleasure (e.g., identification
with one character at one moment and sympathy with the character at next moment
or feeling both at the same time). However, there is an individual difference in the
role expectation, which is why a motion picture cannot be enjoyed by everyone. The
role performance also critically affects the inter role play because a poor intra role
play from the actual actor/actress results in a poor role performance of the implied
actor/actress. A good role performance of the implied actor/actress is expected by the
implied audience.
Interestingly, the role expectation of an implied audience does not have to
exactly match the role performance of an implied actor/actress. In fact, if they match
perfectly, the audience is likely to find the entertainment product boring or
uninteresting. Thus, there should be some fluctuation allowed between the role
expectation and the role performance within the range of a threshold. If the role
performance takes extremes, thus goes beyond the threshold of the audience’s
expectation, the audience cannot understand the story at all and may disengage from
the entertainment product (c.f., role ambivalence in playing out conflicting roles
[Landy, 1993]). This notion is similar to the hypothesis of the habitual strange by
Schlovsky (1917). According to him, the purpose of art is to present the ordinary in
ways that habitual responses from the audience are difficult. In this way, the art may

70
be seen afresh, not passed over, not taken for granted (Oatley, 1994). This is why I
intentionally avoid using the term, matching, for the optimal relationship between the
role expectation and the role performance. In order to optimize the enjoyment of
entertainment, the role performance should be strange enough to invoke curiosity,
interests, or arousal and yet should vary within the threshold of the role expectation
vis-à-vis the optimal role play.

Figure 9. Role-play model for interactive media

The model of role play for interactive media is slightly different from the
previous model although the theoretical components are the same (see Figure 9).
When a user plays a videogame with his or her friends, we can still use the previous
Role Expectations of
Implied Person A
• Beliefs
• Desires
• Knowledge
• Emotions
• Understanding
• Experience
Actual Person A
Role Performance of
Implied Character A
• Beliefs
• Desires
• Goals
• Emotions
Inter Role Play
• Optimal matching
• Aesthetic distance
Intra Role Play
• Mental skills
• Role taking

Intra Role Play
• Physical Skills
• Role Performance

71
model to understand the user’s entertainment experience. When a user plays a
videogame alone, then the actual friend who plays roles does not exist. In stead, the
actual user engages in both role taking and role performance.
In this model, role taking requires mental skills, whereas role performance
requires physical skills such as controlling a pad for videogames. In the
entertainment world, there are still the role expectations from an implied user and the
role performance by an implied character, although both are rooted in the same user
in the real world. A user may have certain expectations about the user’s performance
when the user plays a videogame regardless of the user’s actual skills to play. For
instance, when a videogame is too difficult, the skill for intra role play is not good
enough to result in a good role performance. Even in this case, if the user has a low
role expectation because it is the user’s first time to play such a game, then the
optimal role-play state could be established. The user may say that he/she enjoys
playing the videogame although the user’s performance is terrible. In a similar vein,
the optimal role-play state could be established when the videogame is too easy as
long as the inter role play is balanced. In other words, the optimal role-play state is
broken when a user expects a better performance than the user’s actual performance
(e.g., higher role expectation than role performance) and vice versa. In this case, the
imbalance in the optimal role play tends to be caused by intra role-play process for
the role performance (i.e., skills to play) rather than the process for the role taking.
Either way, a user cannot enjoy the entertainment product in case of inter role-play

72
imbalance. That is, the real world becomes more salient than the entertainment world,
which may lead the user to disengage from the entertainment product.
Entertainment Theories, Revisited
Finally, I briefly review how the conceptualized model of role play could
explain some of the limitations in entertainment theories identified previously. First,
people often identify with the opposite gender when they watch television programs.
I state that identification is not identifying an actual audience with the actual
actor/actress but identifying the role expectations from an implied audience with the
role performance by an implied actor/actress. Thus, the above identification process
may work in the following way: (1) people tend to have clearer expectations for the
opposite sex than the same sex because of mating strategies and opportunities for
reproducing their genes (see Buss, 2004); (2) the role performance by the implied
actor/actress (e.g., physical appearance, or personas) is usually excellent; (3) the
optimal role-play state is likely to occur because of the clear role expectation and the
excellent role performance (i.e., balanced inter role play); (4) as a result, people may
develop parasocial relationships with the implied actor/actress of the opposite gender.
The identification with the antagonist can be understood in a similar vein. People
have clear expectations for the antagonist. When the role performance of the
antagonist meets people’s role expectations, identification or sympathy could happen
as a result of the optimal role-play state.

73
Another example is documentaries such as Holocaust stories. For non-
fictional documentaries, background knowledge is required for role taking. That is,
people need to have a historical understanding of World War Two or the Nazis,
especially their collapse, prior to their exposure to Holocaust stories in order to enjoy
them. This historical information may be somewhat incompatible to the content
during entertainment experience (e.g., the war has already ended and yet people are
watching it occur), thus is filtered out temporarily during the entertainment
experience and re-filtered in at the end of the stories to provide the audience with
resolution. In this case the enjoyment of the entertainment product may not be
derived from pleasure but from learning. Nonetheless, if the role performance in the
Holocaust stories goes beyond one’s role expectations (e.g. too brutal), then the
audience just cannot or will not want to watch the documentary even though the
audience knows how the story ended.
To sum up the role-play process, only coherent information from the
audience’s minds in the real world is filtered into an implied self in the entertainment
world semi-automatically. It helps people understand superman’s flying or their
videogame character’s super power. Then, the congruence between the role
expectation and role performance facilitates the optimal role-play state from
entertainment experience. Finally, the audience can have a specific entertainment
experience with regard to the aesthetic distance, ranging from identification (i.e., the
first person view) to sympathy (i.e., the third person view). Players can identify

74
themselves with their characters in videogames or feel sympathy for their characters
when they died (e.g., identification cannot explain this feeling of being dead).
Managing the aesthetic distance is possible by controlling intra role play between
actual self and implied self through the filtering process as seen in children’s pretend
play.
Lastly, it is important to note that each theory I have reviewed in this chapter
conceptualizes the entertainment experience from a unique perspective. Discussing
some of the potential limitations and unintentionally misleading descriptions of the
theories helps develop a new theoretical model of role play that may provide a better
and more inclusive understanding of the entertainment experience. I hope the
conceptualization of the role-play model sheds lights on our understanding of
everyday entertainment experience by provoking new insights.

75
Chapter 4. Role Enactment Based on Gender Stereotypes in interactive
media: Effects of Role Play on the Management of Self-Concept and
Physical Distance in Virtual Reality Environments
As discussed in Chapter 1 (see also the next section), gender-derived social
identity could vary within an individual, depending on social contexts or the salient
gender-role identity at a given moment. Similarly, gender-stereotyping behaviors
such as interpersonal distance management could also vary, based on the salient
gender-role identity. Having said that, gender roles and gender-stereotyping
behaviors in interpersonal distance may serve as a good context for research in which
the basic assumptions of the role-enactment model are investigated: the changes of
people’s self-concept (i.e., gender-role identity) and corresponding behaviors (i.e.,
interpersonal distance) as a product of their role play. In this chapter I report on the
results of an empirical study that consists of two experiments in order to examine the
effects of role play on the self-concept and corresponding behaviors, in the context
of interpersonal distance in interactive media.
There have been many studies on proxemics, the use of space in social
interactions, in various research areas, including the concepts of territoriality and
personal space (Patterson, 1978). Among these, studies have demonstrated gender-
stereotyping communication behaviors in interpersonal distance (Aiello & Jones,
1971; Dosey & Meisels, 1969). Specifically, females express intimacy in social
interaction within same-sex groups more naturally than males do, which often results

76
in differences of interpersonal distance between communication partners. Physical
distance in male-male dyads is greater than in both male-female and female-female
dyads regardless of the types of relationship (e.g., Sussman & Rosenfeld, 1982; Tice,
1992; Uzzell & Horne, 2006). Although there were limitations of measuring
interpersonal distance in early studies (see Patterson & Edinger, 1987 for more
discussion about measurement issues), these patterns of gender stereotyping behavior
in interpersonal distance are known to be consistent.
It is also interesting that recent studies have investigated interpersonal
distance in virtual environments (VEs) and found the same gender-stereotyping
communication behaviors (e.g. Bailenson, Blascovich, Beall, & Loomis, 2003; Yee,
Bailenson, Urbanek, Chang, & Merget, in press). Studies have demonstrated that
studying nonverbal behaviors and training motor skills in virtual reality
environments help enhance our understandings of nonverbal behaviors in face-to-
face environments (Bailenson, Beall, Blascovich, Loomis, & Turk, 2005; Bailenson
& Yee, 2005) as well as enhance arm movements for people post-stroke in their
activities of daily life (Holden, 2005; Viau, Feldman, McFadyen, & Levin, 2004).
The advantages of using virtual reality as a means for research include (1) ability to
measure behavioral patterns precisely by using motion trackers; (2) manipulability to
allow the researcher to induce various scenarios dynamically; (3) interactivities to
motivate the participant to engage in the experiment actively.

