Using Avatars and Agents to Promote Real World Health Behavior Changes

Abstract

This chapter reviews the past and current trends of one particularly intriguing aspect of virtual worlds—virtual representations commonly known as avatars—in the context of health behavior change. Despite the growing interest in the influence of virtual representations on health behaviors, there has been a surprising dearth of research exploring the use of virtual representations as a direct and central vehicle of behavior change. State-of-the-art findings on the use of virtual representations to promote behavior change will be discussed first, followed by a more detailed discussion of recent studies that specifically target health behavior change using virtual representations.

Full text

Wiley STM / Combs, Sokolowski, Banks: The Digital Patient: Advancing Medical Research,
Education, and Practice Chapter 12/ Ahn / filename: ch12.doc
[The Digital Patient: Advancing Medical Research, Education, and Practice], Edited by [C. Donald
Combs, John A. Sokolowski, Catherine M. Banks].
ISBN 0-471-XXXXX-X Copyright © 2015 Wiley[Imprint], Inc.
Chapter 12
Using Avatars and Agents to Promote
Real World Health Behavior Changes
Sun Joo (Grace) Ahn
Introduction
Fluid, dynamic, and infinitely replicable, virtual worlds have been an enticing yet
prohibitively costly platform for health behavior researchers in the past decade. What
difficulties that the researchers of the past have faced in incorporating virtual worlds in
primary and secondary prevention and intervention programs have dramatically
decreased in the recent years with unprecedented advancement in digital media
technology [1] and virtual worlds are enjoying a newfound surge of popularity in both
academic and clinical environments. Digital devices that simulate vivid sensory
information, allowing users in virtual worlds to experience visceral environmental cues,
have now become much more affordable and accessible, opening up new horizons for
incorporating virtual reality technology in people’s everyday lives, outside of
sophisticated laboratories.
This chapter reviews the past and current trends of one particularly intriguing
aspect of virtual worlds—virtual representations commonly known as avatars—in the
context of health behavior change. Despite the growing interest in the influence of
virtual representations on health behaviors, there has been a surprising dearth of research

Wiley STM / Combs, Sokolowski, Banks: The Digital Patient: Advancing Medical Research,
Education, and Practice Chapter 12/ Ahn / filename: ch12.doc
exploring the use of virtual representations as a direct and central vehicle of behavior
change. State-of-the-art findings on the use of virtual representations to promote
behavior change will be discussed first, followed by a more detailed discussion of recent
studies that specifically target health behavior change using virtual representations.
Avatars and Agents
The word “avatar” originates from the Sanskrit word avatara, which means “descent” to
describe an incarnation or a bodily manifestation of an immortal being in Hinduism. In
much the same way, users interact in the virtual world in the form of embodied virtual
identities that mark their presence in the virtual environment [2]. In the past, avatars
typically served as simplistic and static visual markers (e.g., a simple chat icon on AOL
or Yahoo Messenger), much like the virtual equivalent of horses players use to represent
themselves in board games. Over time, avatars have become significantly more complex,
rendered in three dimensional forms with an extensive range of dynamic movements,
photorealistic appearances, naturalistic language, and even the ability to mimic empathy
when interacting with users. The avatars of today are still works in progress—the
humanoids that have evolved to feel and express naturalistic emotions as are often
depicted in popular media are yet figments of imagination. However, the speed at which
avatars have gained technical sophistication forecasts that more realistic, natural, and
affordable avatars may soon become a reality in the near future.
Agents are another form of virtual representation that shares similar features and
capabilities with an avatar, but the two forms are distinguished by the element of control:
avatars are controlled by human users whereas agents are controlled by computer

Wiley STM / Combs, Sokolowski, Banks: The Digital Patient: Advancing Medical Research,
Education, and Practice Chapter 12/ Ahn / filename: ch12.doc
algorithms [3]. Although seemingly similar in their specifications, agents and avatars
yield meaningfully different influences in their interactions with human users [4].
Research studies have demonstrated that the mere perception of interacting with another
human (vs. a computer algorithm) meaningfully affects whether a virtual representation is
successful at influencing an individual’s attitude and behaviors even when the agents and
avatars are performing identical tasks at the same level [5, 6].
A recent meta-analysis examining 32 studies that compared the influence of
agents against avatars concluded that the mere perception of human control elicited
stronger social responses from humans than the perception of machine control [7]. In
particular, this agency effect was stronger when humans were required to form a certain
degree of relationship with the virtual representations by engaging in a competitive or
cooperative task, rather than a neutral task. The effect of agency was also stronger when
the virtual representations were actually controlled by a human rather than a machine,
regardless of perceived agency.
Designing Agents and Avatars for Health Programs
These findings have important implications for the design and implementation of virtual
representations in health prevention and intervention programs. First, avatars that are
controlled by humans are likely to have stronger impacts on health behavior change than
agents that are controlled by machines. Thus, rather than an agent providing a heavily
scripted intervention, an avatar delivering naturalistic responses is likely to be much more
effective in changing health behaviors.

