Multitasking, Presence & Self-Presence on the Wii

Abstract

Previous research has shown that presence, self-presence, and multitasking all affect task performance, but these mechanisms have been tested in separate experiments. This study is one of the first to concurrently examine how the feeling of presence and self-presence while playing Nintendo Wii affects the ability to perform on various multitasking tasks. In a 2 x 2 between-subjects design, participants played Wii Tennis either with a Mii they created to resemble themselves or with a generic Mii. To make participants self-aware of their virtual or real selves, they were told that they were videotaped either on the screen or both on-screen and in the physical room. These two manipulations, Mii similarity and self-awareness, were found to significantly affect the participants' feelings of self-presence or presence, respectively, in the expected directions. These manipulations also significantly affected participants' behavioral performance on various multitasking tasks, including memory encoding and retrieval, and problem solving. These data demonstrate the importance of presence management and strategic avatar choice in multitasking scenarios.

Full text

Multitasking, Presence & Self-Presence on the Wii
Rabindra Ratan, University of Southern California
Michael Santa Cruz, University of Southern California
Peter Vorderer, VU University Amsterdam
Abstract
Previous research has shown that presence, self-presence,
and multitasking all affect task performance, but these
mechanisms have been tested in separate experiments. This
study is one of the first to concurrently examine how the feeling
of presence and self-presence while playing Nintendo Wii
affects the ability to perform on various multitasking tasks. In
a 2 x 2 between-subjects design, participants played Wii Tennis
either with a Mii they created to resemble themselves or with a
generic Mii. To make participants self-aware of their virtual
or real selves, they were told that they were videotaped either
on the screen or both on-screen and in the physical room.
These two manipulations, Mii similarity and self-awareness,
were found to significantly affect the participants’ feelings of
self-presence or presence, respectively, in the expected
directions. These manipulations also significantly affected
participants’ behavioral performance on various multitasking
tasks, including memory encoding and retrieval, and problem
solving. These data demonstrate the importance of presence
management and strategic avatar choice in multitasking
scenarios.
Keywords--- Presence, Self-Presence, Multitasking, Avatar,
Self-Awareness, Nintendo Wii
1. Introduction
Consider, briefly, that you are sitting on a couch, playing a
video game in which your character is struggling to vanquish a
seemingly unassailable enemy, when suddenly your phone, in
the real world, rings. It’s the pizza delivery person, lost and
asking for directions. Instead of pausing the game, you
continue your battle, simultaneously guiding your sword
toward your enemy and the pizza toward your home. Left
swing for the armor, “Right turn on Main Street.” But as the
skirmish heats up, does your ability to direct the delivery
person waiver? As your character sustains damage, sending a
twang of empathy through your real-world heart, do you
temporarily forget about the rumblings of your real-world
stomach?
As interactive media technology becomes more pervasive
in the world around us, the need to balance our virtual and real-
world endeavors becomes more apparent. The hypothetical
situation above illustrates that finding such balance is
dependent on various factors, including level of engagement
with the virtual environment, also known as presence, and level
of connection to one’s virtual representation, also know as self-
presence. The current work examines these two factors,
presence and self-presence, in an attempt to contribute to an
understanding of how concurrent virtual and real-world
endeavors are negotiated. Such an understanding will
hopefully be applicable to the design and implementation of
virtual and real environments in ways that improve efficiency,
safety and enjoyment within such environments.
2. Presence and Self-Presence as Results of
Avatar Similarity and Self-Awareness
While various scholars have debated the definition and
value of the concept of presence [1, 2, 3, 4, 5, 6, 7], the present
study does not intend to contribute to such debates but instead
will provide empirical research based on a definition of the
concept that has emerged from such debates relatively
unscathed. Namely, presence is treated here as the perception
(or misperception) that a virtual experience is actually a real
experience. Much presence-oriented empirical research utilizes
high fidelity displays to create virtual environments that can be
described as “virtual reality”, but presence is not constrained by
technology, and individuals can feel high levels of presence
even while reading a book [5]. Hence, the present study
examines the concept of presence using commonly available
technologies that are usually not described as virtual reality, a
television and a Nintendo Wii video game console.
The present study is particularly focused on a subcategory
of presence known as self-presence, treated here as the extent
to which a user feels that her avatar, or virtual representation of
self inside a virtual world, is an extension of herself [1].
Previous research has shown that variance in representations of
virtual selves results in differences in self-presence, such that
virtual selves that are more similar to real selves induce greater
feelings of self-presence than those that are less similar to real
selves [8, 9]. Consistent with such research, in the present
study the level of similarity between individuals and their
avatars was varied as a means of manipulating the amount of
self-presence that these individuals experience. Level of
similarity was controlled by either having participants use an
avatar they created in their own likeness or a generic looking
avatar that was assigned to them. Participants in the former
category were expected to experience more self-presence than
participants in the latter. Further, because self-presence is
intrinsically tied to the overarching concept of presence,
participants in the former category were expected to experience
more presence than participants in the latter.
PRESENCE 2007
167/388

Hypothesis 1a: Participants who use an avatar they have
created in their own likeness will report feeling more self-
presence than participants who use a pre-made avatar that is
assigned to them.
Hypothesis 1b: Participants who use an avatar they have
created in their own likeness will report feeling more presence
than participants who use a pre-made avatar that is assigned to
them.
