142
476

 

ITHS_

1" m-_...__“_ § h..—
a 1" 9‘.
l a 3-: Q L‘s RV
l ll . 4. . Tan.
‘ "”L"“t;~na viaie

. " I.
94 r'. :4"? _ ,1’
lati!i~:3811y

 

This is to certify that the
dissertation entitled

THE EFFECT OF TASK DEMAND ON MOOD REPAIR AND
SELECTIVE EXPOSURE TO VIDEO GAMES

presented by

Nicholas David Bowman

has been accepted towards fulfillment
of the requirements for the

PhD. degree in Communication

 

 

(8% JW‘

Major Professor’s Signature

M&\[ 3‘ acre

 

Date

MSU is an Affirmative Action/Equal Opportunity Employer

‘ .9-I‘v-t--I-I-I-o----l-v--I-o-I-

--a£‘!-_:_—

 

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

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5/08 K:lProj/Acc&Pres/ClRC/DateDue.indd

 

THE EFFECT OF TASK DEMAND ON MOOD REPAIR AND SELECTIVE
EXPOSURE TO VIDEO GAMES

By

Nicholas David Bowman

A DISSERTATION

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

DOCTOR OF PHILOSOPHY
Communication

2010

ABSTRACT

THE EFFECT OF TASK DEMAND ON MOOD REPAIR AND SELECTIVE
EXPOSURE TO VIDEO GAMES

By
Nicholas David Bowman

It is argued by entertainment scholars that the potential for video games to
intervene in noxious mood states is heightened by the fact that video game play is a more
demanding task as compared to consuming other forms of media. According to mood
management theory, heightened intervention potential should make video games
particularly well-adept at repairing noxious moods. Moreover, according to selective
exposure theory, this heightened intervention potential should make video game play
more desirable to people experiencing noxious mood states. Although many researchers
have made theoretical claims about the unique attributes of video games as compared to
more traditional media, both in general and in relation to mood management and selective
exposure, these claims have yet to be tested empirically.

To this end, three studies were conducted. Study 1 varied task demand in a video
game by experimentally manipulating the amount of control an individual had over a
mediated environment and found that (a) increased task demand heightens a video
game’s intervention potential, (b) heightened intervention potential enhances the video
game’s ability to relieve boredom and stress, (0) too much task demand has a detrimental
effect on mood repair, and (d) the ability of video games to repair negative affect is a
function of increased task demand, and not simply increased arousal. Study 2 used
behavioral measures of user engagement rather than experimental manipulations of task

demand and found that (a) increased involvement with some game controls has a positive

. influence on post-game play affect and (b) increased involvement with game controls has
a positive influence on affect for bored participants and a negative influence on affect for
stressed individuals. Study 3 focused on selective exposure stemming from expectations
of a video game’s task demand learned from game play, and found that (a) participants in
states of boredom and stress preferred moderate levels of expected task demand, and (b)
this preference was stronger for stressed participants than for bored participants.

By focusing attention on the role of task demand in these processes, this
collection of studies expands previous conceptualization of intervention potential in a
manner that aids efforts to understand the uses and effects of interactive media as related
to mood regulation. Combined, these studies advance our knowledge of mood
management and selective exposure processes related to a specific and increasingly-

popular form of interactive media: the video game.

ACKNOWLEDGEMENTS

Data collection for this series of studies would not have been possible without the
help and dedication from my team of undergraduate research assistants. Special thanks to
Cole Allen, Zach Conti, Alison Couzens, Evelyn Schramm, John Sfreddo, and Jarrett
Smith. Special thanks should also be given to Dr. Charles Atkin, chair of the Department
of Communication, Michigan State University, for providing ample laboratory space and
resources for this study, and Dr. Pamela Whitten, Dean of the College of
CoMunication, Michigan State University, for providing funding for this project.
Finally, I wish to thank the faculty of the 2009 National Communication Association
Doctoral Honors Seminar — Dr. Andrew C. Billings, Dr. Jennings Bryant, Dr. Mary Beth
Oliver — as well as the doctoral students in attendance that provided their insightful
feedback to earlier drafts of this work.

Perhaps most important, I also need to thank John for his friendship and guidance
through the “stupid years’f of my education (and instilling his love of hockey and Detroit
in me), Rene for his bottomless knowledge of research methods (and his bright orange
office, which would drive a lesser man insane), Frank for owning and providing me
access to his Wonder Emporium of Technology (and for being in the Western
Hemisphere for my defense hearing), and Ron for being the no-good, bastard Son of a
Bitch responsible for this whole journey in the first place (and one of the most patient and
helpful men I‘ have ever met); to these four men, I am forever indebted. And — of course —-
my wife, Jennifer, for her support, comments and suggestions throughout the editing

process.

iv

TABLE OF CONTENTS

 

 

 

 

 

 

List of Tables -- - - - - - -- - -- - .............. vii

List of Figures - - - - _ -viii
Introduction - -- - - 1
Mood management and selective exposure: An overview ............................... 2
Intervention potential in different media forms .............................................. S
‘ Study 1 ...... - -- -8
Hypotheses ...................................................................................................... 8
Method .............................................................................................................. 11
Participants .................................................................................................... 11
Design and Procedure .................................................................................... 13
Stimuli/Materials ........................................................................................... 14
Measures ......................................................................................................... 17
Results ............................................................................................................... 21
Induction check ............................................................................................. 21
Hypothesis testing ......................................................................................... 25
Discussion ...................................................................................... ' ................... 29
Conclusions ....................................................................................................... 31
Study 2 -- - -- - - ........... -- -- -- - - 32
Hypotheses ................................. ‘ ................................................................... 32
Method .............................................................................................................. 33

Participants ..................................................................................................... 34 '

Design and Procedure .................................................................................... 34
Stimuli/Materials ........................................................................................... 35
Measures ........................................................................................................ 35
Results ............................................................................................................... 37
Induction check ............................................................................................. 37
Hypothesis testing ......................................................................................... 40
Discussion ......................................................................................................... 43
I’ Conclusions ....................................................................................................... 44
Study 3 ............. - - 46
Hypotheses .................................................................................................... 46
Method .............................................................................................................. 50
Participants .................................................................................................... 50
Design and Procedure .................................................................................... 51
Stimuli/Materials ........................................................................................... 52
Measures ......................................................................................................... 52
Results ............................................................................................................... 53
Induction check ............................................................................................. 53
Hypothesis testing ......................................................................................... 55

Discussion ......................................................................................................... 61

 

 

 

Task demand and selective exposure ................................................. 61
Selective exposure and mood repair ............................................................... 63
- Conclusions ....................................................................................................... 65
General Discussion - - - ...... -- - -- 66
Limitations ..................................................................................................... 67
Future research directions ............................................................................. 69
Concluding Remarks -70
Appendices -- _- - - 71
Appendix A: Consent Form .......................................................................... 72
Appendix B: Video recording consent form ................................................. 73
Appendix C: Pre-test questionnaire ............................................................... 74
Appendix D: Study debriefing form ............................................................... 75
Appendix E: Questions from stress manipulation ........................................ 76
Appendix F: NASA-TLX (Task Load Index) ............................................... 97
Bibliography -_ 98

 

vi

LIST OF TABLES

Table 1. Controls available to the user in each task demand condition, at the start of game
play (Study 1) .................................................................................................................... 17

Table 2. Results of 2 (mood manipulation) x 4 (task demand) AN COVA on mood repair
(Study 1) ............................................................................................................................ 25

Table 3. Descriptive statistics for mood repair by mood manipulation and task demand
condition (Study 1) ........................................................................................................... 26

Table 4. Descriptive statistics and correlation matrix for measured engagement coding
(Study 2) ............................................................................................................................ 37

Table 5. Summary of regression analysis for the effect of measured engagement and post-
game play arousal on mood repair (Study 2) .................................................................... 40

Table 6. Regression results for the effect of measured engagement and post-game play
arousal on mood repair for bored and stressed participants (Study 2) .............................. 42

Table 7. Observed frequency of task demand choices by mood manipulation
(Study 3) ............................................................................................................................ 57

Table 8. Results of 2 (mood manipulation) x 3 (task demand) AN COVA on mood repair
(Study 3) ............................................................................................................................ 59

Table 9. Descriptive statistics for mood repair by mood manipulation and task demand
condition (Study 3) ........................................................................................................... 60

V11

LIST OF FIGURES

Figure 1. Predicted pattern of mood repair scores as a function of task demand condition,
split by mood manipulation with specific hypotheses labeled (Study 1) .......................... 10

Figure 2. Russell et a1. (1989) Affect Grid metric, adapted for use in the current study
(Study 1) ...................................................................................................................... 20

Figure 3. Induction check: Arousal and affect means for mood groups, pre- and post-
induction (Study 1) ........................................................................................................... 23

Figure 4. Observed relationship between task demand condition and subsequent mood
repair (Study 1) ................................................................................................................. 27.

Figure 5. Observed change in post-game play affect across mood manipulation
conditions, from baseline (Study 1) .................................................................................. 28

Figure 6. Observed relationship between mood state and task demand on mood repair
(Study 1) ............................................................................................................................ 29

Figure 7. Induction check: Arousal and affect means for mood groups, pre- and post-
induction (Study 2) ................................. , ......................................................................... 39

Figure 8. Predicted pattern of mood repair scores as a function of task demand condition,
split by mood manipulation (Study 3) .............................................................................. 50

Figure 9. Induction check: Arousal and affect means for mood groups, pre and post
induction (Study 3) ........................................................................................................... 54

Figure 10. Observed post-game play affect scores from Study 1 superimposed on
observed task demand selections from current study (Study 3) ................ . ....................... 56

Figure 11. Observed post-game play affect scores for bored participants in Study 1
superimposed on observed task demand selections for bored participants from current
study (Study 3) .................................................................................................................. 58

Figure 12. Observed post-game play affect scores for stressed participants in Study 1
superimposed on observed task demand selections for stressed participants from current
study (Study 3) .................................................................................................................. 58

Figure 13. Observed relationship between mood state and task demand on mood repair,
separated by mood manipulation condition (Study 3) ...................................................... 61

viii

INTRODUCTION

The ability of entertainment media to serve as a regulator of one’s mood state was
proposed by Zillmann and Bryant (1985). Their affective-dependent theory of stimulus
arrangement — referred to commonly as mood management theory — explains that
individuals are motivated to dissipate noxious mood states whenever possible, and will
make media choices in line with this motivation. The result of this selective exposure to
media fare is mood repair; that is, a marked shift in mood state from noxious to optimal.
This process has been tested with a variety of entertainment media, including television
and film (Bryant & Zillmann, 1984), music (Knobloch & Zillmann, 2002), and Internet
browsing behavior (Mastro, Eastin, & Tamborini, 2003). However, it has not been
examined in any great detail with video games.