Wiley STM / Combs, Sokolowski, Banks: The Digital Patient: Advancing Medical Research,
Education, and Practice Chapter 12/ Ahn / filename: ch12.doc
Having an actual person control a virtual representation may be useful in a variety
of health contexts. For instance, a rich collection of literature points to the fact that
individuals often judge others based on nonverbal cues such as physical appearance or
behavior [8-10]. Indeed, people are often drawn to others perceived as similar to
themselves [11, 12] or simply familiar (e.g., celebrities) [13, 14]. Because virtual
representations afford labor- and cost-effective means of adopting almost a limitless
option of physical appearances and behaviors that may be different from the actual self,
human controllers behind the virtual representation may flexibly adapt to tailor their
avatars for each respective interactant.
Such a scenario is then possible: an Asian female health care provider is
controlling an avatar. A patient of different sex and ethnicity walks in. The health care
provider may then select a male avatar of the patient’s ethnicity and interact with the
patient through the male avatar to elicit positive responses based on perceived similarity.
Similarity is not necessarily confined to physical appearances; the health care provider
may also mimic the behavior of the patient. Behavioral mimicry has been found to
increase a host of favorable attitudes toward the interaction partner, including liking [15].
Alternatively, the avatar could take on the physical appearance of a well-known celebrity
to deliver health messages, and the perceived familiarity triggered by the avatar may
favorably impact persuasion as demonstrated in earlier studies [14].
Despite earlier findings that indicate the superiority of avatars over agents in a
persuasion context, having human controllers positioned for each and every avatar is
likely a costly option. A more labor- and cost-effective solution may be to implement
computer controlled agents. Although the initial development and setup of the agent

Wiley STM / Combs, Sokolowski, Banks: The Digital Patient: Advancing Medical Research,
Education, and Practice Chapter 12/ Ahn / filename: ch12.doc
might be costly, once the infrastructure is established, agents can continue to work at the
same speed and efficiency without the need to eat or rest. These agents may be infinitely
replicable, which would allow patients to receive equal and uniform care across all health
care facilities. Also, for sensitive topics of discussion (e.g., sexual history), patients may
prefer to interact with a machine controlled agent rather than a human controlled avatar
for greater perceived privacy.
Thus, both agents and avatars are viable options for incorporation in health
prevention and intervention programs and the choice to use one over the other should be
made after a careful cost-effectiveness analysis [16]. Although research demonstrates
that avatars are more effective in promoting desired behaviors than agents, the associated
costs in employing a human controller may be prohibitive in a large scale program. On
the other hand, if the program calls for strong and impactful interventions that are more
likely to elicit behavior change, avatars may be well worth the costs. The following
section discusses representative case studies on the systematic investigation of using
agents and avatars in the context of health behavior change to consider the costs and
benefits associated with each type of virtual representative.
Using Agents and Avatars to Promote Health Behavior
Changes
Virtual representations offer a set of novel characteristics that allow researchers and
practitioners to implement new strategies to approach health behavior change that were
difficult or not possible with traditional tools and platforms. These characteristics
include the virtual acceleration of time [17], wherein agents and avatars are able to
transcend temporal boundaries of the physical world to digitally depict events in the past