Drawing from Hull and Levy’s [10] assertion that self-
awareness is a function of self-relevant information in an
individual’s social environment, differences in presence and
self-presence are expected to result from differences in self-
relevant information within either the virtual or real
environment. Hull and Levy manipulated self-relevant
information by either including or not including a mirror in a
room while participants completed a written task and by telling
participants that their answers either would or would not be
scored with feedback provided. The present study used a
similar mechanism. By indicating that the participant was
videotaped on just the screen or on both the screen and in the
physical room, we sought to induce participants to feel more
self-aware within either the virtual or real environment,
respectively. Because this mechanism relates to both the self
and the environment, it was expected to affect presence as well
as self-presence separately. Levels of presence and self-
presence within the virtual environment were expected to vary
depending on whether an individual was induced to feel more
self-aware of the virtual or real self, with participants in the
former category expected to report higher levels than
participants in the latter.
Hypothesis 2a: Participants who are induced to feel self-
aware of the virtual self (i.e., videotaped on the screen only)
will report feeling more presence than participants who are
induced to feel self-aware of the real self (i.e., videotaped on
the screen and in the physical room).
Hypothesis 2b: Participants who are induced to feel self-
aware of the virtual self (i.e., videotaped on the screen only)
will report feeling more self-presence than participants who are
induced to feel self-aware of the real self (i.e., videotaped on
the screen and in the physical room).
3. Multitasking Performance as a Measure of
Presence
The amount of presence or self-presence that an individual
feels affects how such individuals perform on various tasks.
For example, the level of similarity between avatar and
participant in a virtual environment significantly affects self-
presence and plays an important role in where participants
move in a virtual environment, how close they stand to other
avatars, and their willingness to commit embarrassing acts [9].
In the present study, level of presence and self-presence were
expected to affect not only performance within the virtual
environment but performance on multitasking tasks that occur
in the real environment as well.
The use of such a measure is notable because of increasing
evidence that media use is an activity that often occurs
concurrently with other activities [11, 12, 13] and because of
the dearth of media effects research that examines such
multitasking. While multitasking with television and music has
been shown to affect performance on homework-type and other
cognitive activities [14, 15, 16, 17, 18, 19, 20], no empirical
research seems to have been conducted on the effects of
multitasking while using interactive media. According to a
recent investigation of young people, multitasking while using
video games is the most common form of multitasking while
using interactive media (when computer use is segmented into
activities such as homework, IM, email, websites, etc) [12].
The present study delves into this largely unexplored area of
multitasking while using video games with the intent of
contributing theory and methods that may be applicable to
multitasking with other interactive media.
The present study is one of the first to concurrently utilize
the constructs of presence and multitasking as a means of
understanding both concepts. Presence and multitasking are
naturally complementary areas of study because levels of
presence are dependent on cognitive stimuli [5] and the chief
attention-oriented explanations of the effects of multitasking
largely focus on cognitive resources [14, 15, 21, 22, 23]. There
are two prevalent, somewhat conflicting areas of attention-
oriented explanations for changes in quality of performance
while multitasking, and both of these explanations are
applicable to the construct of presence. Limited-capacity
theories posit that performance suffers when people attempt to
utilize more mental resources than are available within their
overall pool of such resources [19]. Applied to presence, self-
presence and multitasking, limited capacity theories would
imply that individuals have a finite supply of resources that can
be allocated to feeling presence or self-presence, so when more
of such resources are required to experience presence or self-
presence in multiple (real or virtual) environments than are
available, the feeling of presence or self-presence suffers in one
or all of the environments. Structural interference theories
posit that performance suffers when similar cognitive tasks
compete for similar mental resources, not when resource
capacity limits are reached [16, 22, 23]. Applied to presence,
self-presence and multitasking, structural interference theories
would imply that there are numerous types or causal
mechanisms of presence and self-presence and that similar
types or mechanisms of presence and self-presence cannot
coexist within multiple (real or virtual) environments.
Both the limited-capacity and structural interference
theories are addressed in the present study. Although increases
in presence or self-presence may lead to higher levels of
cognitive performance [1], such increases are not expected to
augment mental capacity to an extent that would imply that
higher presence or self-presence is associated with improved
multitasking performance. Instead, presence and self-presence
should be negatively related with multitasking performance.
One possible explanation for such a relationship, consistent
with limited capacity theories, is that increased feelings of
PRESENCE 2007
168/388

presence and self-presence within the game environment will
require an increasing proportion of total mental resources,
reducing the amount of resources available for multitasking
tasks, thereby impairing performance on these tasks. Another
explanation, consistent with structural interference, is that
performance on the multitasking tasks suffers when these tasks
require similar resources to those that are being utilized to
facilitate the feeling of presence or self-presence in the game.
In order to account for both of these explanations, various
types of multitasking tasks were utilized in the present study,
including memory encoding and retrieval of various types of
information and brainteaser-type problem solving. Each of
these tasks requires different amounts and types of cognitive
load, but there were no specific expectations of whether
limited-capacity or structural interference theories would
explain the effects. Further, because the research on presence
and self-presence is limited, no distinction is made between
these constructs regarding the expected strength or type of
effects. In other words, performance on any or all of the
multitasking tasks was expected to suffer when there were high
feelings of either presence or self-presence.
Hypothesis 3a: Participants who report feeling the most
presence or self-presence will perform the worst on some or all
of the multitasking tasks.
Because the measurement tools of presence and self-
presence had potentially limited predictive power, the
manipulations themselves were also used to test the effects on
multitasking performance. Again, no distinction is made
between these manipulations regarding the expected strength or
type of effects.
Hypothesis 3b: Participants who use an avatar they have
created in their own likeness or who are induced to feel self-
aware of the virtual self (i.e., videotaped on the screen only)
will perform worse on some or all of the multitasking tasks
than participants who use a pre-made avatar that is assigned to
them or who are induced to feel self-aware of the real self (i.e.,
videotaped on the screen and in the physical room).