Entertainment scholars assert that the experience of video game play is unique
among other media forms, and that features of technology responsible for this uniqueness
afford video games great mood management potential. For example, Grodal (2000)
suggests that video game differ significantly from ‘passive’ media such film and
television viewing in part because of the increased attention and physical engagement I
required for the ongoing experience of video game play to continue. In order to play a
video game, users have to pay careful attention to the game, make mental maps of game
environments, make note of objects and landmarks for future reference, and coordinate
visual attention with motor behavior (Grodal, 2000). Unlike more passive media such as
film, a video game proceeds only through the player's motivation to continue; this

continuation requires a user’s focused attention (Tamborini et al., 2004).

The present investigation begins by asserting that the unique demands placed on
users during video game play is in part responsible for the enhanced mood management
potential of video games. I maintain that, for individuals in a noxious state, playing video
games that place a greater task demand will result in greater mood repair. Moreover, I
assert that both (a) users in a noxious state will choose to play video games that offer a
high level of task demand and (b) the level of task demand in the games selected by these
individuals will predict mood repair.

To my knowledge, prior to this investigation no research has examined with any
scientific rigor suggestions that the increased task demand produced by a video game is
responsible for their mood management capacity, and that individuals in a noxious state
will choose games with greater task demand. The proposed set of studies will examine
these assertions in greater detail, specifically looking at how one key dimension of the
mood management process - intervention potential — can be used to explain why some
media forms might have a greater capacity to influence mood repair than others. More
specifically, this paper argues that video games, due to their greater intervention
potential, will be more effective in the mood repair process than other media forms such
as television when controlling for other mood management-relevant dimensions.

Mood management and selective exposure: An overview

Bryant and Davies (2006) recognize four dimensions mood management that
should be considered when examining the role of media use in mood regulation: arousal
regulation (via the principle of excitatory homeostasis), behavioral affinity, hedonic

valence, and intervention potential. Each is defined and discussed below.

Arousal regulation is understood as the ability for a medium to increase or
decrease an individual’s felt arousal. With respect to media, it refers to the tendency of
individuals to choose media that will help them achieve an optimal level of arousal.
Bryant and Zillmann (1984) showed evidence of this concept by experimentally inducing
boredom or stress in a group of undergraduate students before affording them an .
opportunity to watch a series of television programs. Consistent with their predictions,
participants in the boredom induction chose to watch exciting programming (e. g.,
highlights from a football game, an action-adventure show, and a game show) whereas
participants in the stress induction chose to watch more relaxing programming (e.g.,
segments of a travel documentary, orchestra concerts, and a nature program). Recent
work by Mastro, Eastin, and Tamborini (2002) replicated these results with Internet
search behavior, finding that bored individuals tended to surf the Internet more rapidly
than stressed individuals, and interpreting this behavior as consistent with the notion that
bored participants engaged in more highly interactive sessions that afforded greater
stimulation.

Behavioral affinity refers to similarity between message content and one’s current
affective state. For example, if an individual was in an aggressive (perhaps even hostile)
mood state, a violent boxing match would be understood to have a high level of
behavioral affinity, whereas a non-tendentious romantic comedy would have a relatively
low level of behavioral affinity. Selective exposure logic would predict that individuals in
an aggressive mood would avoid watching the boxing match, preferring the non-
tendentious romantic comedy in hopes of dissipating their noxious mood state. Work by

Zillmann, Hezel, and Medoff (1980) found support for this claim, reporting that study

participants who were provoked prior to viewing television (thus creating negative
affective states) avoided watching situational comedies that featured tendentious humor,
and action drama programming.

Hedonic valence is understood to be the general pleasurable or unpleasurable tone
of a message. One can point to the prototypical Hollywood ‘buddy comedy’ (e. g., Turner

and Hooch) as a genre with a generally pleasurable tone, and the blood-soaked ‘slasher’
film (e. g., Friday the 13th) as a genre with a generally unpleasurable tone. Similar

examples of pleasurable and unpleasurable tone can be easily thought of in music. The
aforementioned work of ‘Knobloch and Zillmann (2002) on music preferences perhaps
demonstrates this element most clearly, as participants induced into bad moods — using
the same bogus feedback induction as Zillmann et al. (1980) — chose to listen to music
rated as ‘energetic-joyful’ and were more decisive in their music choice. Thus, the tone of
a message can be understood to influence media choice based on one’s current mood
state.

Although the above-listed dimensions are important to understanding the mood
management process, the dimension most central to the current series of studies is
intervention potential, defined as the medium’s ability to capture an aroused individual’s
attentional resources (Bryant & Davies, 2006). Generally, it is argued that messages with
. higher intervention potential are more likely to distract an individual from the root cause
of their noxious mood state, thus hastening the mood repair process. Prior research
examining mood repair has demonstrated that the extent of media exposure’s intervention
potential is influenced by attributes found in message content (Zillmann, Knobloch, &

Yu, 2001; Knobloch, Hastall, Zillmann, & Callison, 2003). The current series of studies

differs from research focusing on content attributes, and proposes that the extent of
exposure’s intervention potential is also influenced by specific attributes of different
media forms — in this case, attributes that differ inherently between video games and
television.
Intervention potential in diflerent media forms

Both Bryant and Davies (2006) and Grodal (2000) maintain that the highly
interactive nature of video games and the user involvement required for the production of
its dynamic content demands more of a user’s attentional resources than other media
forms, such as television or film. If this is true, mood management logic would suggest
that this increased demand on a user’s attentional resources should results in greater
intervention potential. As such, we might expect the user involvement required by video
games to increase the medium’s ability to intervene in noxious mood states, and therefore
aid in the mood repair process. Vorderer (2000) offers a similar claim, arguing that video
games demand both cognitive and tactile engagement, and thus increase their intervention
potential over other media forms. Likewise, Klimmt and Hartmann (2006) note that video
games are specifically designed to demand steady streams of input from the user in order
for the game to progress, and therefore should have a higher intervention potential
relative to less demanding media, such as television. Beyond simply turning on the
television set and selecting a channel, one is hard-pressed to identify another point of
entry in which the user has any agency over television content in real-time. Tamborini
and Bowman (2010) advance similar logic with regard to the presence enhancing
attributes of video game technology. They hold that the vividness and interactivity

inherent in video game technology makes playing games an engaging and absorbing

experience. Although they focus on how these attributes should increase feelings of
presence, these engaging and absorbing attributes make playing video games precisely
the type of experience that should have great intervention potential. For example, while
traditional narrative media such as television and film require the user only to set the TV
tuner on a particular channel or start a VHS or DVD in order to View a complete
narrative, video games require near-constant user feedback from one point to the next.
The active user involvement required by video games commands a much higher level of
engagement than other media. This engagement not only prompts experiences of
presence, but also results in a high capacity for the video game to intervene in.
rumination.

Although the above-mentioned scholars have speculated that attributes of video
game technology afford it greater intervention potential than media such as television,
neither the claim of video game’s inherent superiority to television nor the processes
claimed to underlie this superior intervention potential have been empirically
demonstrated. To this end, three studies are proposed that aim to demonstrate how one
key difference between media forms — task demand — can influence mood repair and
selective exposure processes. The first study will vary levels of task demand to examine
the effect of task demand on mood repair processes. The second study probes the same
question as the first, using measures of actual user engagement during video game play
(i.e., the extent to which the respondent physically manipulates the game controls) to
predict mood repair process. Finally, the third study examines one’s selective exposure to
media environments that vary in the amount of task demand on individuals in order to

determine if individuals in a noxious state will selectively expose themselves to

environments with greater task demand and, subsequent to this, if mood repair resulting
from these choices is increased with the selection of media environments higher in task
demand.

As a whole, these studies examine a feature of video games thought to distinguish
this technology. It attempts to answer questions asking if the requirement for users to
actively engage in the production of dynamic media content when playing video games —
that is, the heightened level of task demand in video games — significantly increases the
intervention potential of this technology as compared to media such as television and

film, making them particularly well-suited to serve in a mood management capacity.

STUDY 1

The first study offers a comparison of mood repair under different conditions of
task demand created by the requirement for the user to actively engage in the production
of the dynamically changing media content. This study examines the claim that video
games are superior to television in their mood repair capacity, and that this superior
capacity results from the increased task demand created by the need to actively engage in
video game play. In this case, intervention potential is corollary to the level of task
demand in each video game play condition.

Hypotheses

The logic of this study begins with the assumption that video games have a higher
task demand than television (cf. Klimmt & Hartmann, 2006). From this, I propose to
examine existing claims that the heightened task demand in video games increases their
mood repair capacity, controlling for other mood management-relevant factors. As a
result of the intervention potential created by differences in task demand, I predict the
following (Figure 1, H1):

H1: For people in noxious mood states (boredom or stress), an increase in task

demand will increase mood repair.

When examining the effect of user-demand produced intervention potential on
mood repair, it would be imprudent to ignore the possibility for other recognized
dimensions of mood management, such as arousal regulation, to affect mood repair.
Though somewhat peripheral to the task demand hypotheses central to this study (H 1),
the potential influence of physical engagement and arousal regulation should not be

overlooked. When experimentally manipulating task demand, there is a natural confound

between physical and cognitive resources. Game play that requires more actions should
also be expected to increase an individual’s felt arousal. As such, arousal is measured and
controlled for in tests of H1.1

As boredom is a noxious state caused in part by very low levels of arousal, I
expect that any increase in felt arousal resulting from the higher levels of physical
engagement produced by a game requiring an increased number of actions should
produce a more positive effect on mood repair for those in a state of boredom than those
in a state of stress. Thus, bored individuals will experience a positive effect on mood
repair as a result of the intervention potential produced by heightened task demand as
well as a positive effect on mood repair resulting from arousal regulation. Conversely,
since stress is a noxious state from elevated levels of arousal, though stressed individuals
should experience the same positive effect on mood repair resulting from intervention
potential, they should not experience the same a positive effect on mood repair resulting
from arousal regulation. Thus, I tender the following hypothesis (Figure 1, H2): ‘

H2: The positive impact of task demand on mood repair will be greater for those

in a boredom condition than for those in a stress condition.

Notably, H2 predicts only a main effect for noxious state such that mood repair is
greater for bored individuals than it is for stressed individuals. Because the comparatively
greater mood repair predicted for those in the boredom condition results from elevated
arousal, the degree to which mood repair in the boredom condition surpasses repair in the

stress condition should mirror levels of arousal created by the game play conditions. As

 

l . . . . .

This becomes important specrfically when consrdermg the nature of boredom and arousal, the two
noxious moods states under investigation in the current study that can also be understood as
understimulation (boredom) and overstimulation (stress).

such, because physical activity and its resultant arousal should increase as task demand
increases, the extent to which mood repair in boredom conditions exceeds repair in stress
conditions should also increase as task demand condition increases. This leads to my
third hypothesis (Figure 1, H3):
H3: For those in a noxious mood state condition (i.e., boredom or stress), there
will be a disordinal, nonsymmetrical interaction between mood state and task
demand on mood repair. The increased mood repair predicted as task demand
condition goes from low to high will be greater for those in the boredom condition
than those in the stress condition.

Figure 1. Predicted pattern of mood repair scores for bored versus stressed players as a
fimction of task demand condition.

 

’8

8::

:6

>.

.‘2

Q.