Wiley STM / Combs, Sokolowski, Banks: The Digital Patient: Advancing Medical Research,
Education, and Practice Chapter 12/ Ahn / filename: ch12.doc
or future from the present point in time. Another major characteristic that distinguishes
virtual representations from traditional media platforms is high interactivity, or the
medium’s capacity that allows users to influence the form and/or content of the mediated
experience in real time [18]. It is generally agreed that the best interactive medium
mimics the interactive dynamics of face-to-face communication [19]. Using interactive
virtual representations that afford naturalistic social interactions is likely to heighten
users’ engagement and involvement [20], as well as foster more positive attitudes or
liking towards the interaction content [21].
The initial scientific foray into using agents and avatars to impact behavior in the
physical world began with the Proteus Effect [22, 23], which found that when individuals
are aware of the visual characteristics and traits of an avatar that he or she is embodying,
and understand what others will expect of their behaviors based on those characteristics
and traits [24], the experience of embodying that particular avatar will encourage them to
continue to engage in the expected behaviors in the physical world. For example, when
individuals were given tall avatars to embody in a virtual world, they were more
aggressive during a negotiation task in the physical world than individuals who were
given shorter avatars. This is likely a result of conforming to normative expectations that
taller people are more confident and more likely to be in positions of power [25].
Combined with the novel media characteristics that afford users some experiences
that were difficult or impossible with traditional media platforms, the ability of agents
and avatars to influence behaviors in the physical world introduces new horizons for
research on the incorporation of virtual representations in health promotion campaigns as
vehicles of behavior change. The implication is that even the most creative and

Wiley STM / Combs, Sokolowski, Banks: The Digital Patient: Advancing Medical Research,
Education, and Practice Chapter 12/ Ahn / filename: ch12.doc
fantastical virtual experiences with agents and avatars may yield effects that transfer into
the physical world to affect individuals’ attitudes and behaviors, giving researchers the
freedom to explore any health context of their desires. Despite such potentials, few
studies have systematically investigated agents and avatars in the context of health
behavior change. Among the collection of literature on agents and avatars in the health
realm, the following sections review some of the representative and state-of-the-art
studies that specifically focus on how interactions in the virtual world with virtual
representatives affect health behaviors in the physical world.
Vicariously Experiencing Future Negative Health
Consequences through Agents and Avatars
Individuals are likely to hold a “rosy” view of distant futures, conceptualizing their
futures in a positive light. Kahneman and Lovallo [26] attribute this overly optimistic
assessment of future outcomes to an isolation error: people tend to think of the future as
an isolated event, independent of past and present events, and base their forecasts of the
future on plans and scenarios of success rather than on accurate past results. Greater
temporal distance of an event, relative to the present, enhances the perceived positivity
towards it [27]. Because future negative health consequences may take some time to
manifest following present behaviors, the large temporal distance is likely to encourage
unrealistic and inaccurate levels of optimism in thinking about the health issue. For
instance, smoking a cigarette today will not immediately lead to lung problems the next
day; rather, the detrimental effect of smoking may require years to manifest. The
temporal distance between the cause (smoking) and effect (lung problems) renders this
relationship abstract and opaque, leading individuals to assume an optimistic outlook for

Wiley STM / Combs, Sokolowski, Banks: The Digital Patient: Advancing Medical Research,
Education, and Practice Chapter 12/ Ahn / filename: ch12.doc
their own health in the future. Consequently, this “rosy view” phenomenon is one major
barrier to successfully communicating health risks and changing present health behaviors.
One solution to the rosy view phenomenon is personal experience. Studies have
demonstrated that going through the actual experience at the moment reduces the
unrealistic level of optimism that individuals perceive about future events [28] and that
recent experiences are given more weight in deciding one’s susceptibility to risk than
distant experiences [29]. However, with future negative health consequences, it would
be unrealistic to prescribe personal experiences of negative health outcomes (e.g., lung
problems) to reduce overly optimistic future forecasts. These negative health outcomes
are often irrevocable or fatal.
Using agents and avatars to digitally render future negative health consequences
allows individuals to vicariously, but vividly, experience negative outcomes without
having to incur actual damages to their physical and mental health. In the virtual world,
time becomes a more fluid concept; once created, an agent or an avatar may be digitally
manipulated to dynamically shift their appearances. For example, a virtual representation
of an individual in his or her 20s may be created and then rapidly aged to depict the same
person in his or her 60s [30]. Although 40 years has passed in the virtual world, this may
take only a few seconds in the physical world. More importantly, these virtual
experiences are sufficiently realistic to influence behaviors in the physical world.
In one of the first studies to observe how using agents and avatars to virtually
depict future negative health consequences may influence health relevant behaviors in the
physical world, participants were asked to watch a virtual simulation of an agent with
photorealistic resemblance to themselves gain weight by eating candy or lose weight by