To briefly recap, the present study examines the causal
mechanisms and task-related effects of presence and self-
presence with the aim of contributing to the understanding of
how virtual
[M1]
should be designed and implemented. The two
expected causal mechanisms of presence and self-presence are
level of similarity between individuals and their avatars, and
self-awareness of the real and virtual self. The manipulated
feelings of presence and self-presence are expected to affect
multitasking performance. Using the design described below,
the present study tested the hypotheses related to these causal
mechanisms and effects.
4. Method
4.1 Participants
32 females (8 in each condition) and 32 males (8 in each
condition) participated in the study at a large, western
American university, and spanned the ages of 18-30.
Participants who were in the introductory communication class
received course extra credit while other students received no
compensation. Participants were randomly assigned to one of
the four experimental conditions resulting from crossing the
avatar similarity and self-awareness variables. Differences
regarding compensation were balanced evenly across
conditions.
4.2 Apparatus
The main device used in this study was the Nintendo Wii
console, the latest generation in home entertainment systems
created my Nintendo, and the game Wii Tennis. Due to the
unique nature of its control interface, essentially a wireless
controller with 5 degrees of freedom (pitch, yaw, roll, x-axis, z-
axis), playing the Wii requires much more physical movement
than playing with most other video game consoles. Because
much of this physical movement is directly correlated with the
avatar’s movement on-screen, the Wii has the potential to
induce significant amounts of presence within players. For
example, when playing Wii tennis, the game used for this
study, the player tosses up the ball to serve by moving her arm
upwards and then swings her arm forward as if the controller
were a racquet to hit the ball. These movements are quite
similar to the movements required in real tennis and so the
opportunity for a player to perceive this virtual experience as
real is significant.
The Wii with Wii Tennis were also chosen as the primary
device for this study because of the feature that allows players
to design their own avatar, known as a “Mii”, and then use this
avatar in the game. Mii’s can be customized in a variety of
ways, including height, body mass, face shape, skin/eye/lip
color, hairstyle and facial features. Given this vast range of
options, Mii’s can be created to look quite similar, though
cartoonish, to their real-life counterparts.
Standard equipment also used in this study included a
computer, 46-inch television, video camera, and stopwatch.
The computer was used to administer the questionnaires to the
participants. The output from the Wii was displayed on the
television. The video camera, used to induce self-awareness
but never actually turned on, was stationed in the back of the
room. The stopwatch was used to administer the multitasking
questions. The room in which the study was conducted and
equipment used is depicted in Figure 1, including the Nintendo
Wii (1), Wii controller (2), television with Mii customization
screen displayed (3), and video camera (4).
PRESENCE 2007
169/388

Figure 1. Room and Equipment
4.3 Design
The design included two between-subjects variables:
avatar similarity (similar or dissimilar) and self-awareness
(virtual or real self). Gender was distributed evenly and
participants were randomly assigned to one of the four
conditions. The 2 x 2 matrix depicting these assignments can
be found in Table 1.
Self-Aware:
Virtual Self
Self-
Aware:
Real Self
Avatar and
Participant
Similar
16
Participants
16
Participants
Avatar and
Participant
Dissimilar
16
Participants
16
Participants
Table 1. Study Conditions
In order to control for possible effects from the act of
creating a Mii, all participants created a Mii in their own
likeness. Participants in the similar-avatar condition played the
game with the Mii they had created. Before beginning the
games, these participants were told, “You will be playing with
your Mii.” Participants in the dissimilar avatar condition
played the game with a generic-looking Mii that was assigned
to them. In this latter condition, avatar gender always matched
the participants’ gender. Before beginning the games, these
participants were told, “You will be playing with,” for females,
“Amanda” or “Stacie,” and for males, “Larry” or “Bob”.
Participants in this condition were randomly assigned to one of
two Miis. Images of these generic-looking Mii’s can be found
in Figure 2.
Figure 2. Generic-looking Miis
The researcher administered the self-awareness
manipulation verbally by reciting a scripted statement that also
reinforced the avatar similarity condition. Participants in the
virtual self-aware condition were told that while they played
the game, the video camera would record their performance on
the screen only. Participants in the real self-aware condition
were told that while they played the game, the video camera
would record their performance on the screen as well as in the
physical room. Reflecting the avatar similarity condition, for
participants in the virtual self-aware and avatar similar
condition, the researcher pointed to the video camera and said,
“We have set up that camera right there,” and then pointed at
the screen and said, “to record you in the game.” For
participants in the real self-aware and avatar similar condition,
the research pointed to the video camera and said, “We have set
up that camera right there,” then pointed to the participant and
said, “to record both you,” and then pointed to the screen and
said, “and your Mii while you play the game.” For participants
in the virtual self-aware and avatar dissimilar condition, the
researcher pointed to the video camera and said, “We have set
up that camera right there,” and then pointed at the screen and
said, “to record Amanda in the game.” Note that depending on
participant gender and random selection the researcher could
have also said “Stacie”, “Bob”, or “Larry”. For participants in
the real self-aware and avatar dissimilar condition, the research
pointed to the video camera and said, “We have set up that
camera right there,” then pointed to the participant and said, “to
record both you,” and then pointed to the screen and said, “and
Amanda while you play the game.” Again, the specific name
depends on participant’s gender and random selection.
4.4 Procedure
After participants arrived at the laboratory and reviewed
their informed consent information, they filled out a
questionnaire that asked basic demographic information and
then learned how to properly wear and manipulate the Wii
PRESENCE 2007
170/388

controls. The researcher then instructed the participant to
design a Mii and requested, “Please make this Mii look as
much like yourself as possible.” After customizing their Mii
for no more than 5 minutes, participants learned how to play
Wii tennis. The researcher read instructions on how to serve,
return and direct the ball and then demonstrated these functions
while playing a few volleys with the computer. According to
condition, participants were then told which Mii they would be
playing with and the function of the video camera.