0)

E Boredom
9”

:3 - - Stress
o

o..-

 

 

Low Medium 1 High

User demand condition

Finally, there is reason to suspect that the beneficial effect of task demand on
mood repair will peak at the point where a game becomes too demanding, and frustration
starts to set in (cf. Wolf & Perron, 2003). If this is the case, we might expect extreme
levels of task demand to counteract the mood repair process. Yet the point at which task

demand becomes noxious and starts to outweigh the mood repair benefits of intervention

10

is difficult to predict, as available research offers no insight on this matter. This makes it
is hard to predict where or if this point would be reached in any experimental induction of
task demand. As such, I began this study predicting positive linear effects of task demand
on mood repair, but I recognize the potential curvilinear influence of task demand on
mood repair and its possible effect on all three hypotheses. This potential curvilinear
relationship will be examined in data analyses in addition to tests of the predicted linear
effects.
METHOD

This study experimentally manipulated the intervention potential of media
exposure by way of task demand (i.e., the extent to which the user is required to
physically manipulate a video game’s controls) and observed subsequent differences in
mood repair on respondents placed in noxious states (conditions of boredom versus
stress). In a 2 (mood state) x 4 (task demand) between-subjects experimental design,
participants were randomly assigned to conditions of boredom or stress, and then asked to
play a flight simulator video game programmed to vary (low, low + cognitive task,
medium, high) in task demand.
Participants

Participants (N = 172) were recruited from a large, Midwestern university and
offered course credit and the chance for a $100 cash prize for participating in the study.
The sample contained 79 males and 93 females, with an average age of 21 years, five
months, and approximately 81 percent of respondents indicated that they were majoring
in communication or a related field (i.e., retailing, advertising, or public relations).

Notably, data collection was restricted here to a convenience sample of college students;

11

however, because the study examines how media exposure impacts mood repair in people
experiencing common noxious mood states (in this case, boredom and stress), I have no
reason to believe that this population possesses any unique elements that would be
expected to have an extraneous effect on the proposed hypotheses. In fact, an argument
could be made that college students would a more applicable population than others, as
this population represents a substantial portion of the gaming community (Jones, 2003).
To determine a proper sample size for this study a priori, a meta—analysis of affect
regulation research by Augustine and Hemenover (2008) was consulted. This meta-
analysis provided an average effect size calculation for pleasant distraction mood repair
strategies — such as those specified by mood management research according to
Parkinson and Totterdall’s (1999) affect regulation strategy taxonomy — as d = .41, which
- corresponds to an effect size f of .19 (effect size f is the proper measure of effect size for
analysis of variance (ANOVA) testing such as the analysis performed in this study).
However, it should be noted that this effect size measure was calculated from mood
management research using what would be classified in the current study as low task
demand medium — such as magazine articles, music, and television — which would be
expected to have a lesser effect on mood repair than high demand medium such as Video
games. Although research involving higher demand media is sparse, a secondary analysis
of preliminary data from Chen and Raney (2009) using video games as an agent for mood,
repair provides an average effect size for change in positive and negative mood after
game play of f = .31; this effect size measure was used to determine the sample size
needed for the current study. When comparing eight experimental groups using ANOVA

techniques with a < .05 one-tailed, a statistical power of B = .80, and an effect size of f =

12

.31, a minimum of 20 participants were needed per condition (N = 160); after all data was
collected, the final sample size of N = 172 was achieved.
Design and Procedure

Upon entering the lab, participants reviewed and signed informed consent forms
(Appendices A and B). Once consent was obtained, participants were given a sterilized
ear clip to affix to their right ear lobe (a heart rate measurement device) and asked to
complete a questionnaire measuring perceived video game skills and demographic
characteristics (Appendix C). After the questionnaire was completed and heart rate
measures stabilized, participants played or viewed the flight simulator for five minutes to
become familiar with the game. Following this session, participants were subjected to
either the boredom or stress mood induction. After the induction, participants completed
a mood measure (as an induction check) and then played (or viewed) the video game for
approximately two to three minutes. During game play, task demand was measured with a
distractor task. Once game play was finished, participants completed another set of mood
measures (to measure change in mood since induction). Finally, participants completed a
questionnaire containing measures of perceived task demand and overall game evaluation
before being fully debriefed (Appendix D); the game evaluation measures were part of a
separate data collection and thus are not discussed further in this manuscript. The study
lasted about one hour in total.

The four levels of task demand (low, low+, medium, high) were created by
varying the number of actions required to play the video game. In the low demand
condition, game play was set at auto-pilot on to simulate watching a television program

(i.e., no actions required). In the medium demand condition, play was set at auto-pilot

l3

half on (i.e., some actions required). In the high demand condition, play was set at the
auto-pilot off (i.e., full actions required). Finally, concern that participants in the low task
demand condition would simply ignore the video game entirely led to the inclusion of a
fourth condition labeled low+ task demand, in which a cognitive demand element was
added to the low demand condition described above. The logic for the addition of the

low+ demand condition was as follows. The goal of the low task demand condition was

to simulate the level of demand found in television viewing.2 If participants exposed to
Avideo game play in demonstration mode simply ignored the game, this would not be an
accurate corollary to television viewing, which surely contains a cognitive or information
processing element (cf. Lang, Bolls, Potter, & Kawahara, 1999). The inclusion of the
low+ condition was intended to add such a cognitive processing element to the low
demand condition.
Stimuli/Materials

Mood inductions. Consistent with prior research (Bryant & Zillmann, 1984;
Mastro et al., 2002), participants were induced into either boredom or stressful affective
states. Each induction required participants to perform a particular task for 20 minutes.
For the boredom induction, participants were given a large box of metal washers, and
asked to thread the washers onto a length of string. For the stress induction, participants

were asked to complete a booklet of difficult logic puzzles designed to exceed the talents

 

2 Some might question why a video game set on auto-pilot was used to simulate television viewing in the
low task demand condition instead of merely showing similar footage from a television show. Both options
were considered but, ultimately, the auto-pilot setting was used for the low-demand condition for two
reasons. First, the use of a television viewing condition with unique content would introduce unwanted
content differences in the experimental conditions. Second, although it can be argued that playing a game
in auto-pilot is not the same as watching television, the level of task demand required to play the game in
auto pilot can be equated with the task demand required for viewing television (discussed later in the
manuscript). As such, use of the auto-pilot condition allows us to test claims that task demand increases
intervention potential of game play over media experience that closely simulate television viewing.

14

of the participants (Appendix E). Furthermore, participants in the boredom induction
were left to their own volition, whereas participants in the stress induction were under
constant pressure fi'om an experimenter to perform better. The validity of these
inductions — as well as the amount of time allocated to achieve the desired boredom
(under-stimulated) or stressful (over-stimulated) states — has been demonstrated in prior
research (Bryant & Zillmann, 1984; Mastro et al., 2002).

Video game. The video game played in this study was Lock-0n: Modern Air
Combat, released by Ubisoft in November 2003 and promoted as “an ultra-realistic
[combat flight] simulator with faithfully rendered physics, weather, and avionics”
(Gametap, 2009, para. 1). The game is played using the Saitek X36F flight stick and
X35T throttle in tandem with a standard PC keyboard and mouse, based on how the
. game’s controls are configured. Lock-0n was particularly well-suited for this study for
two reasons: the game has fully programmable flight controls (which allow the
experimenter to turn on or off any number of game controls) and variable auto-pilot
capability (which allow the experimenter and the user to give varying amounts of game
control to the video game itself). In all experimental conditions, participants are asked to

pilot the same Russian Su-27 Flanker, a jet fighter built in the late 19705 for the Soviet
military by Sukhoi Design Bureau. The jet is known for its agility and ease-of-
maneuverability, and variants of the aircraft are still manufactured today.3 In terms of its
use in the study, the Su-27 is offered as a basic training aircraft in the video game for

practice on maneuverability and basic flight principles.

 

3 Russian Su—27 aircraft were used as recently as the 2008 South Ossetia War (Chang, 2008).

15

Task demand induction. All participants began playing the flight simulator game
at the same starting point, with the aircraft configured for a final approach toward the
landing strip. For the low task demand condition, participants played the game with full
auto-pilot engaged and all user controls turned off; that is, the game did not require any

input from the user in order to progress from flight to landing, akin to television

viewing.4 For the medium task demand condition, participants were in command of the
flight controls used to control the speed and direction of the plane (the joystick, throttle,
and rudder), while the simulator automatically controlled all other avionics for the
participant. These avionics include the landing gears (used to safely land the plane on a
landing strip), landing flaps (used to help increase drag to bring the plane to a safe '
landing speed), airbrakes (used to help bring the plane to a safe ground speed by
increasing drag), wheel brakes (used to help slow the speed of the plane once on the

landing strip), and drogue chute (a small parachute used to aid in slowing the plane on the

ground).5 For the high task demand condition, participants were in fiill control of all
simulator flight controls with no assistance from the computer. In terms of task demand,
the high task demand condition required participants to control both the speed and
direction of the aircraft along with all other flight operations. Whereas participants in the
medium task demand condition will need only to use the flight controls to land the plane
safely, participants in the high task demand condition will need to use both the flight

controls and all avionics to land the plane safely. The low+ task demand condition was

 

As game play 1n the low task demand condition IS set to auto-p1lot in order to more closely Simulate TV
viewing, these participants did not actually “play” the game. For lack of a better word, the word “play” is
used throughout this paper to represent exposure in all three conditions.

5 . . . . .
These avromcs were actually be controlled by a remote keyboard in the possess1on of the pr1mary
investigator and hidden from view, although their control appeared automated to the research participant.

l6

simply a replication of the low task demand condition with the addition of a
memorization task intended to increase cognitive demand without increasing interaction
with the video game. This memorization task required participants to take mental note of
all of the avionic settings in the aircraft in preparation for an exam on aircraft landing
techniques to be taken at the end of the study; the exam was never actually administered.
Notably, all game play sessions were restricted to five minutes. Table 1 lists the different
controls made available to the participant in each of the three experimental conditions.

Table 1. Controls available to the user in each task demand condition, at the start of game
play.

 

 

Low task demand* Medium task demand High task demand
Flight controls Flight controls Flight controls
0 [none] 0 Joystick o Joystick
o Throttle o Throttle
o Rudders o Rudders
AViOHiCS Avionics Avionics
0 [none] 0 [none] 0 Airbrake

o Landing flaps
o Landing gear
0 Drogue chute
0 Wheel brakes

 

'"T he low+ task demand condition replicates the controls of the low task demand
condition, with the addition of the memorization task.

Measures

Distractor task. Task demand was measured using a distractor task consisting of a
small black box with a red button and red LED. Participants were asked to press a red
button in response to an audio cue (a loud, digitized beep) to activate the LED, and their
response time was measured to a precision of 1/ 100 of a second. Slower reactions were
indicative of greater task demand. Similar distractor tasks have been used in research on
driver safety (Strayer & Johnston, 2001; Nunes & Recarte, 2002) in which research

participants respond to a visual distraction. However, as video game play has been shown

17

to increase one’s ability to pay attention to visual distracters (e.g., Green & Bavelier,
2003), participants in the study were asked to respond to an audio cue. The audio cue
technique has been used successfully in research on cognitive capacity in response to
media messages (e.g., Lang, Bradley, Park, Shin, & Chung, 2006). The audio cue was
played eight times during game play, and the reliability of measurement for these reaction
time responses was a = .812.