Wiley STM / Combs, Sokolowski, Banks: The Digital Patient: Advancing Medical Research,
Education, and Practice Chapter 12/ Ahn / filename: ch12.doc
eating carrots [31]. After watching the virtual simulation, participants were subjected to
an unobtrusive measure of candy consumption in the physical world. Candy
consumption was influenced by an interaction of two variables: sex and presence, or the
degree to which participants felt that they were right there in the virtual world [32]. Male
participants who felt high presence ate more candy whereas female participants who felt
high presence suppressed this behavior and ate less candy. The authors posited that this
finding was a result of virtual imitation, wherein participants modeled the behavior he or
has observed from his or her agent, particularly when the perceived presence was high.
However, as the agents were shown to eat both candies and carrots, it was not clear which
observed virtual behavior was affecting candy consumption in the physical world.
Building on these preliminary results, a recent set of studies investigated the
transfer of virtual world effects to physical world behaviors in the context of soft drink
consumption [33-35]. If the earlier study posited that individuals would model and
imitate the behaviors observed by the agent, this set of studies argued that agents and
avatars may be used to reduce two types of psychological distances—temporal and social.
Reducing the perceived temporal distance between the present health behavior and the
future negative health consequence is likely to render the causal relationship concrete
[36]. Also, reducing the perceived social distance between the individual and the given
health issue is likely to promote perceived personal relevance and involvement with the
issue [36].
The results from this set of studies revealed that perceived social distance could
be successfully reduced by tailoring the information to the audience [33]. Tailoring may
be as simple as changing the verbiage in a traditional pamphlet to create the illusion that

Wiley STM / Combs, Sokolowski, Banks: The Digital Patient: Advancing Medical Research,
Education, and Practice Chapter 12/ Ahn / filename: ch12.doc
the pamphlet was created specifically to target an individual. In virtual worlds, an agent
or an avatar may be tailored to bear photorealistic resemblance to an individual, so that
the individual may feel as if the vicarious virtual experiences are actually happening to
him or her. Tailoring in both modalities reduced social distances, increased levels of
involvement with the health issue at hand, and ultimately led to greater intentions to
adopt the desired health behavior (i.e., reducing soft drink consumption) immediately
following experimental treatments [33].
The results also revealed that perceived temporal distance could be successfully
reduced by coupling traditional health pamphlets with virtual simulations that feature
agents and avatars depicting future negative health consequences [33]. A virtual
simulation was created to show an agent dynamically gaining weight as it continued to
consume soft drinks in a virtual world, wherein two minutes of virtual time was
equivalent to two years of physical time. By accelerating the passage of physical time in
the virtual world, participants who were exposed to the virtual simulation perceived
shorter temporal distances between their present health behaviors and future health
consequences. The reduced temporal distance, in turn, increased the perceived
imminence of risks related with soft drink consumption, ultimately leading to lower
consumption of soft drinks one week following the experimental treatment compared to
participants who were not exposed to the virtual simulation. At this point, the effect of
tailoring that was observed immediately following experimental treatments dissipated
and only the effect of watching the virtual simulation remained influential.
A following study explored the effect of virtual doppelgängers [37], agents with
photorealistic resemblance to individuals, to investigate the underlying mechanisms

Wiley STM / Combs, Sokolowski, Banks: The Digital Patient: Advancing Medical Research,
Education, and Practice Chapter 12/ Ahn / filename: ch12.doc
driving health behavior change in the same context of soft drink consumption [34].
Virtual doppelgängers create an interesting social phenomenon wherein a virtual entity
that has photorealistic resemblance to an individual may look like an individual but not
act like him or her because the agent is being controlled by an algorithm. Participants in
the study were exposed to a virtual simulation showing either virtual doppelgängers or an
unfamiliar agent gaining weight as a result of consuming soft drinks regularly for two
years, depicted in two minutes in the virtual world. Results indicated that virtual
doppelgängers were more effective than unfamiliar agents in increasing the perception of
presence as well as self-relevant thoughts in the virtual simulation. Watching an agent
that looks like the self consume soft drinks and become obese made participants feel as if
he or she were truly undergoing the experience and encouraged them to think about
themselves in the context of soft drink consumption. Heightened presence and self-
relevant thoughts, in turn, led to increased personal relevance to the issue of soft drink
consumption and obesity.
Finally, different modalities used to deliver a health message about soft drink
consumption and obesity were compared to determine the most effective message
modality in the promotion of health behaviors [35]. Results indicated that compared to
strictly statistical information, print narratives, and pictures, the virtual simulation of an
agent gaining weight as a result of soft drink consumption over the years was best able to
highlight the risks involved with soft drink consumption and actually reduce consumption
one week following experimental treatments.
Taken together, these studies indicate that agents and avatars may serve as a
powerful vehicle of health behavior change by depicting future negative health