Next, participants were instructed to remember the
following list of facts, words, and numbers. The research read
this list slowly in the order presented below.
Facts
• A dime has 118 ridges around the edge.
• A cat has 32 muscles in each ear.
• A dragonfly has a life span of 24 hours.
• A goldfish has a memory span of three seconds.
• Cats sleep an average of 16 hours per day.
Words
• Cantankerous
• Interspersed
• Rejuvenate
• Mollify
• Trampoline
Numbers
• 690
• 695
• 130
• 214
• 340
Participants were then instructed to play a series of single-
match games against the computer and to listen to and respond
verbally to the researcher while they were playing. Starting
from the participants’ first serve, at 30-second intervals the
researcher asked the participant to remember (recognize) the
information that had been read earlier. These questions were
phrased in the format of, “Question, true or false, a cat has 42
muscles in each ear?” and “Question, yes or no, do you
remember hearing the word ‘mollify’?” The stopwatch was
paused in between tennis matches as well as while the
participants responded to questions during the matches. After
being asked to remember one fact, one word and one number,
participants were told an additional fact in the format of, “New
fact, there are about 1500 active volcanoes world-wide.”
Following the next 30-second interval, participants were asked
to pause and read a brainteaser-type word problem.
Participants were given different time limits for different word
problems, ranging from 30 to 90 seconds, depending on
problem difficulty. This entire process – fact question, word
question, number question, new fact presentation, and word
problems – was repeated five times while the participant played
tennis. All of the new facts and word problems, with time
limits, are listed below:
New Facts
• There are about 1500 active volcanoes worldwide.
• The winter of 1932 was so cold that Niagara Falls
froze completely solid.
• A snail can sleep for three years.
• There are 293 ways to make change for a dollar.
• February 1865 is the only month in recorded history
not to have a full moon
Word Problems
• How can you throw a ball as hard as you can and have
it come back to you, even if it doesn't hit anything,
there is nothing attached to it, and no one else catches
or throws it? (30 secs to respond)
• The day before yesterday I was 25 and next year I will
be 28. (This is true only one day in a year.) When was
I born? (60 secs to respond)
• How many flowers do I have if all of them are roses
except two, all of them are tulips except two, and all
of them are daisies except two? (30 secs to respond)
• If you had ONLY a 5-litre and a 3-litre bowl and
unlimited access to water. How would you measure
exactly 4 litres (to be contained in one or both of the
bowls)? (90 secs to respond)
• I am looking at somebody's photo. Who is it I am
looking at, if I don't have any brothers or sisters and
the father of that man on the photo is the son of my
father? (45 secs to respond)
After this process had been iterated 5 times, the
participants were told to stop playing and were asked some
additional questions about specific characteristics of the game
and about the new facts they were told while playing. The
former included the three questions, “Did your Mii have legs
while you were playing?”, “Did your Mii have arms while you
were playing?”, and “What did the sign in the back of the
tennis court say?” The latter were quite similar to the fact
questions asked during the game and so they are not listed
below. The participants then completed a post-questionnaire
that measured their feelings presence and self-presence, their
video gaming and multitasking habits, and their qualitative
attitudes about their Mii and the self-awareness condition. The
items from this questionnaire that were used in the analysis are
described in the following section. Finally, before leaving the
room, the participants were debriefed about the purpose of the
study and the minor deception utilized.
5. Measures
Reliability analysis was used to create a subjective
measure of presence that included eight questions to which the
participants responded on 5-pt likert scales. In these questions
PRESENCE 2007
171/388

the participant indicated the extent of agreement with the
statements, “The game looked realistic,” “I felt like I was in the
tennis game,” “The game felt realistic,” “I felt engaged in the
game,” “I felt bored while playing the game” (reverse coded),
“I gripped the controls tightly while playing,” “I used grand
arm movements while playing,” and “I swung the controller
like it was a real racket.” Cronbach’s alpha for this composite
measure was .802.
Reliability analysis was also used to create a subjective
measure of self-presence that included five questions to which
the participants responded on 5-pt likert scales. In these
questions the participant indicated the extent of agreement with
the statements, “I liked the Mii I played with,” “I felt connected
to the Mii I played with,” “The Mii I played with was an
accurate representation of me,” “The Mii I played with looked
like a type of person I would want to be,” and “The Mii I
played with looked like me.” Cronbach’s alpha for this
composite measure was .791.
As described above, multitasking performance included
measures of the number of facts, words and numbers the
participants remembered correctly and the number of word
problems the participants answered correctly. There was not
enough agreement between any of these items to create a
composite measure of multitasking performance.
6. Results
An analysis of variance was used to test the effects of Mii
similarity on self-presence and presence, with the between-
subjects Mii similarity manipulation as the independent
variable and the subjective composite measures of self-
presence and presence as the dependent variables. Supporting
Hypothesis 1a, there was a significant main effect for Mii
similarity and self-presence, F (1, 62) = 13.56, p < 0.01, η
p
2
=
0.19, with participants who used a similar Mii reporting more
self-presence (mean = 15.77, SD = 3.66) than participants who
used a dissimilar Mii (mean = 12.22, SD = 3.77). There was no
main effect for Mii similarity and presence, F (1, 62) = 0.00, p
= 0.96, η
p
2
= 0.00, so Hypothesis 1b was not supported.