77
However, the studies are unable to fully explain the direction of causalities
and the reason why the same gender-stereotyping behaviors appear in a new
relationship, human-computer interaction (HCI), in a new context, VEs, because they
passively observed behavioral patterns (e.g., participants in a laboratory or avatars in
massively-multiplayer online role-playing games) without manipulating scenarios for
social interactions in VEs: Studies lack either interactivity or manipulability among
the advantages of VEs.
Nonetheless, these consistent findings of gender-stereotyping interpersonal
distance in both F2F and computer-mediated environments inspired me to
empirically test the role-play perspective in the context of interactive media. Given
the literature review and discussions in Chapter 1, it is possible that certain
nonverbal communication behaviors such as interpersonal distance differ as a
function of gender self-concept but may be susceptible to changes based on role play.
In other words, if an individual plays a certain gender role naturally in social
contexts (e.g., role play in interactive media), the gender-role identity may become
highly salient, which, in turn, may influence the individual’s self-concept and
corresponding behaviors. Therefore, the purposes of the study reported here are (1)
to investigate the effects of role play on the management of the self-concept and
interpersonal distance by actively manipulating gender-role play in interactive media;
(2) to extend existing interpersonal theories in face-to- face (F2F) context to the
understanding of human-computer interaction (HCI; a new relationship) in VEs (i.e.,

78
a new context); (3) to investigate the social use of physical space in interpersonal
communication in an ecologically realistic experiment using VEs as a tool; and (4) to
discuss the design of interfaces for VEs and videogames based on points-of-view
(the second experiment). This study is unique in that I used an ecologically valid way
to measure interpersonal distance, using a motion tracker in VEs and induced the
salience of a certain gender-role identity to the participants through role-play
scenarios and altering points-of-view.
Gender Identity and Gender-Stereotyping Communication Behaviors
Gender-derived social identity
Researchers started to realize that gender identity is one of the social
identities that could vary even within the same gender group or within the same
person across occasions (Cameron & Lalonde, 2001; Lindeman & Sundvik, 1995;
Palomares, 2004.; Uzzell & Horne, 2006; see also Chapter 1). For example, a man
may feel strong masculinity or manhood when his identity as a man becomes highly
salient and may feel little masculinity when his gender identity is not salient or the
opposite gender identity becomes highly salient. If we extend this argument further,
we may be able to accentuate female-gender identity salience within a group of men.
That is, even the opposite gender identity may become highly salient within an
individual, depending on how the individual internalizes social interactions that
induce the salience of the opposite gender-role identity naturally. Based on the above
discussion and Chapter 1, the following hypotheses are proposed:

79

H1-1a: People will identify themselves with the character more positively
when their own gender matches the gender of the character that they have
played (e.g., role salience).
H1-1b: Nevertheless, people will identify themselves with the opposite gender
character positively (i.e., the average score on identification will be greater
than the score of three in a seven-point scale).
H1-2: People will perceive the importance of their own biological sex to the
self-concept (i.e., centrality) more significantly when their own biological sex
matches the gender of the character that they have played (e.g., self-concept).

Gender difference in interpersonal distance
Studies have demonstrated gender-stereotyping behaviors in interpersonal
distance between participants and the same gender confederate in the sitting distance
(Sussman & Rosenfeld, 1982; Tice, 1992), between avatars in a massively
multiplayer online role-playing game (Yee et al., in press), and between participants
in dyadic interactions (Uzzell & Horne, 2006). Their results support the conclusion
that male-male dyads maintain greater physical distance than female-female dyads.
In addition, Uzzell and Horne (2006) found that gender role accounts for more of the
variation in interpersonal distance than biological sex. Although they simply
measured participants’ gender-role identity prior to the experiment, rather than

80
actively inducing gender-role identity, their findings suggest that gender-stereotyping
interpersonal behaviors vary as a function of gender-role identity that could become
highly salient based on a prescribed role in social contexts. Thus, the following
hypotheses are proposed:

H1-3: People assigned to the male role will maintain greater physical
distance between themselves and the computer agent than people assigned to
the female role, regardless of their biological genders (e.g., communication
behaviors)

Finally, Sussman and Rosenfeld (1982) found cultural differences in
interpersonal distance. In their study, Venezuelan participants from relatively high
contact cultures of Latin America maintained greater interpersonal distance than
Japanese participants who represented low contact cultures did. Therefore, it is
necessary to control for possible cultural effects in this study, in order to examine
gender stereotyping behaviors in interpersonal distance without any bias from
cultural differences.
Method
Experimental design
A 2 (subject gender: male vs. female) x 2 (gender-role play: male role vs.
female role) between-subjects analysis of covariance (ANCOVA) design with

81
subjects’ ethnicity as a covariate
3
was used to test the hypotheses in a laboratory
environment. The two levels of the gender-role variable (i.e., male role vs. female
role) was manipulated by creating two versions of a virtual reality scenario in that
participants were randomly assigned to act out either a male or female character (see
Appendixes). A total of 56 undergraduate students (28 males and 28 females)
enrolled at the University of Southern California participated in the experiment.
Procedures
Pre-experiment. Upon participants’ arrival at a laboratory an administrator
gave each participant brief instructions about (1) their participation in a new
experimental animation movie-making with a three-dimensional virtual reality (VR)
system; (2) their role in the movie and how they were to act out the assigned
character based on a pre-written script; and (3) how to use a head-mounted display
(HMD). Participants’ gender role was randomly assigned upon their arrival. In order
to make the participants accept their gender role in a natural way, a 3D animation
movie-making scenario was set up with two pre-written scripts (i.e., one for the
male-male dyads condition and the other for the female-female dyads condition) for
gender salience (see the appendix for actual scripts used in the experiment).
Participants were provided enough time to read the scripts alone at least twice and to
be familiar with their role in order to make the assigned gender-role identity highly
salient. After reading the script by themselves, participants were asked to practice

3
ANCOVA design with subjects’ ethnicity as a covariate was used to statistically control for possible
cultural effects on interpersonal distance.

82
their role with an administrator by taking roles in turn and reading out the script
aloud.
During experiment. Participants wore a HMD and engaged in conversation
with their virtual friend in a bar-like VR environment. Participants played their
assigned gender role (e.g., either male or female, regardless of their biological sex)
and read aloud the script that was also displayed on the bottom of the screen in the
HMD. Participants were instructed to spend as much time as they wanted and to
freely move around in the VR environment.
Post experiment. Participants completed a short self-report survey about their
experience. Then, participants were debriefed about the deception; that is, that the
study was not about a new 3D movie making but about investigating people’s
behavioral patterns in virtual environments.
Manipulation
An immersive, three-dimensional virtual bar scene was created, including a
piano, tables, and other details, around which the four virtual agents were either
standing or seating (see Appendixes). Participants could see the four agents,
including one agent that played the role of the participants’ best friend, with whom
they engaged in conversation. The HMD contains a separate display monitor for each
eye to make users perceive 3D stereo and a panoramic 6 degrees of freedom (DOF)
effect. One magnetic tracker was attached to the HMD to track participants’
movements in the VR system.

83
Gender-role salience was manipulated by the five-minute long scripts and
role- play scenario. Reading out the script at least twice before engaging in social
interaction would make the assigned gender-role identity highly salient at a given
moment, not to mention that stronger perception of the gender salience was expected
as participants interacted with the virtual agent. The script for male-male dyads was
written based on the script from a movie, “Hitch.” Similarly, the female-female
dyads script was written based on an episode from the television series, “Sex and the
City” (see Appendixes).
Measures
First of all, the positional data of a participant during the interaction were
logged into a database system in real time whenever a speech was made by either a
participant or the computer agent, in terms of the x and y planes. Later, a series of
paired x and y positions of the participant was used to compute the average shortest
distance between the participant and the agent based on a formula: average of the
square root of x square plus y square (c.f., the location of the computer agent was
fixed and served as a zero point). The unit of physical distance is the inch.
Secondly, a paper-based self-report survey was used to measure participants’
identification with the gender role (i.e., role-identity salience) and subjective
importance of their own gender to their self-concept (i.e., centrality). The measure
for perceived identification with the gender role that participants have played is
modified from the three-factor model of gender-derived social identity that was

84
tested in five different studies, including two studies for gender-derived social
identification, with a total of 1078 respondents (Cameron, 2004). The measure
consists of two factors, ingroup ties and ingroup affect. Each factor consists of four
sub-questions with a seven-point Likert scale, with one being strongly disagree and
seven being strongly agree (see Appendixes for the questions). The final index is the
average score on the eight items (Cronbach’s α = .87). Higher scores indicated a
greater sense of identification with the gender role. Subjective importance of being a
man or woman to the self-concept was measured by four items used to measure
centrality in studies by Cameron (2004). Higher scores indicated a greater perceived
importance of their own gender to the self-concept. Alpha reliability was .70 for the
matching gender and gender-role conditions and .45 for the mismatching gender and
gender-role conditions. The relatively low scores are likely to be due to the nature of
the experimental design. That is, a group of male participants played a female
character, and a group of female participants played a male character in spite of the
fact that gender-role identity is highly correlated with biological sex.
Lastly, participants’ attitude toward new technologies was measured using
four items with a seven-point Likert scale (Cronbach’s α = .92). Higher scores
indicate more positive attitudes toward new technologies.
Results
A 2 (gender of subjects: male vs. female) x 2 (gender role: male role vs.
female role) factorial ANCOVA with subjects’ ethnicity as a covariate was used to

85
test all hypotheses (see Table 1). Subjects’ ethnicity was statistically controlled in
order to eliminate any possible cultural factors that might have influenced the
management of physical distance in dyadic relationships. Subjects’ attitude toward
new technologies was not included in the final analysis because of its non-significant
relationship with dependent variables.
Consistent with Hypothesis 1-1, the participants identified themselves with
the gender role that they have played more positively when the gender of the
character matched their own biological sex (M = 5.13, SD = 1.01 for male
participants; M = 5.37, SD = 0.67 for female participants) than when the gender role
mismatched their sex (M = 4.11, SD = 1.12 for male participants; M = 4.60, SD =
1.16 for female participants), F(1, 51) = 11.27, p<.01 (see Figure 10). However, even
in the mismatched conditions (e.g., a man played a female role; and a woman played
a male role) the participants identified themselves with the opposite gender role
positively (M = 4.11, SD = 1.12, t(13) = 3.69, p < .01; M = 4.60, SD = 1.16, t(13) =
5.15, p < .001, respectively).