Wiley STM / Combs, Sokolowski, Banks: The Digital Patient: Advancing Medical Research,
Education, and Practice Chapter 12/ Ahn / filename: ch12.doc
consequences. Without incurring actual harm to personal health, individuals are able to
observe sufficiently realistic simulations of what the future might have in store for them
if they were to continue their present health behaviors. The observation of accelerated
changes in the virtual representations’ health is able to meaningfully reduce the temporal
and social distances perceived between the health risk and the self. Consequently,
individuals feel that the risk may be more imminent and more personally relevant than
they had originally thought, and ultimately adopt desirable health behaviors in the
physical world.
Interactive Agents and Avatars for Health Behavior
Change
If merely observing the vicarious experience of future negative health consequences
occurring to agents and avatars are powerful enough to change health behavior, the
ability to directly interact with the agent or avatar is likely to amplify these favorable
effects. In one of the earliest studies looking at how interactions with avatars could lead
to differences in health behaviors in the physical world, participants were given either a
photorealistically similar self avatar or an unfamiliar avatar to interact with in a virtual
world [38]. When the participants exercised in the physical world, the avatar exercised
with them using synchronous head and body movements in the virtual world. Results
indicated that when participants interacted with a self avatar that exercised with them,
they engaged in more exercising than when they interacted with an unfamiliar avatar.
These effects persisted for up to 24 hours following the experimental treatments.
In the past, such studies had to be conducted in a highly controlled laboratory
setting to deliver interactive experiences with a virtual representation because the

Wiley STM / Combs, Sokolowski, Banks: The Digital Patient: Advancing Medical Research,
Education, and Practice Chapter 12/ Ahn / filename: ch12.doc
experimental set up required state-of-the-art digital devices to track and render
participants’ movements. Recently, however, the development of consumer grade
electronics, such as video game consoles, has gradually increased the accessibility and
affordability of interactive media, allowing individuals to interact with virtual
representations in the comforts of their own living rooms [39]. The newer video game
consoles such as the Microsoft Kinect Xbox and the Nintendo Wii are equipped with
sensors and accelerometers that allow players to use naturalistic body movements to
control their avatars in the game. This development has introduced a novel genre of
gaming called exergames, which require players to use body movements to progress
through the game [40]. Several recent studies have demonstrated that interacting with the
avatars in these exergames results in increased physical activity [41] as well as weight
reduction [42], particularly when playing with others rather than playing alone. Although
exergames still contribute to the overall number of hours individuals spend in front of
screens, which is positively linked to negative health outcomes [43], they help to
substitute what would otherwise have been completely sedentary screen time with low to
moderate levels of physical activity [44].
Although the bulk of studies looking at agents and avatars fail to reflect this, not
all virtual representations are required to take on human forms. In one of the few studies
that explored the effect of non-human virtual representations on health behaviors in the
physical world, researchers investigated the potential of using a virtual pet to promote
physical activity in children [45]. The American Heart Association released a scientific
statement in 2013 regarding pet ownership and cardiovascular risk [46], noting that

Wiley STM / Combs, Sokolowski, Banks: The Digital Patient: Advancing Medical Research,
Education, and Practice Chapter 12/ Ahn / filename: ch12.doc
owing a pet, a dog in particular, significantly increases physical activity levels of the
pet’s owner, thereby reducing the risks for cardiovascular diseases and obesity.
Guided by the framework of social cognitive theory [47], the virtual pet was a dog
designed to systematically promote physical activity in children through goal setting,
vicarious experiences, and positive reinforcement. In the study, children’s physical
activity was measured with an activity monitor that was synchronized with each virtual
dog so that each child was paired with a unique pet displayed on a television screen
mounted on a kiosk. The kiosk set up allows for the virtual pet to be mobile, following
the children wherever needed rather than the children having to come to a specified
location to participate in the program.
The underlying logic was that as children engaged in physical activity in the
physical world, the virtual dog would also stay active with them in the virtual world,
reaping the health benefits. When compared with children in the control group who were
given an identical computer system with the same functionalities but without the virtual
dog, children who interacted with the virtual dog engaged in approximately 1.09 more
hours of physical activity daily. Self-report survey data revealed that interacting with the
virtual dog led children to feel confident about their abilities to set and meet physical
activity goals, which in turn, heightened their beliefs that physical activity is good for
them. The increase in physical activity belief ultimately led to an increase in physical
activity.