An analysis of variance was used to test the effects of self-
awareness on presence and self-presence, with the between-
subjects self-awareness manipulation as the independent
variable and the subjective composite measures of presence and
self-presence as the dependent variables. Supporting
Hypothesis 2a, there was a significant main effect for self-
awareness and presence, F (1, 62) = 4.28, p < .05, η
p
2
= 0.07,
with participants induced to feel self-aware in the virtual
environment reporting more presence (mean = 29.78, SD =
4.51) than participants induced to feel self-aware in the real
environment (mean = 26.83, SD = 6.70). There was no main
effect for self-awareness and self-presence, F (1, 64) = 0.47, p
= .50, η
2
= 0.01, so Hypothesis 2b was not supported.
Interaction effects between the independent variables and
the presence and self-presence dependent variables were also
tested. There was no interaction effect for Mii similarity by
self-awareness with presence as the dependent variable, F (1,
62) = 1.92, p = 0.17, η
p
2
= 0.03, nor for Mii similarity by self-
awareness with self-presence as the dependent variable, F (1,
62) = 0.03, p = 0.87, η
p
2
= 0.00.
None of the simple correlations between the composite
measures of presence and self-presence were significantly
related to multitasking performance. Hence, Hypothesis 3a
was not supported at all. The insignificant correlation results
can be found in Table 2 below.
In-Game Fact Memory
Presence
Pearson's Correlation = -.12, p = .36
Self-Presence
Pearson's Correlation = -.14, p = .30
In-Game Word Memory
Presence
Pearson's Correlation = .10, p = .45
Self-Presence
Pearson's Correlation = -.25, p = .06
In-Game Number Memory
Presence
Pearson's Correlation = .05, p = .72
Self-Presence
Pearson's Correlation = -.17, p = .21
Word Problem Performance
Presence
Pearson's Correlation = -.18, p = .17
Self-Presence
Pearson's Correlation = -.05, p = .70
Memory of Game Characteristics
Presence
Pearson's Correlation = -.03, p = .84
Self-Presence
Pearson's Correlation = .20, p = .17
Post-Game Fact Memory
Presence
Pearson's Correlation = -.06, p = .63
Self-Presence
Pearson's Correlation = -.15, p = .26
Table 2. Insignificant Multitasking Correlation Results
An analysis of variance was used to test the effects of the
Mii similarity and self-awareness manipulations on the
multitasking performance dependent variables. Regarding
performance on the word problems, partially supporting
Hypothesis 3b, there was a significant main effect
1
for self-
awareness, F (1, 62) = 4.1, p < .05, η
p
2
= 0.07, with participants
induced to feel self-aware in the real environment answering
more word problems correctly (mean = 1.69, SD = 1.23) than
participants induced to feel self-aware in the virtual
1
Gender was used as a covariate there was a weak main effect between
gender and word problem performance, F (1, 62) = 3.9, p = .05, η
2
= 0.06, with
males answering more word problems correctly (mean = 1.69, SD = 1.26) than
females (mean = 1.13, SD = 1.00).
PRESENCE 2007
172/388

environment (mean = 1.13, SD = 1.04). There was no main
effect for Mii similarity, F (1, 62) = 0.20, p = .65, η
p
2
= 0.00,
and no interaction effect for Mii similarity by self-awareness, F
(1, 62) = 0.82, p = 0.18, η
p
2
= 0.03.
Using an analysis of variance, the effects on multitasking
performance were also tested with the two independent
variables (Mii similarity and self-awareness) and the dependent
variable that reflects the participants’ ability to remember
(encode) a series of new facts while playing the game and
recognize these new facts after playing. There were no main
effects for self-awareness, F (1, 62) = 0.45, p = .47, η
p
2
= 0.01,
or for Mii similarity, F (1, 62) = 1.41, p = .24, η
p
2
= 0.02.
Partially further supporting Hypothesis 3b, there was a
significant interaction effect for Mii similarity by self-
awareness, (1, 62) = 4.33, p < .05, η
p
2
= 0.07, and the means
and confidence intervals of performance by condition for this
effect are illustrated in Figure 3.
Figure 3. Post-Game Fact Memory
Using an analysis of variance, the effects on multitasking
performance were also tested with the two independent
variables (Mii similarity and self-awareness) and the dependent
variable that reflects the participants’ ability to remember a list
of numbers before playing the game and recognize (retrieve)
these numbers while playing. There were no main effects for
self-awareness, F (1, 62) = 0.28, p = .60, η
p
2
= 0.01 or for Mii
similarity for males, F (1, 62) = 2.53, p = .12, η
p
2
= 0.04. There
was no interaction effect for Mii similarity by self-awareness, F
(1, 62) = 0.91, p = .34, η
p
2
= 0.02. However, splitting the data
according to gender and rerunning the test did yield one
significant result with regard to number recognition, albeit
unreliable because of the low N. Partially supporting
Hypothesis 3b, for females only, there was a significant
interaction effect for Mii similarity by self-awareness, F (1, 30)
= 6.18, p = .02, η
p
2
= 0.18. The means and confidence intervals
of performance by condition for this effect are illustrated in
Figure 4.
Figure 4. Female In-Game Number Memory
None of the other multitasking effects tested with an
analysis of variance provided significant results.
7. Discussion
The findings outlined above provide some evidence for the
basic premise of this study. Namely, feelings of presence and
self-presence can be manipulated and such manipulations can
affect performance on multitasking tasks. A detailed
explanation of these findings follows.
7.1 Manipulating Presence and Self-Presence
The expectation that feelings of presence and self-presence
are dependent on whether participants are assigned to a
character that is similar or dissimilar to themselves was
partially fulfilled. Participants who used a Mii that was
dissimilar to themselves reported significantly less self-
presence than participants who used a Mii that was similar to
themselves. However, feelings of presence were unaffected by
character assignment. Hence, Hypothesis 1a was supported but
Hypothesis 1b was not. The latter finding is not surprising.