86
Table 1
ANCOVA results in Experiment 1
Means and standard deviations F values and effect sizes
Subject gender Gender role Main effects Interaction
effects

Dependent variables
Male Female Male
role
Female
role
Subject
gender (G)
Gender role
(R)
G x R
Physical distance 65.73
(24.71)
56.43
(18.14)
70.55
( 24.92)
51.61
(16.84)
2.00
η
2
= .04
14.18**
η
2
= .22
0.12
η
2
= .00
Identification with
the character
4.62
(1.17)
4.98
( 1.01)
4.87
( 1.10)
4.74
(1.11)
1.45
η
2
= .03
0.16
η
2
= .00
11.27**
η
2
= .18
Centrality 5.37
(1.07)
5.20
(1.03)
5.00
(0.99)
5.56
(1.03)
0.24

η
2
= .01
4.06*
η
2
= .07
2.59
η
2
= .05

Note.
+
p<.10, two-tailed. *p<.05, two-tailed. **p<.01, two-tailed.

87
4.11
5.13
5.37
4.6
1
2
3
4
5
6
7
Male Role Female Role
Male
Female

Figure 10. Identification as a product of gender role and biological sex
Note. Higher scores indicate that participants identified themselves more positively
with the gender role. The index is an average score of eight items with a seven-point
Likert scale (Cronbach’s α = .87).
F(1, 51) = 11.27, p<.01

Hypothesis 1-2 was not supported. The participants perceived the importance
of their own biological sex to their self-concept equally across the matching and
mismatching conditions, F(1, 51) = 2.59, n.s. Instead, I found the main effect of
gender role on centrality. The participants in the female role group felt a greater
sense of centrality (M = 5.43, SD = 1.09 for male participants; M = 5.70, SD = 0.99
for female participants) than the participants in the male role group (M = 5.30, SD =
1.08 for male participants; M = 4.70, SD = 0.83 for female participants), after
controlling for their biological sex, F(1, 51) = 4.06, p<.05 (see Figure 11). It is

88
worth noting that the main effect of gender role on centrality is primarily driven by
female participants, the pattern of which is consistent with Hypothesis 1-2.
5.43 5.3
5.7
4.7
1
2
3
4
5
6
7
Male Role Female Role
Male
Female

Figure 11. Main effect of gender role on centrality: Experiment 1
Note. Higher scores indicate that participants perceive the importance of their own
gender more often and more seriously. The index is an average score of four items
with a seven-point Likert scale (Cronbach’s α = .70 for the matching gender and
gender-role conditions; Cronbach’s α = .45 for the mismatching gender and gender-
role conditions).
F(1, 51) = 4.06, p <.05

Consistent with Hypothesis 1-3, the participants in the male-role group
maintained greater physical distance (M = 70.55, SD = 24.92) than the participants in
the female-role group (M = 51.61, SD = 16.84), F(1, 51) = 14.18, p<.001, after
controlling for their biological sex. However, the difference in the management of
physical distance between male participants and female participants (i.e., main effect

89
of participants’ biological sex) was not significant, F(1, 51) = 2.00, n.s., after
controlling for the gender role that they have played.
Discussion
First of all, participants identified themselves even with the opposite gender
role positively (the average scores on identification were above the median score of
four and statistically greater than the score of three) although they did not identify
with the opposite gender role as much positively as they did with the same gender
role. It makes sense that people cannot perfectly identify themselves with roles that
they are not familiar with, at least at the conscious level. Nevertheless, the finding
implies that playing a role through social interaction can make the role identity
highly salient at a given moment even when the role identity has not been a part of
the self prior to the role play.
Secondly, role play influenced the perceived importance of biological sex to
the self-concept (i.e., centrality) only for female participants. The average score on
centrality was lower when female participants played the male role, compared to
when they played the female role. This finding implies that playing a gender role
influences how people perceive the importance of their biological sex to self-
definition either more seriously when the role matches their biological sex or less
seriously when the role mismatches their sex, which suggests that role play can
affect the self-concept in the same way as it influences the role identity. However,
the relationship between role play and the self-concept was only supported for

90
female participants, and not for male participants. For male participants, it was
important to be men even when they played the female role. A possible explanation
for this particular reaction from male participants is virtual homophobia. Studies
have demonstrated the greater magnitude of homophobia among men in that overt
affection was regarded as less appropriate in male-male relationships than female-
female relationships (Morman and Floyd, 1998). Similarly, a significantly more
negative relationship between homophobia and comfort with same-sex touch was
found for men (Roese, Olson, Borenstein, Martin, and Shores, 1992). The magnitude
of negative relationship between homophobia and evaluations of same-sex
affectionate touch was significantly increased for men (Floyd, 2000). The male
participants in the experiment might have been reluctant to state that being men is
not important to their self-concept, especially after playing the female role. They
might have felt pressure to complete the survey in the way that emphasized the
importance of masculinity in their lives, partly due to homophobia (e.g., social
desirability bias in a self-report survey). Or it is also possible that male participants
were not willing to give up privileges embedded within the ideology of the dominant
male identity. In any case, the results from this experiment cannot directly support
these arguments because I did not measure any of the possible explanations in the
experiment. Therefore, future research needs to consider more complex gender-
related ideology for further discussions on this topic.

91
Finally, participants in the male-role group maintained greater distance than
participants in the female-role group, regardless of their biological sex. Even when
participants played the opposite gender role, they showed gender-stereotypic
behaviors in accordance with the opposite gender role that they played during the
experiment. It is important to note that gender-role identity, not biological sex,
accounted for the management of physical distance in the virtual environment (R
2

= .32), which is consistent with findings in a previous study done in the F2F context
(see Uzzell & Horne, 2006). This finding implies that communication behaviors such
as interpersonal distance could vary as a function of role play in social contexts. It
also provides empirical evidence that gender identity is one of the role identities and
varies depending on the degree to which a certain gender-role identity becomes
salient through role enactment.
Taken together, the results support the utility of the role-play perspective to
understand the effects of engaging in interactive media such as playing videogames,
not to mention our daily activities. For example, playing a brutal role in violent
videogames could make the violent role identity highly salient during the play, which
is likely to affect players’ self-concept and behaviors in a way that reflects the
violent role (e.g., aggressive thoughts and physical aggression). In a similar vein,
playing an instructive role in educational videogames could influence players’ self-
concept and behaviors to conform to the positive role identity. Thus, the role-play
perspective serves as a good alternative theoretical framework that could explain

92
both positive and negative consequences of using interactive media. Finally, it is
worth mentioning that communication behaviors based on role play seem to come
out unconsciously. The result from participants’ psychological identification showed
that participants did not identify with the opposite gender role as much as they did
with the matching gender role, which was measured by self-report (i.e., the
significant difference in identification between the opposite gender-role condition
and the matching gender-role condition). However, participants’ behaviors in
interpersonal distance showed no difference between when they played the opposite
gender role and when they played the matching gender role. Participants showed
typical gender stereotyping behaviors in distance management when they played the
male role (i.e., greater distance) and the female role (i.e., shorter distance), regardless
of their own biological sex. The results together suggest that people do not notice
how their behaviors are influenced by the salient role identity at the conscious level,
measured by self-report, but do show behaviors in accordance with the salient role
identity at the unconscious level, measured by observing behaviors. Having said that,
it is no wonder that we often see third-person effects in media studies (see Perloff,
1993 for the review of third-person effect studies).
Experiment 2
Results from Experiment 1 showed that role play in interactive media could
influence role-identity salience, the corresponding self-concept and behaviors at least
temporarily. The purposes of Experiment 2 are to reconfirm the effects of gender-

93
role play on role-identity salience and the management of the self-concept and
interpersonal distance by replicating Experiment 1 with larger samples, as well as to
investigate a new important factor, virtual embodiment, that may also influence
interpersonal process in interactive media (see the next section). Therefore,
Experiment 2 tests the same four hypotheses investigated in Experiment 1.

H2-1: People will identify with a character more positively when their own
gender matches the gender of the character (e.g., role salience).
H2-2: People will perceive the importance of their own biological sex to the
self-concept (i.e., centrality) more significantly when their own biological sex
matches the gender of the character that they have played (e.g., self-concept).
H2-3: People assigned to the male role will maintain greater physical
distance between themselves and the male-computer agent than people
assigned to the female role with female-computer agent.