Mission feedback. Along with the behavioral measure of task demand, the NASA-
Task Load Index (NASA-TLX, Appendix F) was used as a self-report measure of
subjective workload assessment. This six-item, 20-point scale is designed for use in
measuring workload in human-machine interactions (N ASA-TLX, n.d.), and has been
used in prior research on flight simulations (of Moroney, Reising, Biers, & Eggemeier,
1993; Rueb, Vidulich, & Hassoun, 1992; Schweingruber, 1999). Sample items from the
scale were: “How much mental and perceptual activity was required?” and “How much
physical activity was required?” One item from the scale designed to measure perceived
performance had a negative effect on scale reliability, and thus was dropped from
subsequent analysis. The reliability of the remaining five-item scale was a = .811.

Mood repair. Mood repair was measured using a pre-test/post-test administration
of an adapted version of the Affect Grid (Russell, Weiss, & Mendelsohn, 1989), see
Figure 2. The scale asks participants to visually map their current mood state in the
semantic space between positive and negative affect (the x-axis) and high or low arousal
(the y-axis) using a 9 x 9 grid, with the square at the center of the grid representing a
“neutral, average, everyday feeling” (Russell et al., 1989, pp. 501). My adaptation to this

scale deals with a change to the labeling of the bottom-left quadrant of the scale which

18

was originally labeled as “Depression” in prior research. As this label is the conceptual
equivalent of boredom in research on mood repair (characterized as a negative affective
state in which one is experiencing low levels of arousal), I have re-labeled this quadrant
as “Boredom” in the scale provided to participants in my study. The scale has been
validated in prior research as a measure of mood and affect (Russell et al., 1989; Killgore,
1998; Swindells, MacLean, Booth, & Meitner, 2007), and the pre-test/post-test
implementation of the scale has been established as a valid measure of mood change

(e.g., De Petrillo & Winner, 2006; Eich & McCaulay, 2000).

19

Figure 2. Russell et al. (1989) Affect Grid metric, adapted for use in the current study.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

HIGH
STRESS AROUS AL EXCITEMENT
UNPLEASANT PLEASANT
FEELINGS FEELINGS
BOREDOM SLEEPINESS RELAXATION

Felt arousal. Although the Affect Grid takes into account measures of self-
reported arousal, a subsequent physiological measure of arousal was taken using the Cat-
Eye PL-6000, a heart rate monitor that measures in real time the wearer’s current pulse
rate in beats per minute. Heart rate has been used as an indicator of arousal response to
media (e.g., Malmstrom, Opton, & Lazarus, 1965; Snook-Luther, 1990; Riddle, 2004) as
it shares a strong positive correlation with skin conductance measures, another indicator
of arousal used in media research (e.g., Lang et al, 1999). It has been argued by some
(e.g., Freeman et. al, 2008) that arousal-induced increases in heart rate might be
obfuscated by attention-induced decreases in heart rate as the heart muscle is involved

both sympathetic (i.e., the ‘fight or flight’ response responsible for increased heart rate as

20

indicative of arousal) and parasympathetic (i.e., the ‘rest and digest’ response responsible
for decreased heart rate as indicative of attention) nervous system. However, prior
research by Kahana, Gopher, Grunwald, Iani, and Lavie (2004) examining these variables
in flight simulators has reported an overall increase in heart rate as a function of both
arousal as well as increased work load; the construct of work load in their study is the
conceptual equivalent of task demand as defined in the current research. Heart rate was
recorded at 10 second intervals for the duration of video game play and the mood
manipulation, and the score was converted as a “change from baseline” heart rate
measure, with a score of “0” equaling no fluctuation from baseline, positive scores
indicating increased heart rate from baseline, and negative scores indicating decreased
heart rate from baseline.

Unfortunately, measured arousal via the heart rate monitor did not differ
significantly between mood manipulation conditions as would be expected, t(170) = .026,
ns. Average heart rate for bored participants after mood manipulation was M = .045
above baseline, SD = 5.73, whereas average heart rate for stressed participants was M =
.024, SD = 5.08. Moreover, I found bored participants to actually have a higher heart rate
than stressed participants, which was opposite of my expectations based on both prior
research using the mood inductions as well as data from the self-reported arousal
measures. In addition, the heart rate measure had no significant correlation with the self-
reported measures of arousal (r = .019 with pre-game play arousal, r = .003 with post-

game play arousal). This data, coupled with previously—mentioned issues regarding the

21

use of heart rate as a valid measure of arousal (e. g., Freeman et. al, 2008), led me to drop

the heart rate measures in subsequent data analysis.6

Video game skill. Participant’s perceived video game skill was assessed using the
lO-item, seven-point Likert-scaled Game Playing Skill scale (GaPS; Bracken & Skalski,
2006). Sample items from this scale include: “I am a good video game player” and “I
often win when playing video games against other people.” The reliability of this lO-item
scale was a = .967. Sample mean for GaPS was M = 3.67, SD = 1.59, and an unexpected

significant difference was found across task demand conditions, F (3,168) = 2.706, p =
.047, n2 = .046. Tukey’s honest significant difference (HSD) test reported that

participants in the low demand condition (M = 3.11, SD = 1.62) self-reported a lower
level of game skill than participants in the high demand condition (M = 4.02, SD = 1.51).
Thus, video game skill was used as a covariate in all analyses.
RESULTS

Induction checks

Mood. The mood inductions used in this study were found to significantly affect
arousal and affect levels in the predicted direction.7 The means are reported in Figure 3.
For participants in the boredom condition, the induction produced the expected
significant shift in both arousal, t(109) = 12.0, p < .001, and affect, t(109) = 11.4, p <

.001. For participants in the stress condition, the induction again produced the expected

 

6 . .

For the sake of completeness, all reported analyses 1n the study were run usmg heart rate as a control
variable along with the self-reported arousal measure; the variable was not a significant covariate, and the
results of these re-analyses did not differ from those reported here. Thus, they are not reported.

There was no Slgmficant difference in either pre-inducuon arousal, t(221) = -.1 18, ns, or pre-induction
affect, t(221) = 1.46, ns, between mood conditions.

22

significant shift in both arousal, t(112)= -3.88,p < .001, and affect, t(112) = 12.1,p <
.001. Thus, I conclude that my mood manipulations were successful in inducing feelings
of boredom and stress in my participants.

Figure 3. Induction check: Arousal and affect means for mood groups, pre- and post-

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

induction.
HIGH
STRESS AROUSAL EXCITEMENT
Post-Manipulation [a
Arousal = 6.24b . . LI
Affect = 2.86., Pre-Manrpulatron
Arousal = 5.23a
\ b Affect = 5.53al
UNPLEASANT \ PLEASANT
FEELINGS \) FEELINGS
y/ Pre-Manipulation
Arousal = 5.19a __
Z/ Affect = 5.9021

 

 

 

 

 

 

Post-Manipulation
Arousal = 2.84b
Affect = 3.091,

BOREDOM SLEEPINESS RELAXATION

 

 

 

 

 

 

 

 

 

 

 

 

 

a = boredom, b = stress
Note: Means with different subscripts differ significantly at p = .05 or greater.

Task demand. An induction check was also performed on the task demand
conditions using scores from the distractor task measure. For the distractor task, a
univariate analysis of variance (ANOVA) reported a significant difference between task
demand condition and response time, F (3,168) = 20.2, p < .001 , 112 = .27. As expected,
participants in the low task demand condition had the fastest reaction times to the

distractor task (M = 1.74 seconds, SD = .41), followed in order by those in the low+ (M =

23

2.58, SD = 1.48), medium (M = 2.81, SD = 1.69), and high task demand conditions (M =
4.48, SD = 2.48); this trend followed a linear pattern, F linear(l,168) = 53.2, p < .001.

Although differing significantly from the low and high task demand conditions, post-hoc
analyses using Tukey’s HSD test found that reaction times for participants in the low+
and medium task demand conditions did not differ significantly from each other. This
suggests that the added cognitive demand from viewing video game footage with a
memorization task was not significantly more distracting than playing the video game
with a moderate level of control.

Data from the mission feedback scale told a similar story albeit not as
pronounced, as participants in the low task demand (M = 7.13, SD = .533) and low+ task
demand (M = 7.32, SD = .546) conditions evaluated these conditions as less demanding

as those participants in the medium task demand (M = 10.3, SD = .55) and high task

demand (M = 11.8, SD = .493) conditions, F (3,167) = 19.3, p < .001, n2 = .26. In contrast

to the self-reported task demand measures, post-hoc analysis using Tukey’s HSD showed,
no significant difference in means between scores in the low and low+ task demand
groups, suggesting that perhaps the added cognitive task did not have a perceptible effect
on task demand. Notably, these two measures of task demand shared significant
correlation (r = .241, p = .001). These results are used to conclude that the a priori task
demand conditions indeed differed significantly in the predicted direction. Moreover, as
results regarding the relative difference in task demand resulting from the low+ demand
condition were not made clear from my induction check, all four task demand conditions

were retained in data analysis.

24

Hypothesis testing

My hypotheses predicted that (H1) increased task demand will result in greater
mood repair for those in noxious mood states (H2) mood repair would be greater for
bored participants than for stressed participants, and (H3) a significant interaction
between mood induction and task demand conditions would exist such that mood repair
would be greater for bored participants than for stressed participants. To examine these

predictions, an omnibus 2 (mood manipulation) x 4 (task demand condition) AN COVA

. 1...:__“‘~m£

was performed with post—game play affect as the dependent variable, and. post-game play

arousal and perceived video game skill as covariates.8 ANCOVA results are presented in
Table 2, descriptive statistics for mood repair as a function of mood manipulation and
task demand condition are presented in Table 3, and resultant pattern of means by mood
manipulation and task demand condition are plotted in Figure 4.

Table 2. Results of 2 (mood manipulation) x 4 (task demand) ANCOVA. on mood

 

 

repair.*
. ss df MS F p 112

Video game skill 14.9 1 14.9 4.53 .035 .027
POSt’game play- 7.35 1 7.35 2.23 .137 .014
arousal
Task demand 24.6 3 8.20 2.49 .062 .044
Mood manipulation 15.0 I. 15.0 4.56 .034 .027
“51‘ demaf‘d by. 10.6 3 3.52 1.07 .362 .019
mood mampulatron
Error 533.1 162

 

*covariates in this analysis are post-game play arousal and perceived video game skill

 

8 Pre-game play affect was not used as a covariate, as these scores did not differ significantly across task
demand conditions, F (3,162) = 1.95, p = .124 or mood manipulation conditions, F (1,162) = .128, p = .721.

25

Table 3. Descriptive statistics for mood repair by mood manipulation and task demand
condition.

 

Task demand condition

 

 

 

Low Low+ Medium High
M SD M SD M SD M SD
Boredom 4323a 1.83 5_4()b 1.39 5.33c 2.09 5.54b 1.91
Stress 443:1 1.72 4,7oa 1.66 5201» 1.91 4,32a 2.04

 

Note: Means with different subscripts per row difler at p < .05- level or greater using
T ukey ’s honest significant diflerence (HSD) post—hoc test.