Presence is an umbrella concept to which self-presence belongs
in addition to other factors. This finding indicates that the
other factors that affect presence most likely influence presence
more significantly than self-presence. However, such a
conclusion based on a lack of significant results is not
dependable. Instead, it is fair to conclude that these results do
not contradict this potential explanation.
The former result, that avatar assignment and self-presence
are significantly related, corroborates previous research and
notably implies that level of self-presence can be manipulated
through avatar assignment. Further, in their open-ended
responses, many participants in the dissimilar Mii condition
noted a lack of connectedness to the Mii based on this
dissimilarity, for example, by stating, “Because I didn't use the
PRESENCE 2007
173/388

one I made, I didn't feel any sense of connection to the one I
played with.” However, some participants in the opposite
condition indicated that the Mii’s similarity to themselves did
not affect their performance, for example, by stating, “After I
created the Mii character I did not pay much attention to his
looks. I was simply focused on winning the matches and
answering the question correctly.” Hence, the effects of Mii
similarity on self-presence sometimes goes unnoticed by the
user, implying that choosing an avatar in order to induce an
appropriate level of self-presence, depending on environmental
constraints, should not always be left up to the user. Future
research should empirically examine whether users understand
the relationship between avatar similarity and self-presence and
how systems should assign avatars in either case.
The expectation that feelings of presence and self-presence
are dependent on whether participants are induced to feel self-
aware of the virtual self (i.e., videotaped on the screen only) or
of the real self (i.e., videotaped on the screen and in the
physical room) was partially fulfilled. Participants who were
induced to feel self-aware of the virtual self reported
significantly less presence than participants who were induced
to feel self-aware of the real self. However, feelings of self-
presence were unaffected by the self-awareness inducement.
Hence, Hypothesis 2a was supported but Hypothesis 2b was
not. The latter finding is not surprising because the inducement
of self-awareness focuses on the salient environment, not on
representation of the self. The former implies that level of
presence can be manipulated by inducing self-awareness within
real or virtual environments, but it should be noted that the
effect size is relatively small (.07).
As with the previous finding, the effects of videotaping
went unnoticed by many participants, indicated by statements
such as, “I was initially uncomfortable wondering where the
video of me playing might end up being played (in case I
[stunk]!) but forgot its presence pretty quickly.” Self-
awareness can be manipulated through many other mechanisms
besides videotaping and could be an important facet of
developing media technologies that adapt to fluid
environmental constraints. Future research should test the
effects of such additional mechanisms of inducing self-
awareness, such as by referring to facets of the real or virtual
environments.
7.2 Multitasking Performance
The expectation that multitasking performance is
dependent on the feelings or manipulations of presence and
self-presence was partially fulfilled. There were no significant
differences in multitasking performance between participants
who reported feeling the most presence or self-presence and
participants who reported feeling the least presence or self-
presence. So, Hypothesis 3a was unsupported. The
manipulations of presence and self-presence did significantly
affect performance on some of the multitasking tasks, so
Hypothesis 3b was partially supported.
The lack of a relationship between multitasking
performance and feelings of presence or self-presence could
partially be explained by flaws in these composite measures
that prevent them from fully representing the concepts they
attempt to target. Further, these measures are based on self-
report items, the reliability of which as a measurement of
presence and self-presence is questionable. An alternative
explanation may be rooted in the lack of validity of these
concepts themselves. As noted in the beginning of this paper,
the definitions and operationalizations of presence and self-
presence are still hotly debated. As the findings indicate, the
definitions of presence and self-presence utilized for this study
may not have been consistent with the multitasking component
of the inquiry. Yet, it is still notable that the manipulated
independent variables caused differences in these measures of
presence and self-presence and also caused differences in
multitasking performance. While it is possible that the latter
changes occurred due to some other mechanisms besides
presence or self-presence, such a mechanism would most likely
be closely tied to these concepts. So although the measures of
presence and self-presence were not related to multitasking
performance, because the manipulations that affected presence
and self-presence also affected multitasking behavior, the
potential applications of this study regarding the design of
virtual environments is not hindered. Future research may find
that concepts closely related to presence and self-presence
affect multitasking performance, or alternatively, the definition
of presence and self-presence may adapt in ways that allow for
these concepts to measurably affect multitasking performance.
Participants who were induced to feel self-aware of their
virtual self (i.e., videotaped on the screen only) performed
worse on the word problem task than participants who were
induced to feel self-aware of their real self (i.e., videotaped on
the screen and in the physical room). This finding is consistent
with the notion that more cognitive resources are allocated to
the environment in which individuals feel more present. The
small effect size (.07) should be noted and it should be
emphasized that the word problem task is not entirely a
multitasking task because participants paused the game while
they completed this task. Hence, it seems that when
participants are induced to feel self-aware in the physical room,
they are more willing to break their feeling of presence in the
virtual world in order to complete the task. Otherwise, they are
more anxious to return to the virtual world and hence dedicate
fewer cognitive resources to the task. In order to test this
explanation, future research should examine the differences
between performance on various multitasking tasks that occur
while the game is either paused or continuous.
Contributing further mixed support for Hypothesis 3b are
the interaction effects found for remembering a series of facts
and a series of numbers while playing. Participants who played
with a similar Mii remembered more facts when induced to feel
self-aware of their real self than when induced to feel self-
aware of their virtual self, but participants who played with a
dissimilar Mii remembered more facts when induced to feel
self-aware of their virtual self than when induced to feel self-
aware of their real self.