In addition to the above hypotheses, Experiment 2 also examines an
important contextual factor that may moderate media effects in interactive media:
virtual embodiment.
Virtual Embodiment
Baumeister and Muraven (1996) recognize the basis of the self as “having a
physical body, experiencing reflexive consciousness, having interpersonal

94
connection, and belonging to small groups, and exercising the executive functions of
decision-making and self-regulation” (p. 406). Given the above statement, having a
physical body (i.e., embodiment) is important for recognizing the self and identity.
Although embodiment is a loaded term, Lee, Jung, Kim, and Kim (2005) simply
define physical embodiment as bodily presence. Similarly, I define virtual
embodiment as a visually presented body of a user’s character in interactive media.
Given the definition, the concept of virtual embodiment is similar to the concept of
points-of-view. In the first person view, people cannot see the bodily presentation of
their own characters, which constitutes a virtually disembodied condition. On the
other hand, people can see the bodily presentation of their own characters in the third
person view, which constitutes a virtually embodied condition. With regards to
points-of-view, a recent videogame study about violence (Farrar, Krcmar & Nowak,
2006) has demonstrated an interesting finding that players were more focused and
involved, thus felt greater hostility and physical aggression when they played in the
third person view than in the first person view. A possible explanation is that seeing
bodily presentation of players’ own characters may help them immerse themselves
into the role that they are playing (e.g., seeing a character wearing an army uniform
with holding a rifle may help a player to play the role of a soldier). Highly involved
role play is likely to make role identity extremely salient, as well as to influence
player’s self concept and behaviors that reflects the portrayed role identity. To sum
up, altering points of view could affect the degree to which the role identity becomes

95
salient, which may intensify or attenuate communication behaviors embedded within
the role identity. Thus, the following hypothesis is proposed:

H2-4: Discrepancies of distance between male and female roles will be
greater in the third-person view (the embodiment condition) than in the first-
person view (the disembodiment condition).

Method
Experimental Design and Procedures. The same 2 (gender role: male vs.
female) x 2 (points of view: first-person view vs. third-person view) between-
subjects analysis of covariance (ANCOVA) design with subjects’ ethnicity as a
covariate was used in the second experiment. The biological sex of participants was
disregarded for the second experiment because of its non-significant effect on
interpersonal distance, after controlling for the gender role effects as shown in the
first experiment as well as previous studies (e.g., Uzzell & Horne, 2006). A total of
84 female-undergraduate students enrolled at a major university in the west coast of
the U.S. participated in the second experiment.
Manipulation. In addition to the previous VR system, two different points of
view were created for virtual embodiment (see Appendixes). In the first-person view
(i.e., the disembodiment condition) participants could see only the four computer
agents without self-representation (i.e., their own avatar) just like the conditions in

96
Experiment 1, whereas in the third-person view (i.e., the embodiment condition)
participants could see their own character (e.g., a photo of the face next to each script
line on the bottom of the screen, a back of the head and part of the shoulder on the
screen: Over-the-shoulder looking view). The position of participants’ own character
was automatically altered in subtle ways so as not to block the view of the computer
agent during the interaction (only either slightly to left or right without changing the
distance between the participants’ own character and the computer agent).
Measures. All dependent variables were measured in the same way as in
Experiment 1 (i.e., physical distance units of inches; identification with the gender
role [Cronbach’s α = .77]; centrality [Cronbach’s α = .78 and .58 for the matching
gender condition and mismatching gender condition respectively]; and attitude
toward new technologies [Cronbach’s α = .88]).
Results
A 2 (gender role: male vs. female) x 2 (points of view: first person vs. third
person) factorial ANCOVA with subjects’ ethnicity as a covariate was used to test all
hypotheses (see Table 2). Again, attitude toward new technologies was not included
in the final analysis because it was not significantly correlated to any of the
measured variables.
Hypothesis 2-1 was supported. The participants identified themselves with
the gender role that they played more strongly when the gender of the character
matched their own biological sex (i.e., the female-role condition; M = 5.34, SD =

97
0.66) than when the gender role did not match their sex (i.e., the male-role condition;
M = 4.76, SD = 0.86), F(1, 79) = 11.81, p<.01. Again, the average score on
identification in the mismatching condition was positive (M = 4.76, SD = 0.86, t(41)
= 13.28, p < .001).
Consistent with Hypothesis 2-2, the participants perceived the importance of
their own biological sex to their self-concept more significantly when they played
the matching gender role (M = 5.67, SD = 1.01) than when they played the
mismatching gender role (M = 5.07, SD = 1.13), F(1, 79) = 6.62, p<.05 (see Figure
12).
4.77
5.68
5.36
5.65
1
2
3
4
5
6
7
Male Role Female Role
First Third

Figure 12. Main effect of gender role on centrality: Experiment 2
Note. Higher scores indicate that participants perceive the importance of their own
gender more often and more seriously (Cronbach’s α = .78 for the matching gender
condition; and Cronbach’s α = .58 for the mismatching gender condition).
F(1, 79) = 6.62, p < .05

98
Table 2
ANCOVA results in Experiment 2
Means and standard deviations F values and effect sizes
Points of View Gender role Main effects Interaction
effects

Dependent
variables First Third Male
role
Female
role
Points of
view (V)
Gender role
(R)
V x R
Physical distance 51.11
(9.48)
75.13
(11.34)
67.92
( 16.26)
58.33
(14.23)
142.11**
η
2
= .64
22.67**
η
2
= .22
3.33
+

η
2
= .04
Identification with
the character
5.01
(0.95)
5.08
( 0.66)
4.76
( 0.86)
5.34
(0.66)
0.19
η
2
= .00
11.81**
η
2
= .13
2.23
η
2
= .03
Centrality 5.23
(1.11)
5.51
(1.10)
5.07
(1.13)
5.67
(1.01)
1.51

η
2
= .02
6.62*
η
2
= .08
1.88
η
2
= .02

Note.
+
p<.10, two-tailed. *p<.05, two-tailed. **p<.01, two-tailed.

99
Consistent with Hypothesis 2-3, the participants in the male-role group
maintained greater physical distance (M = 67.92, SD = 16.26) than the participants in
the female-role group (M = 58.33, SD = 14.23), F(1, 79) = 22.67, p<.001.
54.07
81.78
48.15
68.5
40
45
50
55
60
65
70
75
80
85
90
First Third
Male Role
Female Role

Figure 13. Interaction effect on physical distance
Note. Physical distance as a function of gender role and points of view, controlled
for subjects’ ethnicity (the unit is the inch).
F(1, 79) = 3.33, P < .10

Hypothesis 2-4 was marginally supported. In addition to the significant
difference in interpersonal distance between the male and female role group, the
discrepancies of the physical distance between male and female roles were greater in
the third-person view (M = 81.79, SD = 8.34 for the male-role group; M = 68.50, SD
= 10.07 for the female-role group) than in the first-person view (M = 54.07, SD =

100
8.32 for the male-role group; M = 48.15, SD = 9.83 for the female-role group), F(1,
79) = 3.33, p=.07 (see Figure 13).
Discussion
The results indicate that participants identified themselves with the matching
gender role more strongly than with the discrepant gender role. Nevertheless,
participants identified even with the opposite gender role, as shown in the results
from Experiment 1. In addition, participants perceived the importance of being
women to their self-concept more seriously when they played the female role than
when they played the male role. Taken together, playing a gender role in interactive
media not only makes the gender-role identity highly salient, but also affects the
gender self-concept in the way that conforms to the salient gender-role identity.
Consistent with the results from Experiment 1, participants maintained
greater physical distance when they played the male role than when they played the
female role. This finding reconfirms the utility of the role-play perspective: playing a
male role influences people’s behaviors in interpersonal distance in that they
maintain greater distance, regardless of their biological sex. The gender role,
together with points-of-view, accounted for the most variance in interpersonal
distance (R
2
= .68).
Results also showed that the gender difference in interpersonal distance
became more magnified when participants played the gender role in the third-person
view than when they played the role in the first-person view. This finding is

101
consistent with recent findings that third-person view or seeing self-representations
changes players’ behaviors to conform to the portrayed selves in interactive media
(e.g., Farrar, Krcmar & Nowak, 2006; Yee & Bailenson, in press). The result also
implies that contextual factors such as points of view can affect players’
psychological and behavioral consequences after playing interactive media, in
addition to the content of media such as violent or instructive videogames. Based on
this finding, I argue that seeing bodily presentation of one’s own character helps a
person become immersed into the role being played, which tends to result in more
stereotyped behaviors that conform to the salient role identity.
It is very interesting to note that the main effect of gender role on centrality is
mainly driven by participants in the first person view (see Figure 12). Female
participants in the third person view showed a very similar pattern as male
participants demonstrated in Experiment 1: no difference in centrality between the
male role and female role groups. This result is quite opposite to the result from
behavioral data in that participants showed more magnified gender-stereotyping
behaviors in interpersonal distance in the third person view. These rather conflicting
results imply possible differences between short term and long term effects of
playing interactive media (see Williams & Skoric, 2005). During play, seeing a
bodily representation of own character seems to help players immerse themselves
into the role, which results in behaviors that conform the salient role identity at the
unconscious level. After play, however, players may suddenly realize that the highly