Wiley STM / Combs, Sokolowski, Banks: The Digital Patient: Advancing Medical Research,
Education, and Practice Chapter 12/ Ahn / filename: ch12.doc
Conclusion
In an era of digital media technology, people consume health information in ways that are
very different from the past. For young people, in particular, the Internet is one of the
most sought out sources for health information [48]. Atkin [49] argues that choosing a
channel appropriate for a specifically targeted audience will maximize the effect of health
information. With the increasing ubiquity of interactive and mobile digital technology in
our homes, it may be a timely endeavor to reexamine not just the content of health
information but also how it is being disseminated.
Agents and avatars offer a dynamic yet highly controllable means to deliver
health information in a novel, involving way. Offering a wide range of strategic tools
that take advantage of novel media characteristics, such as the virtual acceleration of time
and interactivity, agents and avatars yield powerful impacts in the virtual world that
transfer to the physical world to change health behaviors. Observing and interacting with
these virtual representations allows individuals to feel as if they are genuinely present in
the situation [31, 34, 35], heightens their confidence about achieving health goals [45],
and encourages them to think of the health risk as a personally relevant, important, and
imminent event [33, 34]. Ultimately, these underlying mechanisms drive desired health
behaviors that persist longer over time than the same health information delivered
through more traditional channels, such as statistical information, print, or pictures [31,
33-35, 38, 45]. These efforts may even be combined with gaming mechanisms to replace
overall sedentary time with physical activity while playing video games [40-42].
As agents and avatars offer different strengths and weaknesses, researchers and
practitioners should administer an extensive analysis of cost-effectiveness to select the

Wiley STM / Combs, Sokolowski, Banks: The Digital Patient: Advancing Medical Research,
Education, and Practice Chapter 12/ Ahn / filename: ch12.doc
more appropriate form of virtual representation in the given context. An alternative
option would be to consider an agent-avatar hybrid, which capitalizes the strength of
programmable features and algorithms of agents while still being guided by a human
controller [50, 51]. This crossover design would offer health interventions that have
greater impact on human behavior but remain cheaper to operate and manage.
There is still much work to be done to harness the dynamic flexibility that agents
and avatars offer to implement systematic primary and secondary prevention and
intervention programs. However, the state-of-the-art research introduced in this chapter
confirms the potential of virtual representations to serve as a vehicle of health behavior
change. Health issues often involve an intricate and complex web of individual and
environmental factors, and avatars and agents may not be a panacea for all these issues.
Yet, with the rapid advancement of digital technology transforming our traditional norms
and patterns of communication, these virtual representations, whether they are human or
animal form, hold much potential in the realm of health interventions that has yet to be
discovered.
Key Terms
agents
avatars
presence
virtual representations
interactivity
acceleration of time
health behavior
physical activity
exergames
References
[1] Schnipper M. The state of virtual reality. The Verge. 2014.
http://www.theverge.com/a/virtual-reality/intro.

Wiley STM / Combs, Sokolowski, Banks: The Digital Patient: Advancing Medical Research,
Education, and Practice Chapter 12/ Ahn / filename: ch12.doc
[2] Ahn SJ, Fox J, Bailenson J. Avatars. In: Bainbridge WS, ed. Leadership in Science
and Technology: A Reference Handbook. SAGE Publications; 2011: 695-702.
[3] Bailenson J, Blascovich J. Avatars. In: Bainbridge WS, ed. Encyclopedia of Human-
Computer Interaction. Great Barrington, MA: Berkshire; 2044: 64-68
[4] Blascovich, J. Social influence within immersive virtual environments. In: Schroeder
R, ed. The Social Life of Avatars: Presence and Interaction in Shared Virtual
Environments. London: Springer-Verlag; 2002: 127-145.
[5] Lim S, Reeves B. Computer agents versus avatars: Responses to interactive game
characters controlled by a computer or other player. International Journal of Human-
Computer Studies. 2010; 68: 57-68.
[6] Okita SY, Bailenson JN, Schwartz DL. The mere belief in social interaction improves
learning. In: Barab S, Hay K, Hickey D, eds. Proceedings of the 8th International
Conference for the Learning Sciences. Mahwah, NJ: Erlbaum. 2008: 132-139
[7] Fox J, Ahn SJ, Janssen J, Yeykelis L, Segovia K, Bailenson JN. Avatars versus
agents: A meta-analysis quantifying the effect of agency. Human Computer Interaction.
In press
[8] Sigelman L, Sigleman CK, Fowler C. A bird of a different feather? An experimental
investigation of physical attractiveness and the electability of female candidates. Social
Psychology Quarterly. 1987; 50, 32-43.
[9] Rosenberg SW, McCafferty P. The image and the vote: Manipulating voter’s
preferences. Public Opinion Quarterly. 1987; 51: 31-47.
[10] Todorov A, Mandisodza AN, Goren A, Hall CC. Inferences of competence from
faces predict election outcomes. Science. 2005; 308: 1623-1626.