PRESENCE 2007
174/388

Further, regarding remembering numbers, it appears that
female participants who played with a similar Mii recognize
more numbers when induced to feel self-aware of their virtual
self than when induced to feel self-aware of their real self, but
female participants who played with a dissimilar Mii recognize
more numbers when induced to feel self-aware of their real self
than when induced to feel self-aware of their virtual self. This
latter finding is tenuous because of the low number of female
participants in each condition (eight), but is still interesting to
speculate about, especially because the effect is in the opposite
direction of the previous finding. The effect regarding number
recognition is consistent with Hypothesis 3b within the
dissimilar Mii condition and the effect regarding remembering
new facts is consistent with Hypothesis 3b within the similar
Mii condition. Namely, in these conditions, inducing self-
awareness of the real self improves multitasking performance.
Yet, is not immediately apparent why inducing self-awareness
of the real self detracts from multitasking performance in the
opposite Mii similarity conditions. Future research should
reexamine the relationship between Mii similarity and
multitasking performance to see if this curious interaction
effect persists.
The opposite direction of the findings between the fact
encoding and number recognition tasks is also curious and may
relate to the different cognitive mechanisms required by the
two tasks. The number task requires multitasking between the
game and memory retrieval of the numbers. The facts task
requires multitasking between the game and the memory
encoding of facts for retrieval after the game. It is possible that
encoding and retrieval utilize different types or methods of
acquiring cognitive resources and that these differences are
magnified by the feeling of presence or self-presence.
Although no specific hypothesis was constructed with regard to
such interaction effects, these findings do lend support to a
structural interference explanation of multitasking effects
because it seems that different types of activities with similar
levels of cognitive load affect multitasking in different ways.
This claim is further supported by the findings that only some
of the multitasking measures were related to the independent
variables. Future research should examine how various types
of cognitive mechanisms are affected by feelings of presence
and self-presence while multitasking.
7.3 Shortcomings
Although the present study provides some notable findings
regarding presence, self-presence and multitasking, there are
some important shortcomings in this study’s design. First, the
inducement of self-awareness relied on participants’ believing
that they were being videotaped. Videotaping can cause a
variety of effects in addition to self-awareness, and it is
possible that these other effects contributed to the participants’
multitasking performance and feelings of presence and self-
presence. Unfortunately, this study did not include a
manipulation check to ensure that the video camera’s focus on
the screen or on the screen and physical room induced self-
awareness to a greater extent than other psychological
mechanisms that could have affected the results. Future
research should include such manipulation checks and as
suggested earlier, should use different methods of inducing
self-awareness and reinforcing virtual or real identities.
An additional issue with the study design is that playing
with a similar Mii is confounded with playing with a Mii that
the participant built. Conversely, participants who played with
a dissimilar Mii were disallowed from playing with the Mii that
they built. This may be a problem if playing with something
one has built vs. something one has not built is a more
significant manipulation than playing with a character that is
similar vs. dissimilar to oneself. For example, the participants
who played with the generic Mii may have been disappointed
or frustrated that they could not play with the Mii that they
built. Future research should include a condition in which
some participants play with an avatar that they built to be
dissimilar from themselves and a condition in which some
participants play with an avatar that they did not build that is
similar to themselves, in addition to the same conditions from
this study.
Another flaw in this study relates to the participants’ skill
levels. Namely, not all of the participants began the study at
the same gaming or multitasking skill levels and this was not
controlled for in the data analysis. Instead, it was assumed that
such differences would be negligible because the participants
were randomly assigned to condition. However, the number of
participants may not have been large enough to dampen
potential effects. Further, the variance in gaming skill level
was likely too large to randomize out because of the lack of
availability of the Nintendo Wii at the time the study was
conducted. Therefore, it is possible that participants who were
already highly skilled before participating had an easier time
playing the game, allowing more resource allocation for the
secondary tasks. Future studies should exclude participants
with extreme gaming skill levels and should test and control for
extreme multitasking or cognitive abilities.
A final design flaw of this study is that participants who
did not play with the Mii they built were randomly assigned to
one of two generic-looking Miis, and although these Miis were
typical game characters, there may have been something
specific about their appearance or names that affected
participants’ performance. A better design would have had
participants in this condition play with the Miis that
participants in the opposite Mii similarity condition had
previously built and used. This way all of the same Miis would
have been used in both conditions, ensuring that there were no
effects from the specific way that these Miis looked.
Conclusions
The novel study described here concurrently examines the
concepts of presence, self-presence and multitasking, and
illustrates that such examinations are a potentially promising
avenue of research. The findings that presence and self-
PRESENCE 2007
175/388

presence can be manipulated by inducing self-awareness of the
virtual or real self, or by avatar assignment, respectively,
corroborates previous research and presents these concepts as
valuable indicators of technology users’ psychological states.
The finding that these same manipulations can affect
performance on particular multitasking tasks provides support
for the structural interference explanation of multitasking
effects and indicates that such manipulations might be
strategically used to inform the design of virtual environments.
Yet, while the same manipulations affect feelings of
presence and self-presence and affect multitasking
performance, no relationship was found between multitasking
performance and the feelings of presence or self-presence.
Hence, the study has not provided a link between the
psychological states in question and the participants’ behaviors.
This implies that either the measures of presence and self-
presence are flawed or that there is some additional factor that
is affecting the relationship between the manipulations and
multitasking that is closely related to presence or self-presence.
The fundamental conclusion then is that more research must be
conducted in order to develop the conceptual definitions of
presence and self-presence in ways that facilitate better
operational measures of these concepts.
Although this study should not serve as an archetype for
research on presence, self-presence, and multitasking, namely
because of the small effect sizes, shortcomings in design, and
somewhat unintuitive results, it instead should be considered a
pioneering attempt at understanding how these concepts
interrelate and what future research should be conducted.