102
salient role identity that is not a part of “true” selves. Then, the discrepancy between
the extremely salient role identity and players’ self-concept could be magnified at the
conscious level, which is likely to reinforce players to come back to their “true”
selves, instead of producing a lasting impact on the self-concept. It may be the
reason why many people have not become extremely aggressive, given the huge
number of people who play violent videogames. Although the difference between
short-term and long-term effects is a very important issue in media studies, there has
not been enough research, especially in the context of interactive media. For future
research, we need more carefully designed longitudinal studies that examine short-
term and long-term effects of engaging in interactive media.
General Discussion and Conclusions
In summary, results from Experiment 1 assess the utility of the role-play
perspective to understanding interpersonal process in interactive media. Playing a
role through social interactions influences the salience of the role identity, as well as
changes in the self-concept and corresponding communication behaviors in a way
that conforms to the salient role identity. More specifically, participants maintained
greater physical distance between themselves and the computer agent in VEs when
they played the male role, regardless of their own biological sex. The results support
(1) the assertion that gender identity is a role identity that varies within an individual,
depending on role play through social interactions; (2) the Computers Are Social
Actors (CASA) paradigm positing that people treat non-human machines, including

103
media characters or computer agents as if they were social actors; and (3) the utility
of interpersonal theories to understanding our attitudes and behaviors in computer-
mediated environments. The gender difference in perceived centrality after playing
the opposite gender role suggests that future research needs to deal with more
complicated gender ideology to understand possibly different gender effects of
playing interactive media.
The findings of Experiment 2 confirm the results of Experiment 1 by
successful replication of the first experiment with a larger sample size. As
hypothesized, female participants who played the male role indicated positive
identification with the opposite gender role, perceived their biological sex o be less
important to their lives and maintained greater physical distance, compared to female
participants who played the female role. In addition to reconfirming the findings of
Experiment 1, results showed that gender-stereotyping behaviors in interpersonal
distance became more magnified when participants played the role in the third-
person view than when they played the role in the first-person view. Although it was
marginally significant, this finding is consistent with previous findings in videogame
studies: seeing a bodily representation of own character helps players be more
focused and involved in videogames and changes players’ behaviors to conform to
the image of self-representation (i.e., salient role-identity). The result implies that
altering points of view can affect players’ experience and behaviors during and after
playing videogames, which, I believe, provides industry practitioners with strategic

104
implications for their interface designs. However, somewhat conflicting results from
the psychological measure of centrality and behavioral measure of physical distance
suggest a possible difference between short-term and long-term effects of playing
interactive media. Future research needs to address this important issue of short-term
and long-term effects of our virtual experience in interactive media.
There are a couple of limitations in this study. The study may not have been
able to detect small effects because of its small sample size, especially in Experiment
1. Further, there was much noise in the behavioral data, which resulted in relatively
large standard deviations. With regard to the measure of centrality, it could have
been measured twice, pre-experiment and post-experiment, in order to compare
changes in participants’ self concept in more precise ways. However, the survey was
not administered prior to the experiment in order to prevent participants from
noticing any clue that the experiment was about gender. Indeed, participants did not
know the true purpose of the study until they were debriefed. In order to see changes
in the self-concept, future research needs to measure pre and post self concept in a
subtle way so that participants cannot notice the purpose of the research. In addition,
it will be very interesting to see reactions from participants when the computer agent
moves forward or backward during the interaction in future research. In this way, we
would be able to test whether the increased gender salience is in response to social
interactions or in response to mere role enactment in instructional sets. Finally, the
role-play perspective needs to be tested in different forms of interactive media,

105
including videogames for more theoretical and practical implications. Nonetheless, I
believe that the current study successfully addresses the utility of interpersonal
theories and the role-play perspective to understand a new relationship of HCI in a
new context of computer-mediated environments.

106
Conclusion
In short, I proposed a new theoretical model of role play to enhance our
understandings of virtual experience in interactive media, based on discussions in the
first three chapters. The multi-layering model of role enactment could provide
explanations on how people engage in role play by creating multiple perceptual
layers and managing aesthetic distance among those layers to identify or sympathy
with implied characters. The model could also elucidate possible motivations to
engage in role play from its evolutionary nature, as well as the consequences of role
enactment in terms of players’ role-identity salience, changes in self-concept and
corresponding behaviors. In Chapter 4, the results from two experiments provided
preliminary empirical evidence to support the utility of the role-play perspective by
demonstrating how playing a gender role in interactive media could influence
players’ gender self-concept and physical distance management. In addition, results
showed that altering points of view in interactive media could magnify behaviors in
the way that conforms to the salient role identity.
Taken together, the role-play model could provide an additional theoretical
framework to understand the effects of role enactment based on role-identity
stereotypes in interactive media (e.g., effects of playing videogames), in addition to
other theories such as social learning and priming theory. Finally, the current
dissertation just proposed a basis to conceptualize the role-play model and discussed
its potential to enhance our understandings of virtual experience. The empirical study

107
in the final chapter is just the beginning of research in the role-play perspective, thus,
cannot fully explain the proposed model in the previous three chapters. We surely
need more sophisticated theoretical and empirical research to examine the multi-
layering features of the role enactment model and its relationship with other
entertainment theories in future.

108
References
Adams, G. R., & Marshall, S. K. (1996). A developmental social psychology of
identity: Understanding the person-in-context. Journal of Adolescence, 19,
429-442.

Aiello, J. R., & Jones, S. E. (1971). Field study of the proxemic behavior of young
school children in three subcultural groups. Journal of Personality and Social
Psychology, 19, 351-356.

Ang, I. (1985 [1982]). Watching Dallas: Soap opera and the melodramatic
imagination (D. Couling, Trans.) London: Methuen.

Axelrod, R. & Hamilton, W. D. (1981). The evolution of cooperation. Science, 211,
1390-1396.

Bailenson, J. N., Beall., A. C., Blascovich, J., Loomis, J., & Turk, M. (2005).
Transformed Social Interaction, Augmented Gaze, and Social Influence in
Immersive Virtual Environments. Human Communication Research, 31, 511-
537.

Bailenson, J. N., Blascovich, J., Beall, A. C., & Loomis, J. M. (2003). Interpersonal
distance in immersive virtual environments. Personality and Social
Psychology Bulletin, 29, 1-15.

Bailenson, J. N., & Yee, N. (2005). Digital chameleons: Automatic assimilation of
nonverbal gestures in immersive virtual environments. Psychological Science,
16, 814-819.

Baron-Cohen, S. (1996). Mindblindness: An essay on autism and theory of mind.
MIT Press.

Basil, M. D. (1996). Identification as a mediator of celebrity effects. Journal of
Broadcasting & Electronic Media, 40(4), 478-495.

Bateson, G. (1972). A theory of play and fantasy. In G. Bateson (Ed.), Steps to an
ecology of mind (pp. 177-193). Aylesbury, England: Chandler.

Baumeister, R. F. (1987). How the self became a problem: A psychological review
of historical research. Journal of Personality and Social Psychology, 52(1),
163-176.

109
Baumeister, R. F. (1992). Neglected aspects of self theory: Motivation, interpersonal
aspects, culture, escape, and existential value. Psychological Inquiry, 3(1),
21-25.

Baumeister, R. F., & Muraven, M. (1996). Identity as adaptation to social, cultural,
and historical context. Journal of Adolescence, 19, 405-416.

Biocca, F. (1997). The cyborg's dilemma: Progressive embodiment in virtual
environments. Journal of Computer Mediated Communication, 3(2).
Retrieved October 1, 2004, from
http://www.ascusc.org/jcmc/vol3/issue2/biocca2.html

Bosson, J. K., Prewitt-Freilino, J. L., Taylor, J. N. (2005). Role rigidity: A problem
of identity misclassification? Journal of Personality and Social Psychology,
89(4), 552-565.

Buss, D. M. (2004). Evolutionary psychology. The new science of the mind (2nd ed.).
Boston: Allyn and Bacon.

Callero, P. L. (1985). Role identity salience. Social Psychology Quarterly, 48, 203-
214.

Cameron, J. E. (2004). A three-factor model of social identity. Self and Identity, 3,
239-262.

Cameron, J. E., & Lalonde, R. N. (2001). Social identification and gender-related
ideology in women and men. Journal of Social Psychology, 40, 59-77.

Carroll, D. W. (1999). Psychology of Language (3rd Ed.). Pacific Grove, CA:
Brooks/Cole Publishing Co.

Clark, H. H. (1996). Using language. Cambridge: Cambridge University Press

Clark, H. H. (1999). How do real people communicate with virtual partners?
Proceedings of 1999 AAAI Fall Symposium, Psychological Models of
Communication in Collaborative Systems, November 5-7th, 1999, North
Falmouth, MA.

Cohen, J. (in press). Audience identification with media characters. In Jennings
Bryant & Peter Vorderer (Eds.), Psychology of entertainment. Mahwah, NJ:
Erlbaum.

110
Collier, P. J., & Callero, P. J. (2005). Role theory and social cognition: Learning to
think like a recycler. Self and Identity, 4, 45-48.