Wiley STM / Combs, Sokolowski, Banks: The Digital Patient: Advancing Medical Research,
Education, and Practice Chapter 12/ Ahn / filename: ch12.doc
[11] Baumeister R. The self. In: Gilbert DT, Fiske ST, Lindzey G, eds. The Handbook of
Social Psychology. Boston: McGraw-Hill; 1998: 680-740.
[12] Bailenson JN, Iyengar S, Yee N, Collins NA. Facial similarity between voters and
candidates causes influence. Public Opinion Quarterly. 2008; 72: 935-961.
[13] Zajonc RB. Mere exposure: A gateway to the subliminal. Current Directions in
Psychological Science. 2001; 10: 224-228.
[14] Tanner RJ, Maeng A. A tiger and a president: Imperceptible celebrity facial cues
influence trust and preference. Journal of Consumer Research. 2012; 39: 769-783.
[15] Chartrand TL, Bargh JA. The chameleon effect: The perception-behavior link and
social interaction. Journal of Personality and Social Psychology. 1999; 76: 893-910.
[16] Making Choices in Health: WHO Guide to Cost-Effectiveness Analysis. In. Tan-
Torres ET, Baltussen R, Adam T, Hutubessy R, Acharya A, Evans DB, Murray CJL.
World Health Organization, Geneva; 2003.
[17] Ahn SJ, Fox J, Dale KR, Avant JA. Framing virtual experiences: Effects on
environmental efficacy and behavior over time. Communication Research. In press.
[18] Heeter C. Interactivity in the context of designed experiences. Journal of Interactive
Advertising. 2000; 1.
[19] Rafaeli S. Interactivity: From new media to communication. In: Hawkins RP,
Wiemann JM, Pingree S, eds. Sage Annual Review of Communication Research:
Advancing Communication Science: Merging Mass and Interpersonal Processes, 16.
Beverly Hills: Sage; 1988: 110-134.
[20] Fortin DR, Dholakia RR. Interactivity and vividness effects on social
presence and involvement in a web-based advertisement. Journal of Business

Wiley STM / Combs, Sokolowski, Banks: The Digital Patient: Advancing Medical Research,
Education, and Practice Chapter 12/ Ahn / filename: ch12.doc
Research. 2005; 58: 387-396.
[21] Sundar SS, Kim J. Interactivity and persuasion: Influencing attitudes with
information and involvement. Journal of Interactive Advertising. 2005; 5(2): 6-29.
[22] Yee N, Bailenson JN. The proteus effect: The effect of transformed self-
representation on behavior. Human Communication Research. 2007; 33: 271-290.
[23] Yee N, Bailenson JN, Ducheneaut N. The proteus effect: Implications of
transformed digital self-representation on online and offline behavior. Communication
Research: 2009; 36: 285-312.
[24] Bem DJ. Self-perception theory. Advances in Experimental Social Psychology.
1972; 6: 1-62.
[25] Young TJ, French LA. Height and perceived competence of U.S. presidents.
Perceptual and Motor Skills. 1996; 82: 1002.
[26] Kahneman D, Lovallo D. Timid choices and bold forecasts: A cognitive perspective
on risk taking. Management Science: 1993; 39: 17-31.
[27] Trope Y, Liberman N, Wakslak C. Construal levels and psychological distance:
Effects on representation, prediction, evaluation, and behavior. Journal of Consumer
Psychology. 2007; 17: 83-95.
[28] Mitchell TR, Thompson L, Peterson E, Cronk R. Temporal adjustments in the
evaluation of events: The “rosy view.” Journal of Experimental Social Psychology. 1997:
33; 421-448.
[29] Murdock, BB Jr. The serial position effect in free recall. Journal of Experimental
Psychology. 1962; 64: 482-488.