My first hypothesis predicted that increased task demand will result in greater
mood repair for those in noxious mood states. Results from the AN COVA analysis show
that the effect of task demand approaches significance but fails to meet the p < .05
criterion, F (3,162) = 2.49, p = .062, n2 = .044, thus suggesting initially that support was
not found for H1. However, as suggested earlier, there was reason to suspect that the
influence of task demand on mood repair might be nonlinear, as too much task demand
might become frustrating and as result offset any positive changes in mood (cf. Wolf &
Perron, 2003). In fact, the possibility that video games with very high task demand may
be counter-productive to the mood repair is suggested in the present data by inspecting
the pattern of mood repair means across conditions of task demand (see Table 3 and
Figure 2). Thus, tests were conducted to investigate the potential for a quadratic

relationship between task demand and mood repair, and indeed a significant curvilinear
relationship was found (k matrix plinea, = .096, pquadmn-c = .035, pcubic = .442). The pattern

of means shows that levels of mood repair increased fiom low (M = 4.40, SD = 1.75), to
low+ (M = 5.05, SD = 1.55), to moderate levels of task demand (M = 5.50, SD = 2.00), at

which point it peaked and subsequently declined for users in the highest task demand

26

condition (M = 4.92, SD = 1.05).9 Thus, there is support for H1 to the extent that
increased task demand has a significant positive effect on mood repair up to a point
before dropping off at the highest level of task demand (see Figure 4).

Figure 4. Observed relationship between task demand condition and subsequent mood
repair.

 

 

6.00 —
g 5.50
is
3 5.05
3 5.00 — 4.92
a
:3
2’"
8
“‘- 4.05
4.00 . . . 1
Low Low+ Medium High

User demand condition

My second hypothesis predicted that mood repair would be greater for bored
participants than for stressed participants. The ANCOVA analyses show support for this
hypothesis, as a significant difference was found in mood repair between mood

manipulation conditions, F (1,162) = 4.56, p = .034, 112 = .027. Although post-game play

affect (M = 4.96, SD = 1.88) was significantly greater than pre-game play affect (M =
3.02, SD = 1.65) for all participants, t(l7l) = -11.9, p < .001, the increase for bored
participants (AM = 2.20) was greater than the increase for stress participants (AM = 1.80),

and these differences were significant from one another, t(170) = 2.27, p = .024. Overall,

 

9 In fact, if we remove the high task demand condition from the ANCOVA analysis, the main effect for
task demand on mood repair is significant, F(2,123) = 3.19, p = .045, n2 = .053, and linear (k matrix punear

= ~013apquadratic = 364)-

27

mood repair was greater for bored participants than it was for stressed participants when
controlling for arousal and video game skill, thus showing support for H2 (also see
Figure 5).

Figure 5. Observed change in post-game play affect across mood manipulation
conditions, from baseline.

I Boredom
I" Stress

Affect scores

 

 

Pre-game play Post-game play

Note: Means with different subscripts within comparison group differ at p = .05 level or
greater.

My third hypothesis predicted that a significant interaction between mood
induction and task demand conditions would exist such that the impact of task demand on
mood repair would be greater for bored participants than for stressed participants (H3).
Results from the ANCOVA analysis showed that the interaction of task demand and
mood manipulation was not significant, F(3,l62) = 1.07, p = .362; thus, H3 is not

supported.

28

Figure 6. Observed relationship between mood state and task demand on mood repair.

 

 

 

6.00 - 5.83
E 5.54
‘5,
>4
.9
3 5.00 —
g Boredom
OD
é - - Stress
o \
“t 4.38 4.32
4.00 T T T 1

 

Low Low+ Medium High

User demand condition

It should be noted that a significant effect was found for video game skill on
mood repair, F(1,162) = 14.9, p = .025, n2 = .027. The partial correlation was calculated

between video game skill on mood repair, controlling for arousal, was r_= .177, p = .021.
This small-but—significant correlation suggests that individuals with higher self-reported
video game skill reported greater mood repair regardless of task demand or mood
manipulation.
DISCUSSION

Study I examined the effects of increased task demand on mood repair for bored
and stressed individuals. All participants in the study regardless of task demand condition
or mood induction experienced some form of mood repair post-game play, but bored
individuals experienced significantly more mood repair than stressed participants. At the
same time, task demand increased mood repair up to a point, beyond which too much
task demand had a counter-productive effect of mood repair regardless of mood

condition. This was evidenced by a significant curvilinear relationship reported for mood

29

repair scores as a function of increased task demand. Finally, there was not a significant
interaction between mood manipulation and task demand condition on mood repair. Data
fiom this study show that (a) increasing the amount of control an individual has over a
mediated environment — such as increasing the number of control inputs 3 user has in a
video game — significantly increases that medium’s intervention potential, (b) this
increase in intervention potential results in an enhanced ability for that medium to relieve
boredom and stress, (0) too much task demand can have a detrimental effect on mood
repair, and (d) the significant ability of video games to repair negative mood states is a
function of increased task demand, and not simply increased arousal.

Perhaps the most notable finding in this study is that the effect of task demand on
mood repair was observed after controlling for self—reported arousal. That is, I found that
the ability of video games to repair negative mood states associated with boredom and
stress was a function of increased task demand afforded by the interactive environment,
and not simply increased arousal as is commonly suggested in the literature (cf. Bryant &
Davies, 2006; Raney, Smith, & Baker, 2006). Of course, this is not to say that the arousal
capacity of video games is not an important contributor to their mood management
capacity; in fact, for individuals experiencing low levels of arousal, video games might
prove to be a most attractive media choice for helping one return to an optimal level of
arousal. However, it is important theoretically to demonstrate that intervention potential
and arousal regulation are separate constructs that can have differential effects on mood
repair, even in situations such as video game play in which it has been previously
assumed that both variables are increased as a function of increased control and

interactivity. Moreover, binding this added intervention potential to features of

30

interactivity (in this case, control inputs) demonstrates how unique features of video
games distinguish user experiences with video games from television and other less
interactive media. Not only do these data show that increased task demand contributes to
the mood management process, they also show how unique features of video games
contribute to. task demand.
CONCLUSIONS

Prior literature has suggested that the experience of playing video games is
qualitatively different than consuming other forms of media. In terms of mood
management and mood repair, it has been proposed that video games — due to their
increased task demand — should result in greater mood repair than other forms of media,
even when controlling for the effects of arousal and perceived video game skill. By
experimentally assigning bored or stressed individuals to various task demand conditions
designed to simulate television viewing or video game play, data from this study supports
the assertion that greater task demand is related to greater mood repair so long as demand
is not too high. Thus, Study 1 provides empirical support for the assertions regarding the
nature of video game play and intervention potential as applied to mood management

theory.

31

STUDY 2

Whereas the first study look at mood repair resulting from experimentally-
controlled levels of task demand, the second study is designed to predict mood repair
resulting from naturally occuning taskdemand, using measures of user engagement in
the production of dynamic media during video game play. Put simply, Study 2 uses
measures of the respondent’s real-time movements in the video game as an indicator of
task demand, and these movement measures are used as a predictor of mood repair.
Hypotheses

The mood repair logic for this study is borrowed in whole from the first study. It
is generally expected that users who are more active in the video game environment -‘
that is, those who are more active in the production of dynamic media content vis-a-vis
their increased manipulation of the game controls — will experience greater mood repair
than those users who are not as active in the video game environment. The first prediction
of Study 2:

H]: For people in noxious mood states (boredom or. stress), the extent to which

users physically manipulate the game controls will positively predict increased

mood repair.

Once again, and as found in Study 1, the natural confound of physical
engagement and arousal cannot be ignored. For bored users, mood repair is expected to
profit fi'om both the arousal regulation and intervention potential benefits that result from
increased task demand, whereas for stressed users, mood repair should profit only from
the intervention potential caused by task demand and not by arousal regulation. Because

of this confound and based on the data from the first study, I should expect the effect of

32

measured engagement on mood repair to be greater for bored users than for stressed
users. Thus, I predict that:

H2: The positive effect of measured engagement on mood repair will be greater

for those in a boredom condition than for those in stress condition.

Finally, in a similar fashion to Study 1, I wonder if there is a difference in how
participants might engage in a video game if given the added cognitive demand of a
memorization task. That is, I wonder if there might there be a difference in measured
engagement (and subsequent mood repair) for participants who play the game with or
without an added cognitive task. It makes sense that playing a video game under varying
levels of cognitive demand might have an influence on how one plays a video game, as
media users have a limited capacity to process media content (cf. Lang et al., 1999).
Moreover, demonstrating a difference — if any -— between a simple game-playing task and
a game-playing task in which one is asked to recall something about the event might be
another way to get at the question of information processing’s effect on mood repair
raised in Study 1. As with the prior study, any potential differences in measured
engagement and mood repair between the two game play scenarios will be examined in
my hypothesis testing.

METHOD

This study used natural measures of user engagement to predict mood repair
stemming from video game play. Following a mood induction, participants were video-
taped playing a flight simulator video game —— the high task demand condition from Study

1 — and their actions coded by two independent coders.

33

Participants

Participants were recruited in the same manner as Study 1, although participants
who took part in Study 1 were excluded from participating in Study 2. Assuming a
relatively large effect size f = .31 in Study 1 (see above discussion), a power analyses was
performed for a multiple regression model with three predictors, 01 < .05, two-tailed, and
a statistical power of B = .80; this analysis provided us with a optimal n = 40 for both of
my mood manipulation conditions. After data collection a total of N = 100 participants
(51 male, 49 female) were used in this study, with n = 47 in the boredom manipulation
and n = 49 in the stress manipulation. Due to errors in video recording discovered during
data analysis, four participants were removed from the sample (three from the boredom
manipulation, one from the stress manipulation), leaving a final sample size of N = 96.
The average age of participants was 21 years, one month, and 69 percent of participants
indicated that they were majoring in communication or a related field (i.e., retailing,
advertising, or public relations). As with Study 1, I am not concerned with the use of
college students as a threat to generalizability (cf. Jones, 2003).
Design and Procedure

The procedure from Study 1 is replicated here in its entirety, with one key
difference — rather than randomly assigning participants to one of three conditions of task
demand for the game play session, all users played the flight simulator in the high task
demand condition from Study 1. The high task demand condition was used in Study 2 so
as not to artificially limit the amount of engagement participants could have while
playing the study (recall that the other three conditions of task demand in Study 1 — low,

low+, and medium — were created by limiting the number of inputs the participant would

34

have control over in the flight simulator). In addition to the high demand condition, just
over half of participants (n = 51) were randomly assigned to the play the game with the
added cognitive task — the memorization task — from Study 1. To record participant’s
engagement in the video game, a video camera was used to record the participant’s
behaviors during game play. The camera was conspicuously mounted to the wall of the
laboratory directly above the computer monitor and angled so that it would record only
the participant’s hands on the game controls (joystick, throttle, and keyboard). All
participants consented to being video-taped prior to the start of the study (see Appendix
B), and no participants indicated being bothered or distracted by the camera during the
experiment.
Stimuli/Materials

Mood inductions. The mood inductions from Study 1 are replicated in Study 2.