Cosmides, L., & Tooby, J. (1994). Origins of domain specificity: The evolution of
functional organization. In L. Hirschfeld & S. Gelman (Eds.), Mapping the
mind: Domain specificity in cognition and culture (pp. 85-116). New York:
Cambridge University Press.

Costall, A., & Leudar, I. (2004). Where is the ‘theory’ in theory of mind? Theory and
Psychology, 14(5), 623-646.

Csikszentmihalyi, M. (1993). The evolving self: A psychological engagement with
everyday life. New York: Basic Books.

Currie, G. (1998). Pretence, pretending and metarepresenting. Mind & Language, 13(1),
33-55.

Deaux, K. (1984). From individual differences to social categories: Analysis of a
decade’s research on gender. American Psychologist, 39, 105-116.

Dias, M. G. & Harris, P. L. (1988). The effect of make-believe play on deductive
reasoning. British Journal of Developmental Psychology, 6, 207-221.

Dosey, M. A., & Meisels, M. (1969). Personal space and self-protection. Journal of
Personality and Social Psychology, 11, 93-97.

Dreyfus, H. (1992). Being-in-the-world. Cambridge, MA: MIT Press.

Farrar, K. M., Krcmar, M., & Nowak, K. L. (2006). Contextual features of violent
video games, mental models, and aggression. Journal of Communication, 56,
387-405.

Fein, G. G. (1981). Pretend play in childhood: A integrative review. Child Development,
52(4), 1095-1118.

Fein, M. L. (1992). Analyzing psychotherapy: A social role interpretation. New
York, NY: Praeger Publishers.

Feshbach, N. D. (1978). Studies of empathic behavior in children. In B. A. Maher
(Ed.), Progress in experimental personality research (Vol. 8, pp. 1-47). New
York: Academic Press.

111
Floyd, K. (2000). Affectionate same-sex touch: The influence of homophobia on
observers’ perceptions. Journal of Social Psychology, 140(6), 774-788.

Freud, S. (1964 [1933]). New introductory lectures on psycho-analysis. In J.
Strachey (Ed. & Trans.). The standard edition of the complete psychological
works of Sigmun Freud (pp. 7-182). London: Hograph Press.

Garvey, C. (1977). Play. Cambridge, Mass: Harvard University Press.

Gliech, U. (1997). Parasocial interaction with people on the screen. In R. Winterhoff-
Spurk & T. H. A. van der Voort (Eds.). New horizons in media psychology:
Research cooperation and projects in Europe (pp. 35-55). Olpaden, Germany:
Westdeustscher Verlag.

Goffman, E. (1959). The presentation of self in everyday life. New York: Anchor.

Goffman, E. (1961). Encounters: Two studies in the sociology of interaction.
Indianapolis, IN: Bobbs-Merrill.

Goleman, D. (1995). Emotional Intelligence: Why it can Matter More Than IQ. New
York: Bantam.

Goodman, D., Bradley, N., Paras, B., Williamson, I. J., Bizzochi, J. (2006). Video
gaming promotes concussion knowledge acquisition in youth hockey players.
Journal of Adolescence, 29(3), 351-360.

Goossens, L., & Phinney, J. S. (1996). Identity, context, and development. Journal of
Adolescence, 19, 491-496.

Green, M. C. & Brock, T. (2000). The role of transporation in the persuasiveness of
public narratives. Journal of Personality and Social Psychology, 79, 701-721.

Green, M. C., Brock, T. C., Kaufman, G. F. (2004). Understanding media enjoyment:
The role of transportation into narrative worlds. Communication Theory,
14(4), 311-327.

Harris, P. L. (2000). The Work of the Imagination. Malden, MA: Blackwell.

Holden, M. K. (2005). Virtual environments for motor rehabilitation: review.
CyberPsychology and Behavior, 8, 187-211.

112
Horton, D. & Wohl, R. (1956). Mass Communication and para-social interaction:
Observation on intimacy at a distance. Psychiatry, 19, 215-229.

Jackowitz, E. R., & Watson, M. W. (1980). The development of object
transformations in early pretend play. Developmental Psychology, 16, 543-
549.

Johnson, D.W. and Johnson, F.P. (1997). Joining together: Group theory and group
skills (6th Ed.). Boston, MA: Allyn & Bacon.

Kavenaugh, R. D. & Harris, P. L. (1994). Imagining the outcome of pretend
transformations: Assessing the competence of normal children and children
with autism. Developmental Psychology, 30, 847-54.

Klimmt, C. K.. & Hartmann, T. (in press). Parasocial interactions and relationship. In
Jennings Bryant & Peter Vorderer (Eds.), Psychology of entertainment.
Mahwah, NJ: Erlbaum.

Landy, R. J. (1993). Persona and performance. New York, NY: The Guilford Press.

Lee, K. M. (2004a). Presence, explicated. Communication Theory, 14, 27–50.

Lee, K. M. (2004b). Why presence occurs: Evolutionary Psychology, Media
Equation, and Presence. Presence: Teleoperators and Virtual Environments,
13, 494–505.

Lee, K. M., & Jung, Y. (2005). Evolutionary nature of virtual experience. Journal of
Cultural and Evolutionary Psychology, 3(2), 161-180.

Lee, K. M., Jung, Y., Kim, J., & Kim, S. R. (2006). Are physically embodied social
agents better than disembodied social agents?: The effects of physical
embodiment, tactile interaction, and people's loneliness in human-robot
interaction. International Journal of Human Computer Studies, 64, 962-973.

Lee, K. M., Jung, Y., Nass, C. (2007, November). Effects of the user choice on
social responses to computer-synthesized speech. Paper to be presented at the
Annual conference of National Communication Association, Chicago, IL,
U.S.A.

Lillard, A. S. (1993). Pretend play skills and the child’s theory of mind. Child
Development, 64, 348–371.

113
Leslie, A. M. (1987): Pretence and representation. The origins of “theory of mind”.
Psychological Review, 94, 412-426.

Lewis, M., & Ramsay, D. (2004). Development of self-recognition, personal
pronoun use, and pretend play during the 2
nd
year. Child Development, 75(6),
1821-1831.

Lindeman, M., & Sundvik, L. (1995). Evaluative bias and self-enhancement among
gender groups. European Journal of Social Psychology, 25, 269-280.

Linton, R. (1936). The study of man. New York: Appleton-Century.

Lombard, M., Reich, R. D., Grabe, M. E., Bracken, C. C., & Ditton, T. B. (2000).
Presence of television: The role of screen size. Human Communication
Research, 26(1), 75-98.

Markus, H. & Nurius, P. (1986). Possible selves. American Psychologist, 41, 954-
969.

Mead, G. H. (1934). Mind, self and society. Chicago: University of Chicago Press.

Morman, M. T., & Floyd, K. (1998). “I love you man”: Overt expressions of
affection in male-male interaction. Sex Roles, 38(9/10), 871-881.

Moskalenko, S. & Heine, S J. (2003). Watching you troubles away: Television
viewing as a stimulus for subjective self-awareness. Personality and Social
Psychology Bulletin, 29, 76-85.

Nass, C., Moon, Y., & Green, N. (1997). Are computers gender-neutral? Gender
stereotypic responses to computers. Journal of Applied Social Psychology,
27(10), 864-876.

Nichols, S., & Stich, S. (2000). A cognitive theory of pretense. Cognition, 74, 115-
147.

Oately, K. (1994). A taxonomy of the emotions of literary response and a theory of
identification in fictional narrative. Poetics, 23, 53-74.

Oately, K. (1999). Meetings of minds: Dialogue, sympathy, and identification, in
reading fiction. Poetics, 26, 439-454.

114
Ohler, P., & Nieding, G. (in press). Why play? An evolutionary perspective. In
Jennings Bryant & Peter Vorderer (Eds.), Psychology of entertainment.
Mahwah, NJ: Erlbaum.

Onishi, K. H., & Baillargeon, R. (2005). Do 15-month-old infants understand false-
beliefs? Science, 308, 255-258.

Palomares, N. A. (2004). Gender schematicity, gender identity salience, and gender-
linked language use. Human Communication Research, 30(4), 556-588.

Patterson, M. L. (1978). The role of space in social interaction. In A. W. Siegman &
S. Feldstein (Eds.), Nonverbal behavior and communication. New Jersey:
Lawrence Erlbaum Association.

Patterson, M. L., & Edinger, J. A. (1987). A functional analysis of space in social
interaction. In A. W. Siegman & S. Feldstein (Eds.). Nonverbal behavior and
communication. New Jersey: Lawrence Erlbaum Association.

Perloff, R. M. (1993). Third-person effect research, 1983–1992: A review and
synthesis. International Journal of Public Opinion Research 5, 167–184.

Perner, J. & Ruffman, T. (2005). Infants’ insight into the mind: How deep? Science,
308, 214-216.

Piaget, J. (1962). Play, dreams, and imitation in childhood. New York: Norton.

Pinker, S. (1997). How the Mind Works. New York, NY: W. W. Norton and
Company, Inc.

Pinker, S. (2002). The blank slate: The modern denial of human nature. New York,
NY: the Penguin Group

Press, A. L. (1990). Class, gender and female viewer: Women’s responses to
Dynasty. In M. E. Brown (Ed.), Television and women’s culture: The politics
of the popular (pp. 144-157). London: Sage.