Further, this study provides a basis for speculation about the
potential benefits of such research. If this future research
corroborates the findings that avatar assignment and self-
awareness affect feelings of presence and self-presence,
thereby impacting the quality of multitasking performance,
media technology applications could be built that incorporate
such findings in order to maximize efficiency, safety and
enjoyment.
Returning to the opening example from this paper in which
you are giving directions over the phone while playing a video
game, imagine that the phone and game could work in concert
to manage your feeling of presence within both mediated
environments by manipulating which environment feels more
salient to you or by changing some facet of how your character
is represented. Imagine more serious multitasking scenarios: a
pilot uses numerous virtual interfaces to monitor the
performance of a space shuttle, a doctor performs multiple
telesurgeries concurrently, or a teenager completes homework
assignments in a virtual world while watching over a younger
sibling. While the present study only provides a small step
toward a contribution of design principles for such
technologies, it is an early step upon which many others will
hopefully tread.
Acknowledgements
We would like to thank the students, Brittany Allen, David
Calderon, John Owen, for assistance in designing and
implementing this study, as well as Olga Chandra for assistance
in editing the paper.
References
[1] F. Biocca. The cyborg's dilemma: Progressive embodiment in
virtual environments. In Journal of Computer-Mediated
Communication, 3(2). September 1997. Retrieved May 30, 2007
from
http://jcmc.indiana.edu/vol3/issue2/biocca2.html
[2] J. Blascovich. Social influence within immersive virtual
environments. In R. Schroeder (Ed.), The social life of avatars.
(pp. 127-145). London: Springer-Verlag. 2002.
[3] C. Heeter. Being there: The subjective experience of presence. In
Presence:Teleoperators and Virtual Environments, 1, 262-271.
1992.
[4] International Society for Presence Research. In The Concept of
Presence: Explication Statement. Retrieved May 30, 2007 from
http://ispr.info/
. 2000.
[5] K.M. Lee. Presence, explicated. In Communication Theory, 14,
27-50. 2004.
[6] M. Lombard, T.B. Ditton. At the heart of it all: The concept of
presence. In Journal of Computer-Mediated Communication,
3(2). 1997.
[7] J. Loomis. Distal attribution and presence. In Presence:
Teleoperators and Virtual Environments, 1, 113-119. 1992.
[8] J.N. Bailenson, A.C. Beall, J. Blascovich, M. Weisbuch, R.
Raimmundo. Intelligent Agents Who Wear Your Face: Users'
Reactions to the Virtual Self. In Lecture Notes in Artificial
Intelligence, 2190, 86-99. 2001.
[9] J. Bailenson, J. Blascovich, R. Guadagno. Self
Representations in Immersive Virtual Environments. In press.
[10] J.G. Hull, A.S. Levy. The Organizational Functions of the Self:
An Alternative to the Duval and Wicklund Model of Self-
Awareness. In Journal of Personality and Social Psychology.
37(5), 756-768. 1979.
[11] M. Bloxham, M.E. Holmes, R.A. Papper, M.N. Popovich.
Middletown Media Studies: Observing Consumers and Their
Interactions with Media. Muncie, IN: Ball State University. 2006.
[12] U.G. Foehr. Media multitasking among American youth:
prevalence, predictors and pairings. Menlo Park, CA: Kaiser
Family Foundation. 2006.
[13] K.L. Schmitt, K.D. Woolf, D.R. Anderson. Viewing the Viewers:
Viewing Behaviors by Children and Adults During Television
Programs and Commercials. In Journal of Communication, 53(2),
265 – 281. 2003.
[14] G.B. Armstrong, G.A. Boiarsky, M.L. Mares. Background
television and reading performance. In Communication
PRESENCE 2007
176/388

Monographs, 58, 235-253. 1991.
[15] G.B. Armstrong, B.S. Greenberg. Background television as an
inhibitor of cognitive processing. In Human Communication
Research, 16(3), 355-386. 1990
[16] G.B. Armstrong, L. Chung. Background Television and Reading
Memory in Context: Assessing TV Interference and Facilitative
Context Effects on Encoding Versus Retrieval Processes. In
Communication Research. 27 (3). 327 – 352. 2000.
[17] J.W. Beentjes, C.M. Koolstra, T.H. van der Voort. Combining
background media with doing homework: Incidence of
background media use and perceived effects. In Communication
Education, 45, 59-72. 1996.
[18] A. Furnham, A. Bradley. Music while you work: The differential
distraction of background music on the cognitive test
performance of introverts and extroverts. In Applied Cognitive
Psychology, 11 (5), 445-455. 1997.
[19] M.M. Pool, C.M. Koolstra, T.H. van der Voort. The Impact of
Background Radio and Television on High School Students'
Homework Performance. In Journal of Communication; 53 (1).
74 – 87. 2003.
[20] M.M. Pool, T.H. van der Voort, J.W. Beentjes, C.M. Koolstra.
Background television as an inhibitor of performance on easy and
difficult homework assignments. In Communication Research,
27, 293-326. 2000.
[21] D. Rohrer, J.T. Wixted, D.P. Salmon, N. Butters. Retrieval from
semantic memory and its implications for Alzheimer's disease. In
Journal of Experimental Psychology: Learning, Memory, and
Cognition, 21, 1127-1139. 1995.
[22] C.D. Wickens. The structure of attentional resources. In R.
Nickerson (Ed.), Attention and performance (Vol.8, pp. 239-
257). Hillsdale, NJ: Erlbaum. 1980.
[23] C.D. Wickens. Processing resources in attention. In R.
Parasuraman & R. Davies (Eds.), Varieties of attention (pp. 63-
102). New York: Academic Press. 1984.
PRESENCE 2007
177/388