Video game. The video game in Study 1 is used again in Study 2, with all auto-
pilot features turned off and all game controls fully operational; this is akin to the high
task demand condition from the previous study.
Measures

Distractor task. Task demand is measured in Study 2 using the same distractor
task measure from Study 1. The reliability of the measurement in this study was 01 = .658.

Mission feedback. The adapted NASA-TLX scale from Study 1 was used in Study
2 (Appendix F). The reliability of the scale in this study was 01 = .600.

Felt arousal. The arousal measure used in Study 2 is identical to the measure
from Study 1, although data from this measure was not used in statistical analysis for

similar reasons as the prior study.

35

Measured engagement. The participant’s actual manipulation of the video game
controls was measured in Study 2 and used in place of the experimentally-induced
discrete task demand conditions from Study 1. To measure user input, a digital camcorder
with a wide-angle lens was used to record all user controls (joystick, throttle, and
keyboard) for the duration of game play. Two independent coders — one male, one female
— viewed the footage and coded each participant’s use of the controls on a five-point
scale. Scale points were: “0” (no visible manipulation of control), “1” (control is being
manipulated, but the manipulation is at a minimum), “2” (control is being manipulated in
a calm, controlled manner), “3” (control is manipulated in an erratic, hurried manner),
and “4” (extreme manipulation of the controller, to the point where the controller itself is
actually moved on the desktop). One code was assigned per category to each participant,
and the unit of analysis for these codes was one game play session.

Before coding the study footage, both coders were given two sets of 10 sample
video recordings of participants playing the video game; this footage was taken from a
random set of participants in Study 1. Once an inter-coder reliability greater than K = .610
was established on each of the three measures (.610 considered to be the bare minimum
standard for claims of ‘substantial agreement’ among coders by Landis & Koch, 1977),
the coders were then asked to view the study footage. '0 Inter-coder reliability was greater
than K = .610 for all three variables: joystick K = .827; throttle K = .856; keyboard K =
.961. Not surprisingly, the three measures of user engagement did not form a uni-

dimensonal measure as evidenced by an extremely low reliability, or = .123. Although the

 

10 Coders spent ~ 45 minutes in training, and the K = .610 threshold was exceeded in both trial runs for all,
three coding categories: joystick (Kn-13] 1 = .625; Kma1 2 = .875), throttle (Kn-13] 1 = .625; Km-312 = .750), and
keyboard (Ktrial 1 = .875; ICU-1312 = .875).

36

joystick and throttle measures shared a small significant correlation with one another (r =
.254, p = .012), the keyboard measure was not significantly associated with either the
joystick (r = .091, ns) or throttle (r = -. 123) measures. Thus, for purposes of data analysis,
all three controls were retained in regression analysis so that their effect on mood repair
could be examined in tandem with and controlling for one another. The descriptive
statistics for all three measures and their correlation matrix are contained in Table 4, with
the means in this table representing the average assigned code per game play session for
all participants.

Table 4. Descriptive statistics and correlation matrix for measured engagement coding.

 

 

M SD Joystick Throttle
Joystick 2.41 .591
Throttle 1.78 .668 .254*
Keyboard l .42 .925 .091 -. 123

 

 

* = correlation significant at p < .05 or higher.

Mood repair. The mood measures used in Study 1 are again used in Study 2.

Video game skill. The video game skill measure used in Study 1 was used in
Study 2. Notably, there was no significant difference in video game skill between
participants in the boredom condition (M = 4.07, SD = 1.54) or the stress condition (M =
3.91, SD = 1.69), t(98) = .466, ns.

RESULTS

Induction checks

Mood. As in Study 1, the mood manipulations used in this study were found to
significantly affect both arousal and affect levels of participants in the predicted direction
(see Figure 7). Notably, there was no significant difference in either pre-manipulation

arousal, t(98) = .130, ns, or pre-manipulation affect, t(98) = 1.43, ns, between mood

37

conditions. By condition, the pre-manipulation means were: boredom affect, M = 5.88,
SD = 1.70; boredom arousal, M= 5.32, SD = 1.77; stressed affect, M= 5.33, SD = 2.06;
stressed arousal, M = 5.27, SD = 2.05.

For the boredom manipulation, post-manipulation affect (M: 3,88, SD = 1.67)
was significantly lower than pre-manipulation affect, t(96) = 5.88, p < .001. Similarly,
post-manipulation arousal (M = 2.86, SD = 1.53) was also significantly lower than pre-
manipulation arousal, t(96) = 7.36, p < .001. The boredom manipulation was successful.
For the stress manipulation, post-manipulation affect (M = 2.88, SD = 1.63) was
significantly lower than pre-manipulation affect, t(lOO) = 6.66, p < .001. Post-
manipulation arousal (M = 6.18, SD = 2.26) was also significantly higher than pre—
manipulation arousal, t(100) = -2.106, p = .038. The stress manipulation was successful.
Moreover, post-manipulation affect did not differ significantly between the two
conditions, t(98) = 1.52, ns, indicating that both mood manipulations induced a similarly-

valenced noxious mood.

38

Figure 7. Induction check: Arousal and affect means for mood groups, pre- and post-

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

induction.
HIGH
STRESS AROUS AL EXCITEMENT
-— Post-Manipulation
Arousal = 6‘18b Pre-Manipulation
—i Affect = 2'88b H Arousal = 5.2721
J\ Affect = 5.332,
UNPLEASANT ‘b\-o PLEASANT
FEELINGS I . FEELINGS
/ Pre-Manrpulation
a / Arousal = 5.32a
/ Affect = 5.888 F
t
Post-Manipulation
Arousal = 2.861,
Affect = 3.881,
BOREDOM SLEEPINESS RELAXATION

a = boredom, b = stress
Note: Means with diflerent subscripts differ significantly at p = .05 or greater.

Task demand. I wondered whether adding a cognitive task to high demand video

game play would result in increased task demand in a similar fashion to the added task

demand of a cognitive task to the low task demand condition in Study 1. This question

was posed in order to examine the relative effect of cognitive tasks in media

consumption. Mean comparison did not reveal as significant difference in either the self-

reported task demand between conditions, t(98) = -1.35, ns, or the behavioral measure of

39

task demand, t(98) = 1.38, ns. Thus, these two task demand groups were collapsed in

hypotheses testing.1 1
Hypothesis testing

The first hypothesis of Study 2 predicted a positive relationship between
measured engagement and mood repair such that regardless of mood manipulation,
participants who were more engaged with the video game (as assessed by independent
coders) should experience greater mood repair. This hypothesis was examined by
regressing mood repair scores on each of the three measured engagement scores assessed
by independent coders to control for the potential influence of post game-play arousal on
mood repair, this measure was included as a second step in a hierarchicallinear
regression model (see Table 5).

Table 5. Summary of regression analysis for the effect of measured engagement and post-
game play arousal on mood repair.

 

 

 

Model 1 Model 2
Variable B SE B B B SE B B
Joystick -.210 .359 -.061
Throttle .023 .314 .008
Keyboard .460 .223 .209*
Post-game play arousal -.266 .122 -.216* -.261 .122 -.211*
Video game skill .257 .124 .205* .207 .126 .165
R2 .091 .134
F for change in R2 4.65* 2.78*

 

* = p < .05
The overall regression model was significant, F (5 ,90) = 2.78, p = .022. Measures

of keyboard engagement were significant positive predictors of mood repair, [3 = .209, p

 

11 Upon further reflection, these null findings are not surprising given the limited capacity for individuals
to be able to process message content discussed earlier in Study 1 (cf. Lang et al, 2006). Whereas in a low
task demand situation the addition of cognitive tasks might be expected to increase demand, in a high task
demand situation there is a ceiling effect caused by the already-high resource demands of the video game.

40

= .042, whereas joystick and throttle controls were not (see Table 6). Moreover, the
measure of post-game play arousal served as significant negative predictors of mood
repair in the full regression model, [3 = -.211, p = .036, and the correlation between post-
game play arousal and keyboard manipulation was not significant, r = .013, ns, indicating
that they served as unique predictors of post-game play affect. Interpreted, greater
keyboard engagement increased post-game play affect, whereas greater post-game play

arousal decreased post-game play affect; both findings are expected from theory. Thus,

when considering measured keyboard manipulation, H1 is supported.12

The second hypothesis predicted that the effect of measured engagement on mood
repair would be greater for bored participants than for stressed participants. To examine
this, separate regression analyses were run for bored and stressed participants, and
differences in the magnitude and direction of beta-weights observed in the two analyses
were visually examined. Specifically, I looked at the beta-weights for all indicators of
user-control based engagement (joystick, throttle, and keyboard) to see if the pattern of
engagement’s effect on mood repair for bored participants differed from the same pattern
observed for stressed participants. The separate regression analyses run for bored and
stressed participants used the same predictor and control variables as above. Table 6

reports the results of these analyses.

 

12 One alternative explanation for the lack of significance associated with the measured joystick and
throttle manipulations could be their relatively low variance. Compared to measured keyboard
manipulation, which had a standard deviation on a five-point scale of .925, the measured joystick

manipulation had a standard deviation of .591, and measured throttle manipulation had a standard deviation
of .668.

41

Table 6. Regression results for the effect of measured engagement and post-game play
arousal on mood repair for bored and stressed participants.

 

 

 

 

 

 

 

 

Bored participants
Model 1 Model 2

Variable B SE B B B SE B B
Joystick .195 .551 I .057
Throttle .360 .404 ' .140
Keyboard .604 .303 .289*
Post-game play arousal -.187 .186 -.150 -. 173 .182 -.139
Video game skill .143 .191 .113 .109 .193 .086
R2 .041 .158

F for change in R2 945 1-53

Stressed participants
Model 1 Model 2

Variable B SE B B B SE B B
Joystick -.569 .468 -.172
Throttle -.833 .514 -.233
Keyboard .162 .326 .073
Post-game play arousal -.349 .159 -.298* -.291 .162 -.249
Video game skill .336 .162 283* .324 .161 .273*
R2 .154 .257

F for change in R2 418* 198*

 

"‘ = p < .05

Firstly, it should be noted that running separate regression models for bored and
stressed participants substantially lowers the power of both tests, as the sample size for
bored participants is n = 47 and the sample size for stressed participants is n = 49. Thus,
while significance tests are presented here, they are presented alongside a discussion of
the relative magnitude and direction of the reported beta-weights.