Reeves, B. and Nass, C. (1996). The media equation: How people treat computers,
television, and new media like real people and places. New York, NY:
Cambridge University Press.

115
Roese, N. J., Olson, J. M., Borenstein, M. N., Martin, A., & Shores, A. L. (1992).
Same-sex touching behavior: The moderating role of homophobic attitudes.
Journal of Non-verbal Behavior, 16, 249-259.

Rubin, A., Perse, E., & Powell, A. (1985). Loneliness, parasocial interaction, and
local television news viewing. Human Communication Research, 12, 155-
180.

Rubin, K. H. (1976). The relationship of social play preference to role taking skills in
preschool children. Psychological Reports, 39, 823-826.

Schlovsky, V. (1917). Art is technique. In D. Lodge (Ed.), Modern criticism and
theory (pp. 16-30). London: Longman.

Shaw, M. E., & Costanzo, P. R. (1982). Theories of social psychology (2nd ed.).
New York: McGraw-Hill

Sheridan, T. B. (1995). Teleoperation, telerobotics and telepresence: A progress
report. Control Engineering Practice, 3, 205-214.

Sherry, J. L. (2004). Flow and media enjoyment. Communication Theory, 14(4),
328-347.

Sogunro, O. A. (2004). Efficacy of role-playing pedagogy in training leaders: Some
reflections. Journal of Management Development, 23, 355-371.

Sperber, D. (2000). Introduction. In D. Sperber (Ed.). Metarepresentations: A
multidisciplinary perspective. New York: Oxford University Press.

Steen, F. F., & Owens, S. A. (2001). Evolution’s pedagogy: An adaptationist model
of pretense and entertainment. Journal of Cognition and Culture, 1(4), 290-
321.

Stryker, S., & Serpe, R. T. (1982). Commitment, identity salience, and role behavior:
Theory and research example. In W. Ickes and E. Knowles (Eds.).
Personality, roles, and social behavior (pp. 199-218). New York, NY:
Springer-Verlag, Inc.

Sussman, N. M., & Rosenfeld, H. M. (1982). Influence of culture, language and sex
on conversational distance. Journal of Personality and Social Psychology, 42,
66-74.

116
Tajfel, H., & Turner, J. C. (1986). The social identity theory of intergroup conflict.
In W. G. Austin & S. Worchel (Eds.). Psychology of intergroup relations (pp.
7-24). Chicago: Nelson-Hall.

Tan, E. S. (1994). Film-induced affect as a witness emotion. Poetics, 23, 7-32.

Tice, D. M. (1992). Self-concept change and self-presentation: The looking glass self
is also a magnifying glass. Journal of Personality and Social Psychology,
63(3), 435-451.

Uzzell, D., & Horne, N. (2006). The influence of biological sex, sexuality and
gender role on interpersonal distance. British Journal of Social Psychology,
45, 579-597.

Viau, A., Feldman, A. G., McFadyen, B. J., & Levin, M. F (2004). Reaching in
reality and virtual reality: A comparison of movement kinematics in healthy
subjects and in adults with hemiparesis. Journal of NeuroEngineering and
Rehabilitation, 1 (11), 1-7.

Vorderer, P. (2001). It’s all entertainment−sure. But what exactly is entertainment?
Communication research, media psychology, and the explanation of
entertainment experiences. Poetics, 29, 247−261.

Vorderer, P., Klimmt, C., Ritterfeld, U. (2004). Enjoyment: At the heart of media
entertainment. Communication Theory, 14(3), 388-408.

Westen, D. (1992). The cognitive self and the psychoanalytic self: Can we put our
selves together? Psychological Inquiry, 3(1), 1-13.

Williams, D., & Skoric, M. (2005). Internet fantasy violence: A test of aggression in
an online game. Communication Monographs, 72(2), 217-233.

Wirth, W., Hartmann, T., Böcking, S., Vorderer, P., Klimmt, C., Schramm, H., & et
al. (in press). A process model of the formation of spatial presence
experiences.

Yee, N., Bailenson, J. N. (in press). The Proteus effect: The effect of transformed
self-representation on behavior. Human Communication Research.

117
Yee, N., Bailenson, J. N., Urbanek, M., Chang, F., Merget, D. (in press). The
unbearable likeness of being digital: The persistence of nonverbal social
norms in online virtual environments. The Journal of CyberPsychology and
Behavior.

Zillmann, D. (1994). Mechanisms of emotional involvement with drama. Poetics, 23,
33-51.

Zillmann, D. (in press). Empathy: Affective reactivity to others’ emotional
experiences. In Jennings Bryant & Peter Vorderer (Eds.), Psychology of
entertainment. Mahwah, NJ: Erlbaum.

118
Appendix A: A Screen Shot of the Virtual Bar (the Male-Role
Condition with the First-Person View)

119
Appendix B: A Screen Shot of the Virtual Bar (the Female-Role
Condition with the First-Person View)

120
Appendix C: A Screen Shot of the Virtual Bar (the Male-Role
Condition with the Third-Person View)

121
Appendix D: A Screen Shot of the Virtual Bar (the Female-Role
Condition with the Third-Person View)

122
Appendix E: Male-Role Script (Participants Played the Role of
Albert)

Albert: Hi Michael!

Michael: Hey, what’s up, Albert?

Albert: Not much. Thank you for coming, Michael. I need your help.

Michael: Are you okay? You look very serious. What’s going on?

Albert: Well, it’s about a girl.

Michael: Oh yeh? Tell me about the story.

Albert: She just got out of a relationship. And I don’t know. I am a little
uncomfortable with dating. You know, I had a lot of bad experiences and been
hurt a lot. Some good ones, but definitely a lot of bad ones…

Michael: Come on, Albert! You are a good guy. So, tell me about her. How did you
meet her?

Albert: Actually, I was buying pajamas for my mother.

Michael: And by that, of course, you mean… you were buying lingerie for another
woman.

Albert: Yes, But you can’t help where you meet somebody. Can you? By the
way, the girl I am talking about gave me her number. And that’s how it
happened and we went out once so far.

Michael: Did you kiss her?

Albert: No! Not on the first date…

Michael: Eight out of 10 women believe that the first kiss will tell them everything
they need to know about a relationship. And believe me, she has definitely thought
about it.

Albert: She has?

123
Michael: Oh, of course. Not that she is going to act on it. But the first kiss is a
monumental big deal for women. I need you wrap your head around this. On your
next date, the girl… what is her name?

Albert: Nicole.

Michael: Okay, Nicole could have her “LAST” first kiss on her next date.

Albert: Wow…how do you do that? Give me a tip, Michael!

Michael: Albert, end of the night, you’re dropping her off at home. Do not miss a
signal. She might be fiddling with her keys. Here is a secret. A woman that does not
want a kiss takes her keys out, put them in the door, goes in the house. A woman that
wants to kiss, she fiddles.

Albert: Wow, how, how do you know all this?

Michael: Because I know… Another advice is that most of guys rush in to take the
kiss. But you are not most guys. The secret to kiss is to go to 90% of the way and
then hold. Wait as long as it takes for her to come the other 10%.

Albert: Okay.. 90-10%. I got it. By the way, how come you are still single? I
mean you have not been in a serious relationship for a long time.

Michael: I am not looking for any kind of serious relationship at all right now. I kind
of enjoy playing games.

Albert: You know what your problem is, Michael? Your’re all about the short
game. You pick your shots based on what you see first. Not what’s necessarily
best for you in the long run.

Michael: All of us are not married to the woman of our dreams and about to have a
baby. It is just not meant for everybody. So please just leave me to my hot, sweaty,
totally varied, wildly experimental short game.

Albert: Honestly, I just hope one day you’re able to experience unconditional
love, and trust, and openness.

Michael: Is this really barroom talk?

Albert: You need to listen to me, man. I am serious. Because when you get to a
place with a woman like that, it’s so beyond anything physical… that when I

124
think back to when I used to run around with you and chase all these really
gorgeous but, you know, shallow women… I don’t know, it’s kind of ridiculous
and vaguely pathetic.

Michael: Come on, I was once in a serious relationship, you know, with Ashley.
When we broke up, I went out of my mind. I wanted to throw myself off of every
building in Los Angeles. When I saw a cab, I wanted drive in front of it, then I would
stop thinking about her… After having gone through all that misery, I learned my
lessons. I will just play around without committing myself to a relationship. And it
works for me. Let’s talk more about the girl you are in love with.

Albert: Have you met someone… and you know right away she was gonna be
important to you? Not just because of her looks, but that X factor… She’s so
sweet, funny, southern. I don’t know what it is about her. I just can’t get her out
of my mind. You know, food has lost its taste. Colors, they seem dull. Things
that used to matter… I don’t know, they just no longer do.

Michael: Man, you are really falling in love with her. I am definitely not a person
that you want to get an advice from in this case. But I can hook you up with another
girl… Think about it until I’m back from restroom.

125
Appendix F: Female-Role Script (Participants Played the Role of Ashley)

ASHLEY: Hi, NICOLE! Thank you for canceling your meeting to see me.

NICOLE: Don’t mention it. I am sorry that James cheated on you. Tonight is just for
us without men

ASHLEY: If I end up old and alone, it is all your fault.