Wiley STM / Combs, Sokolowski, Banks: The Digital Patient: Advancing Medical Research,
Education, and Practice Chapter 12/ Ahn / filename: ch12.doc
[30] Hershfield HE, Goldstein DG, Sharpe WF, Fox J, Yeykelis L, Carstensen LL,
Bailenson JN. Increasing saving behavior through age-progressed renderings of the future
self. Journal of Marketing Research. 2011; 48: S23-S37.
[31] Fox J, Bailenson JN, Binney J. Virtual experiences, physical behaviors: The effect of
presence on imitation of an eating avatar. Presence: Teleoperators and Virtual
Environments. 2009; 18: 294-303.
[32] Lombard M, Ditton T. At the heart of it all: The concept of presence. Journal of
Computer-Mediated Communication. 1997; 3.
[33] Ahn SJ. Incorporating immersive virtual environments in health promotion
campaigns: A construal level theory approach. Health Communication. In press.
[34] Ahn SJ, Fox J, Hahm JM. Using virtual doppelgängers to increase personal
relevance of health risk communication. Lecture Notes in Computer Science. 2014; 8637:
1-12.
[35] Ahn SJ. Using Virtual exemplars in health promotion advertisements to promote
health behavior change: Reducing soft drink consumption. Unpublished manuscript
[36] Trope Y, Liberman N. Construal-level theory of psychological distance.
Psychological Review. 2010; 117: 440-463.
[37] Fox J, Bailenson JN. The use of doppelgängers to promote health behavior change.
CyberTherapy & Rehabilitation. 2010; 3: 16-17.
[38] Fox J, Bailenson JN. Virtual Self-Modeling: The Effects of Vicarious Reinforcement
and Identification on Exercise Behaviors. Media Psychology. 2009; 12: 1-25.
[39] Blascovich J, Bailenson JN. Infinite Reality – Avatars, Eternal Life, New Worlds,
and the Dawn of the Virtual Revolution. New York: William Morrow; 2011.

Wiley STM / Combs, Sokolowski, Banks: The Digital Patient: Advancing Medical Research,
Education, and Practice Chapter 12/ Ahn / filename: ch12.doc
[40] Biddiss E, Irwin J. Active video games to promote physical activity in children and
youth: A systematic review. Archives of Pediatrics and Adolscent Medicine. 2010; 164:
664-672.
[41] Peng W, Crouse J. Playing in parallel: The effects of multiplayer modes in active
video game on motivation and physical exertion. Cyberpsychology, Behavior, and Social
Networking. 2013; 16: 423-427.
[42] Staino AE, Abraham AA, Calvert SL. Adolescent exergame play for weight loss and
psychosocial improvement: A controlled physical activity intervention. Obesity. 2013;
21: 598-601.
[43] Hu FB, Li TY, Colditz GA, Willett WC, Manson JE. Television watching and other
sedentary behaviors in relation to risk of obesity and type 2 diabetes mellitus in women.
The Journal of the American Medical Association. 2003; 289: 1785-1791.
[44] Peng W, Crouse J, Lin J-H. Using active video games for physical activity
promotion: A systematic review of the current state of research. Health Education &
Behavior. 2013; 40: 171-192.
[45] Ahn SJ, Johnsen K, Robertson T, Moore J, Brown S, Marable A, Basu A. Using
virtual pets to promote physical activity in children: An application of the youth physical
activity promotion model. Journal of Health Communication. In press.
[46] Levine GN, Allen K, Braun L, Christian HE, Friedmann E, Taubert KA et al. Pet
ownership and cardiovascular risk: A scientific statement from the American Heart
Association. Circulation. 2013; 127: 2353-2363.
[47] Bandura A. Social Foundations of Thought and Action: A Social Cognitive Theory.
Englewood Cliffs, NJ: Prentice-Hall. 1986.

Wiley STM / Combs, Sokolowski, Banks: The Digital Patient: Advancing Medical Research,
Education, and Practice Chapter 12/ Ahn / filename: ch12.doc
[48] Gray NJ, Klein JD, Noyce PR, Sesselberg TS, Cantrill JA. Health Information-
Seeking Behaviour in Adolescence: The Place of the Internet. Social Science &
Medicine. 2005; 60: 1467–1478.
[49] Atkin C. Impact of Public Service Advertising: Research Evidence and Effective
Strategies. Menlo Park, CA. 2001.
[50] Benford S, Bowers J, Fahlén LE, Greenhalgh C, Snowdon D. Embodiments, avatars,
clones, and agents for multi-user, multi-sensory virtual worlds. Multimedia Systems.
1997; 5: 93-104.
[51] Gerhard M, Moore DJ, Hobbs DJ. Continuous presence in collaborative virtual
environments: Towards a hybrid avatar-agent model for user representation. In Intelligent
Virtual Agents. Berlin, Germany: Springer; 2001: 137-155.