When conducting the regression analyses separately for bored and stressed
participants, the omnibus model for bored participants was not significant, F (4,42) =
1.53, ns, while the model for stressed participants was significant, F (4,44) = 2.98, p =
.021. For bored participants keyboard manipulation was found to be a significant positive

predictor of mood repair, [3 = .289, p = .050, but this finding was not replicated with

42

stress participants, B = .073, ns. Perhaps more interesting, the beta—weights for measured
joystick and throttle manipulation on mood repair were stronger and negative for stressed
participants ([3 = -. 172 for joystick measures, [3 = -.233 for throttle measures), whereas
these same measures were weaker and positive for bored participants ([3 = .057 for
joystick measures, 13 = .140 for throttle measures). Finally, we see that while video game
skill served as a significant positive predictor of post-game play affect for stressed
individuals ([3 = .273), it was not a predictor for bored individuals ([3 = .086). In addition,
the significant negative effect of post-game play arousal on mood repair reported in H1
was not replicated here, although the beta-weights were in the same negative direction for
both analyses. In fact, the negative influence of arousal on affect was stronger for stressed
participants than for bored participants as represented by a greater negative beta-weight
for these participants, which supports the earlier claims regarding the counter-productive
effect of task demand on affect for stressed participants. Combined, these data provide
evidence to suggest how the influence of measured engagement on mood repair might be
expected to differ between bored and stressed individuals. In general, it appears that
increased engagement with video game controls enhances post-game play affect for
bored individuals whereas it inhibits post-game play affect for stressed individuals; this
pattern of results is consistent with for H2. I
DISCUSSION

Whereas Study I examined mood repair from experimentally-induced levels of
task demand, Study 2 examined naturally-occurring engagement in a video game
environment as a predictor of mood repair. The results of this second study show that

while some measures of user engagement — namely, joystick and throttle control

43

manipulation — were not significant predictors of mood repair (albeit they trended in the
predicted positive direction), measures of keyboard manipulation were found to be a
strong and significant positive predictor of mood repair. Interestingly, this pattern was
only found for bored participants; for stressed participants, evidence suggests that
measured user engagement may in fact be a negative predictor of mood repair. Upon
further reflection, it is not surprising that the joystick and throttle controls did not have a
significant association with mood repair, especially given the nature of the flight
simulator control used in this study. In a flight simulator, joystick and throttle controls
require very deliberate and smooth operation that likely requires the user to maintain
rather than manipulate the position of these devices during game play. Conversely, the
keyboard in this particular video game was used to control the many other toggle
functions of the aircraft (such as landing gears, airbrakes, and other controls). Thus, the
user was required to constantly manipulate the keyboard in order to operate these aircraft
functions successfully, akin to a traditional video game controller. These points speak to
a limitation of the study that is highlighted later in the paper; nonetheless, this limitation
does not preclude the significant results found in support of H1 and H2.
CONCLUSIONS

Study 2 was designed to examine mood repair stemming from measured user
engagement in video games. Using behavioral measures of task demand as opposed to
random assignment to task demand conditions as was done in Study 1, data from the
second study shows that increased involvement with some game controls has a positive
influence on post-game play affect. In general, increased engagement of a video game’s

controls had a positive effect on post-game play affect for bored participants and a

44

negative effect on post-game play affect for stressed individuals. Specifically, increased
keyboard engagement (an indicator of greater command of the aircraft) was a significant
positive predictor of post-game play affect for bored participants, whereas increased
throttle and joystick engagement (whose overuse is indicative of lesser command of the
aircraft) was a negative predictor of affect for stressed participants. Perhaps this indicates
that bored participants desire more engagement with their video games, while stressed
participants desired less engagement. These data show that an individual’s mood state
can influence the mood repair associated with the physical actions they perform while

playing a video game.

45

STUDY 3

The final study examines the influence of mood state on selective exposure to
video games known to vary in task demand, and mood repair resulting from these
selections. This study differs from the previous two studies in that its central goal is to
demonstrate how the mood of players can be expected to influencetheir preference for
different video games; or, in this case, different configurations of the same video game
that vary in their predicted potential for mood repair. This is done to see if individuals in
a noxious state will selectively expose themselves to environments with greater task
demand and, subsequent to this, if mood repair resulting from these ‘naturally-occurring’
choices is increased with the selection of media environments higher in task demand.
Hypotheses

From selective exposure theory, and based on data fiom the previous two studies,
it is expected that users experiencing noxious mood. states will choose game conditions
known to have a higher amount of task demand than those not in a noxious mood. Yet as
found in Study 1, data on mood repair suggests a curvilinear pattern in which high levels
of task demand (compared to moderate levels) have a detrimental effect on repair. The
first study found that mood repair was highest at moderate levels of task demand,
decreased slightly at the highest level of task demand, and was significantly lower at the
lowest amount of task demand. As such, if selection is driven by anticipated mood repair,
we cannot assume that a noxious mood will results in an unlimited desire for added task
demand. Since this study replicates the task demand conditions used in the first study, the

game choice hypotheses are modeled after the pattern of mood repair means observed in

46

Study 1. Thus, given conditions in which players are familiar with the task demand levels
available in a video game, I offer my first hypothesis:

H1: If given the choice to play a video game with varying conditions of task

demand, a curvilinear choice pattern will be observed in which individuals in a

noxious mood state will prefer a game with low task demand less than one with

moderate task demand, and will prefer a game with moderate task demand more
than one with high task demand.

Related to this, 1 am also concerned with how this pattern of task demand choices
might differ as a function of the type of noxious mood state an individual is experiencing;
in this case, boredom or stress. Logic from Study 1 proposed that bored individuals might
benefit more from task demand than stressed individuals, and this claim was supported in
the data from that study. Bored individuals benefit not only from the intervention brought
on by increased task demand, but may also benefit from game play’s ability to heighten
arousal. This effect differs for stressed individuals. Although stressed individuals too
should benefit from the intervention brought by increased task demand, game play’s
ability to heighten arousal would not be expected to provide the same added mood-repair
benefit it might provide to bored individuals, particularly if task demand is at a high .
level. For stressed individuals, it was reported in Study 1 that a moderate task demand
game led to greater mood repair than either the low or high task demand game, and that
little mood repair from the high task demand game was apparent. By contrast, the mood
repair primacy of the moderate task demand condition was not apparent for bored
individuals. For these respondents, the moderate task demand game again led to greater

mood repair than the low demand game, but the repair benefits for the moderate demand

47

condition did not differ significantly from those in the high demand condition.

The second hypothesis for Study 3 is based on two previously identified
considerations. First, given that the logic for my game-choice hypotheses is based on the
understanding that selection is driven by anticipated mood repair, I would expect
differences in the game—choice behaviors of bored and stressed individuals to mirror
differences between bored and stressed individuals in the amount of mood repair
expected from games at different levels of task demand. Second, since the current study
replicates the task demand conditions used in Study 1, the game-choice hypotheses are
modeled after the pattern of mood repair observed in the prior study. As such, based on
the assertion that stressed individuals will anticipate greater mood repair in the moderate
task demand condition than the low or high demand conditions and that bored individuals
will not mirror this pattern, I predict that stressed individuals will show a greater bias
toward moderate task demand as compared to bored individuals.

H2: Selective exposure patterns will differ significantly between stressed and

bored individuals such that stressed participants will show a greater preference for

moderate levels of task demand over both low and high task demand, whereas
bored participants will not show the same preference for moderate levels of task
demand over both low and high task demand.

Although the primary focus of my study is to understand selective exposure
patterns stemming from mood states, the current study also examines mood repair
resulting from selective exposure. Study 1 presents logic that explains why bored
individuals should experience greater mood repair from games with increased task

demand than stressed individuals, as both the arousal and intervention potential of

48

increased task demand should benefit bored players, whereas stressed players would
benefit only from task demand’s intervention potential. Data from the first study
supported this logic, as indeed bored individuals experienced significantly greater post-
game play affect than stressed individuals (see Results section, Study 1). The third
hypothesis is based on an understanding that the goal of selective exposure behaviors is
to make media selections in line with motivations to facilitate. mood repair. From this, it
is expected that the highest levels of mood repair will be experienced by individuals who
make video game play selections in line with predicted patterns from H1 (the expectation
that moderate levels of task demand will be most preferred) and H2 (the expectation that
stressed individuals will show a greater demand for moderate levels of task demand than
bored individuals) Moreover, since only bored players should benefit from the arousal
potential of increased task demand due to the counter-productive influence of arousal on
stressed participants (see earlier discussion in Study 1), bored respondents should
experience greater mood repair than stressed respondents making the same choices.
Specifically, post-game play affect scores should be highest for bored participants ~
choosing a moderate level of user repair, and lowest for stressed participants choosing a
low level of user repair. The final hypothesis for Study 3 (also see Figure 8):
H3: For those in a noxious mood state condition (i.e., boredom or stress), there
will be a disordinal, nonsymmetrical interaction between mood state and task
demand on mood repair. The curvilinear relationship between mood state and task
demand in which the highest post-game play affect is found at moderate levels of

mood repair will be stronger for bored participants than for stressed participants.

49

Figure 8. Predicted pattern of mood repair scores as a function of task demand condition,
split by mood manipulation.

8

8:

CU

>4

.9

D.

0)

g —Boredom
9°

:73 - - Stress
o

o.

 

 

Low Medium High

User demand choice

METHOD

Study 3 examines how learned expectations of differential task demand in a video
game might influence selective exposure to that game. A 2 (mood manipulation) x 3 (task
demand) between-subj ects experimental design was employed. Participants played the
same flight simulator video game used in the previous two studies at all three levels of
task demand before being subjected to a randomly-assigned mood manipulation.
Participants

Participants were recruited in a similar manner to the first two studies, although
participants from either Study 1 or Study 2 were excluded from participating in Study 3.
Using an effect size measure of f = .31 (cf. Chen & Raney, 2009, also see earlier
discussion), an expected effect size of 00 = .40 ((0 is the proper effect size measure for
chi-square goodness-of-fit analyses as conducted in the current study) was used to
determine a priori the sample size for this study. Power analyses using (0 = .40 was

performed for a chi-square contingency table with three possible choices, (1 < .05, two—

50

tailed, and a statistical power of B = .80; this analysis provided us with a optimal N = 61;
after all data was collected, a final sample size of N = 64 was achieved for this study.
This sample contained 39 males and 25 females with an average age of 22 years, five
months, and 73 percent of participants self-reported majoring in communication or a
related field. As with the first two studies, data collection was restricted here to a
convenience sample of college students.
Design and Procedure

Upon entering the lab, participants reviewed and signed the informed consent
forms (Appendices A and B). Once consent was obtained, participants were asked to
complete a questionnaire measuring perceived video game skills and demographic
characteristics (the same survey completed in the first two studies, Appendix C). After
completing the questionnaire, participants were given five minutes to learn the controls of
the flight simulator game before playing each version of the video game (each of the
three task demand conditions) for five minutes. The games were played in the following
order of task demand: high, low, and moderate. After reach game play session,
participants were asked to complete a perceived task demand measure. Playing all three
games prior to the mood manipulation was important, as selective exposure processes are
driven by learned expectations regarding the medium’s capacity to repair mood that stem
from an individual’s prior experience with that medium (Atkin, 1985). After playing all
three versions of the game, participants were subjected to either the boredom or stress
manipulation, and then asked to choose one of the three task demand conditions to play
for five minutes. Task demand choice was recorded by the primary researcher, and

participants played the game. Once game play was completed, participants were fully

51

debriefed as to the purpose of the study (Appendix D) and entered into the cash drawing.
The entire procedure lasted about one hour.
Stimuli/Materials

Mood inductions. The mood inductions from Studies 1 and 2 are replicated in
Study 3.