NICOLE: Sweetie, we are all alone even when we are with men. My advice to you is
to embrace that fact… slap on some armor and go through life like I do: Enjoying
men but not expecting them to fill you up.

ASHLEY: Yeh, men cheat for the same reason that dog licks himself. Because
they can. It’s part of their biology. Instead of wasting all this energy
condemning it…maybe it’s time we all get in line with the reality of the situation.

NICOLE: That sounds very empowering, but you’re forgetting one detail. Women
cheat too.

ASHLEY: Yeh, but it’s completely different.

NICOLE: How?

ASHLEY: Because we don’t go around randomly attacking any man we’re
attracted to. We’re not driven by testosterone.

NICOLE: Then, what does drive us according to you?

ASHLEY: Emotions.

NICOLE: You mean hormones.

ASHLEY: No. I mean that little voice inside of me that says, “Mate for life,
mate for life.”

NICOLE: Sweet heart, you can’t listen to every little voice that runs through your
head. It will drive you nuts! The fact is, the act of cheating is defined by the act of
getting caught. One doesn’t exist without the other.

126
ASHLEY: What are you talking about? Are you saying that cheating is
acceptable? Am I expecting too much?

NICOLE: Oh, no no… no.

ASHLEY: Okay. Enough of my story. What about you? Are you seeing anyone,
yet?

NICOLE: Well….

ASHLEY: Yes?

NICOLE: It’s something that started again a few weeks ago. I don’t know if it’s real
so I didn’t want to say anything.. But I am seeing Randy.

ASHLEY: Oh my god! You are seeing Randy again? He was so bad to you! He
broke your heart!

NICOLE: Not really… Maybe, sometimes….

ASHLEY: What makes you think it’s gonna be any different this time?

NICOLE: I don’t know. I’m not sure it is, but it just… kind of feels okay.

ASHLEY: If it feels okay, why are you sneaking behind my back?

NICOLE: Look, we’ve got this chemistry… kind of connection that’s hard to shake.
So, lay off, okay?

ASHLEY: So, now it’s gonna be a casual sex thing?

NICOLE: Maybe

ASHLEY: That’s gonna work… Even I am not that naïve.

NICOLE: You know what? Relationships have declined since women came out of
the cave... looked around and said, “This isn’t so hard.”

ASHLEY: Okay, so you don’t need a man, but do you still want one?

NICOLE: Oh, honey, I want more than one.

127
ASHLEY: I can’t decide whether you represent our future or our demise.

NICOLE: I am the future! In fact, seeing Randy reminded me of how needy I was
and how far I’ve come. We are going out Friday night

ASHLEY: Sweetie! It took you a year to get over him. He treated you like shit.
Why do you wanna go out with him again?

NICOLE: Because he treated me like shit.

ASHLEY: Okay, I’m leaning toward demise.

NICOLE: It’s called revenge. I’m gonna make him want me and I’m gonna drop him
like he dropped me.

ASHLEY: As long as you’ve got a plan.

NICOLE: Not only a plan, I’ve got a dress! You know, I don’t believe the
relationship myth… Unbelievable fairy tales concocted by women to make their love
lives seem less hopeless.

ASHLEY: Except it makes us feel even more hopeless because this fabulous,
magical relationship is never happening to us.

NICOLE: Now, we are talking! Hold on a second… I need to use lady’s room.

.

128
Appendix G: Measures

Instruction: How much do you agree with the following statements? For each
question, please circle the dot that best describes your experience in the virtual
reality system.

I have a lot in common with the virtual character (Albert/Ashley) that I acted out.
Strongly Disagree Strongly Agree
• • • • • • •

I feel strong ties to the virtual character (Albert/Ashley) that I acted out.
Strongly Disagree Strongly Agree
• • • • • • •

I find it difficult to form a bond with the virtual character (Albert/Ashley) that I acted
out.
Strongly Disagree Strongly Agree
• • • • • • •

I don’t feel a sense of being “connected” with the virtual character (Albert/Ashley)
that I acted out.
Strongly Disagree Strongly Agree
• • • • • • •

In general, I’m glad to act out a male/female role in the virtual reality system
Strongly Disagree Strongly Agree
• • • • • • •

I regret that I acted out a male/female role in the virtual reality system.
Strongly Disagree Strongly Agree
• • • • • • •

I don’t feel good about acting out a male/female role in the virtual reality system
Strongly Disagree Strongly Agree
• • • • • • •

Generally, I feel good when I think about myself as a male/female role in the virtual
reality system.
Strongly Disagree Strongly Agree

129
• • • • • • •

Instruction: How much do you agree with the following statements? Please
circle the dot that best describes your answer.

I often think about the fact that I am a male/female.
Strongly Disagree Strongly Agree
• • • • • • •

Overall, being a male/female has very little to do with how I feel about myself.

Strongly Disagree Strongly Agree
• • • • • • •

In general, being a male/female is an important part of my self-image.
Strongly Disagree Strongly Agree
• • • • • • •
The fact that I am a male/female rarely enters my mind.
Strongly Disagree Strongly Agree
• • • • • • •

Instruction: How much do you agree with the following statements? For each
question, please circle the dot that best describes your answer.

In general, I like to try new technology products before they become popular.
Strongly Disagree Strongly Agree
• • • • • • •

In general, I am very excited by new technology products
Strongly Disagree Strongly Agree
• • • • • • •

In general, I like to know information about new technology products.
Strongly Disagree Strongly Agree
• • • • • • •

In general, I like to introduce new technology products to other people.
Strongly Disagree Strongly Agree
• • • • • • •

130

What is your gender?
Male _______ Female _______

What is your age?
__________

What is your ethnicity?

_________________

Abstract (if available)

Abstract Role play prevails in our daily activities from working in a company to playing a videogame. In this regards, conceptualizing the role-play process helps us understand how people interact with other people in both face-to-face and computer-mediated environments, as well as how such social interactions could affect people's self-concept and corresponding behaviors in particular ways. The purpose of my dissertation is to propose a new theoretical model of role play to explicate our virtual experience in media, to discuss theoretical implications of the model, and to provide empirical evidence for the utility of the role play perspective in understanding the effects of role enactment in interactive media. In the first three chapters I discuss a new theoretical framework for the multi-layering model of role enactment. Particularly, I review the literature on social psychology and communication to address the prevalence of role play in our daily activities and the relationship between role play and changes in the self-concept and behaviors in Chapter 1. In Chapter 2, I conceptualize a multi-layering model of role enactment and explain how we engage in role play in videogames based on Clark's multi-layering model (Clark, 1996) and Leslie's metarepresentations (Leslie, 1987). In Chapter 3, I elaborate more on the multi-layering model of role play to enhance our understandings of entertainment experience based on the review of important aspects from the seven entertainment theories. In the final chapter, I report on the results of two experiments in order to provide preliminary empirical evidence to support the utility of the role-play perspective. Not only do the results confirm the usefulness of a new role-play perspective to understand our psychological and behavioral reactions from the role enactment in interactive media, but also they suggest that altering contextual features of interactive media could influence the degree to which people conform to role play.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

The Kronos Project: techniques for streamlining complexity and increasing player immersion in role playing games

PDF

Good Girl VR: straddling installation art, performance art, and identity politics with a virtual environment

PDF

Building personal wellness communities: meaningful play in the everyday life of a network society

PDF

STARWEAVE: a north star towards empathy.

PDF

Understanding human-building-emergency interactions in the built environment

PDF

Cognitive science, belief, and interactive media as a distribution platform for experiences that enhance human development

PDF

Computational narrative models of character representations to estimate audience perception

PDF

Bounded technological rationality: the intersection between artificial intelligence, cognition, and environment and its effects on decision-making

PDF

The role of social media for community-based organizations focused on environmental justice in southeast Los Angeles

PDF

Detecting joint interactions between sets of variables in the context of studies with a dichotomous phenotype, with applications to asthma susceptibility involving epigenetics and epistasis

PDF

Exploring the intersection of occupational engagement and well-being in student musicians: A multi-method study

PDF

Building and validating computational models of emotional expressivity in a natural social task

PDF

Toward counteralgorithms: the contestation of interpretability in machine learning

Asset Metadata

Creator Jung, Younbo (author)

Core Title Role enactment in interactive media: a role-play perspective

School Annenberg School for Communication

Degree Doctor of Philosophy

Degree Program Communication

Publication Date 07/24/2007

Defense Date 05/09/2007

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag entertainment theory,evolutionary psychology,gender,interactive media,interpersonal distance management,multiple layers,OAI-PMH Harvest,pretend play,role enactment,role identity,role play,self-concept,theory of mind,virtual reality

Language English

Advisor McLaughlin, Margaret L. (committee chair), Lee, Kwan Min (committee member), Rizzo, Albert Skip (committee member), Vorderer, Peter (committee member)

Creator Email younboju@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-m657

Unique identifier UC170584

Identifier etd-Jung-20070724 (filename),usctheses-m40 (legacy collection record id),usctheses-c127-522474 (legacy record id),usctheses-m657 (legacy record id)

Legacy Identifier etd-Jung-20070724.pdf

Dmrecord 522474

Document Type Dissertation

Rights Jung, Younbo

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Repository Name Libraries, University of Southern California

Repository Location Los Angeles, California

Repository Email cisadmin@lib.usc.edu