Video game. The same video game used in Studies 1 and 2 is used in Study 3,
with the same task demand conditions used in Study 1 implemented here in Study 3.
Measures

Distractor task. The task demand measure used in Study 3 is identical to the
measure used in Studies 1 and 2. Note that the distractor task measure was not used in
data analysis for this study; rather, the measure was retained in the experimental game
play condition. to more closely replicate the same game play conditions from the first two
studies.

Mission feedback. The NASA-TLX scale from first two studies was used in Study
3 (Appendix F). Average reliability of this scale — with one item removed (see
explanation in Study 1) was a = .72; the scale was administered four times.

Arousal. The arousal measure used in Study 3 is identical to the measure from
Studies 1 and 2, although data from this scale was not used in analysis of Study 3 for
reasons explained in Study 1.

Mood repair. The mood measures used in Studies 1 and 2 are used in Study 3.

Selective exposure. To measure participant’s preference for a particular task
demand setting, the primary investigator recorded which version of the video game (low,

medium, or high task demand) participants chose to play.

52

Perceived game skill. The same measure of perceived game skill, and game
success used in the first two studies is replicated here in Study 3. Notably, perceived
game skill did not differ significantly between bored (M = 4.22, SD = 1.38) and stressed
(M = 4.46, SD = 1.62) participants, t(62) = -.639, ns.

RESULTS
Induction check

Mood. The mood manipulations used in this study were found to significantly
affect both arousal and affect levels of participants in the predicted direction. Notably,
there was no significant difference in either pre-manipulation arousal, t(62) = .395, ns, or
pre-manipulation affect, t(62) = .625, ns, between mood conditions. For the boredom
manipulation, post-manipulation affect was significantly lower than the pre-manipulation
affect, t(30) = 6.92, p < .001, and post-manipulation arousal significantly lower than pre-
manipulation arousal, t(30) = 5.42, p < .001. For the stress manipulation, post-
manipulation affect was significantly lower than the pre-manipulation affect, t(32) = 10.1,
p < .001, although post-manipulation arousal did not differ significantly from the pre-
manipulation arousal above, t(32) = -1.80, p = .081. However, post-manipulation affect in
the stress condition was significantly lower than post-manipulation affect in the boredom
condition, t(62) = 3.19, p = .003, indicating that the stress mood manipulation induced a
more noxious mood state than the boredom mood manipulation. Figure 9 contains a

graphical depiction of the mood manipulation induction.

53

Figure 9. Induction check: Arousal and affect means for mood groups, pre and post
induction.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

HIGH
STRESS AROUSAI. EXCITEMENT
Post-Manipulation
Arousal = 6.64a P M . 1 .
Affect = 2-06b I'C- ampu atlon
Arousal = 5.853 __
\\b Affect = 5-70a
1 .
UNPLEASANT P ANT
FEELINGS fl FELEE‘IISINGS
a Pre—Manipulation
Arousal = 5.5581
Affect = 6.1621
Post-Manipulation
Arousal = 2.771,
Affect = 3.411,
I I I
BOREDOM SLEEPINESS RELAXATION

Notes: a = boredom, b = stress, means with different subscripts across pre- and post-
manipulation conditions differ significantly at p = .05 or greater.

Task demand. Although not an induction per se, participant’s self-reported task
demand was measured after each of the three practice sessions with the video game as
indicated by their scores on the NASA-TLX. Note that the behavioral distractor task
measured was not used during the practice sessions due to concerns that participants
might simply learn to do the task better with multiple iterations. A repeated-measures
AN OVA reports that indeed a significant difference in perceived task demand was found

between the three video game conditions in the predicted direction, F (2,126) = 265.1, p <
.001, n2 = .808. Perceptions of task demand were lowest in the low task demand

condition (M = 3.72, SD = .389), moderate in the moderate task demand condition (M =

54

11.5, SD = .389), and highest in the high task demand condition (M = 12.6, SD = .350).
Thus, I can conclude the participants did perceive the expected significant difference in
task demand between task demand conditions.
Hypothesis testing

Selective exposure. My first hypothesis predicted that, given the choice to play
video games known to vary in task demand, a curvilinear choice pattern would be
observed among individuals in noxious mood states such that individuals would prefer
low amounts of task demand the least, moderate amounts of task demand the most, and
high amounts of task demand less so than moderate but more so that low amounts. To
examine this, I first conducted a chi-square goodness of fit test to see if the observed
pattern of game choice behaviors differed from chance. Then, I examined visually the
observed pattern of game choices and compared this to the pattern of post-game play

affect scores from Study 1. The chi-square goodness of fit showed that indeed the
observed pattern of game choice behaviors differed significantly from chance, x2 (2, n =

64) = 15.6, p < .001. Visual inspection reveals that the selection of game play with
varying levels of task demand appears to follow the predicted curvilinear pattern.
Participants chose the low task demand condition the least (7 observed choices), the
moderate task demand condition the most (32 observed choices), and the high task
demand in the middle (25 observed choices). As the chi-square analysis provides us
evidence that the observed game choice behaviors differed significantly from chance, and
simple observation of the pattern of game choices follows the expected curvilinear

relationship reported in Study 1, I conclude that the findings are consistent with for H1.

55

Figure 10 shows the post-game affect scores from Study 1 superimposed on the observed
task demand selections from the current study.

Figure 10. Observed post-game play affect scores from Study 1 superimposed on
observed task demand selections from current study.

6.00

Mood repair scores

 

Low Medium High

l Observed use demand selection

The second hypothesis predicted a significant difference in selective exposure
patterns between stressed and bored individuals; specifically, that stressed individuals
would prefer moderate task demand over high task demand more so than bored
individuals. To examine this, a chi-square goodness-of—fit test was conducted to
comparing the observed pattern of selected task demand exhibited by. bored participants
with the observed selection pattern exhibited by stressed participants. This test required
us to split the data file into separate groups of n = 31 (bored participants) and n = 33
(stressed participants), see Table 7. As at least one cell had a frequency of less than five,

Yate’s correction was applied to the final chi-square critical value (Yates, 1934).

56

Table 7. Observed frequency of task demand choices by mood manipulation.

 

 

Low Moderate High Total n
Boredom 6 13 12 3 1
Stress l l 9 1 3 33

 

 

1212.141, n = 31) = 18.8, p < .001.

The result of the test comparing task demand choices for bored and stressed

participants revealed that the pattern of choices in each condition varied significantly,
12173164 1, n = 31) = 18.8, p < .001. Although all participants showed aversion to the low

task demand condition, stressed participants showed greater preference for moderate task
demand than did bored participants, and both stressed and bored participants showed
equal preference for high task demand. Thus, I conclude that the findings are consistent
with H2. Figure 11 shows a comparison between the observed task demand selection
frequencies for bored participants in the current study and post-game play affect scores
for bored participants in Study 1. Figure 12 shows the same comparison for stressed

participants.

57

Figure I 1. Observed post-game play affect scores for bored participants in Study 1
superimposed on observed task demand selections for bored participants from current
study.

Mood repair scores

 

 

5' Obsu'ved use demand selection

Figure 12. Observed post-game play affect scores for stressed participants in Study 1
superimposed on observed task demand selections for stressed participants from current
study.

Mood repair scores

 

 

Low Medium High

l Obsu'ved user demand selection

Mood repair. The final hypothesis made predictions about the combined effect of
mood manipulation and task demand choice on post-game play affect. Specifically, this

hypothesis predicted that the curvilinear relationship between mood manipulation and

58

task demand in which the highest post-game play affect is found at moderate levels of

mood repair will be stronger for bored participants than for stressed participants. This

hypothesis was tested using a 2 (mood manipulation) x 3 (task demand) ANCOVA

model, with post-game play arousal, pre—game play affect, and perceived video game

skill as covariates. '3 Table 8 contains the ANCOVA analysis results, Table 9 contains the

descriptive statistics of this analysis, and Figure 13 graphs the descriptives.

Table 8. Results of 2 (mood manipulation) x 3 (task demand) ANCOVA on mood

 

 

repair.*
SS MS F 112

Video game skill 1.34 l 1.34 .285 .596 .005
POSI'game Play .019 1 .019 .004 .949 ~.000
arousal
Pure-game play 11.2 1 11.2 2.38 .129 .041
affect
Tas‘.‘ demand 2.56 2 1.28 .273 .762 .010
chorce
Mood manipulation 3.25 1 3.25 .694 .409 .012
TaSk “16‘“an by. 1.34 2 .668 .143 .867 .005
mood manrpulatron
Error 258 55

 

*covariates in this analysis are post-game play arousal and perceived video game skill

 

3 Pre-game play affect was used as a covariate because of a significant difference in this measure between
bored participants (M = 3.41, SD = 2.13) and stressed participants (M = 2.06, SD = 1.14), t(62) = 3.21, p =

.002 (see Induction check: Mood for Study 3).

59

Table 9. Descriptive statistics for mood repair by mood manipulation and task demand
condition.

 

 

 

Low Medium . High
M SD M SD M SD
Boredom 650a 1.76 5.92a 1.85 6-41a 2.39
Stress 5.00 - 5-16a 2.24 51521 2.27

 

Note: Means with different subscripts per row difler at p < .05 level or greater using
T ukey ’s HSD post-hoc text.

The AN COVA model does not support the proposed hypothesis. There is no main
effect for mood for mood manipulation on affect scores, F (1,55) = .694, ns, nor is there a

main effect of task demand choice on affect scores, F (2,55) = .273, ns. For task demand
choices, neither a linear or quadratic effect was reported (k matrix plinear = .635, pquadmn-c

= .876). Although overall the bored participants (M = 6.23, SD = 2.01) self-reported
greater affect than stressed participants (M = 5.15, SD = 2.18), the effect was not
significant —— no doubt in part due to a lack of statistical power resulting from small
sample sizes (n = 31 for bored participants, n = 33 for stressed participants). It should be
noted that bored participants began with a significantly greater affect score than the
stressed participants (see Induction check: Mood). In fact, bored participants overall
experienced less mood repair than stressed participants (AM bored participants = 2.82,
AM stressed participants = 3.09). Furthermore, for bored participants, post-game play
affect was actually lowest in the medium task demand choice; for stressed participants,
there is almost no perceptible change in affect scores across task demand conditions.
Combined, these data do not provide support for H3. This is surprising given the data
fi'om Study 1, but an alternative explanation for these results is presented in the

Discussion section below.

60

Figure 13. Observed relationship between mood state and task demand on mood repair,
separated by mood manipulation condition.

 

 

 

7.00 4
L 6-50 6.41
8
21:“ 5.92
> 6.00 ~
C6
'8.
Q)
g 5 00 5.00 __ 1,541.6- - _ .5'15 Boredom
c0 , ~ - "'
é - - Stress
O
o.

4.00 I , A

Low Medium High

User demand choice

DISCUSSION

The third study was aimed at predicting selective exposure to video game settings
known to vary in task demand as a result of experimentally-induced noxious mood states.
As predicted, participants demonstrated a clear preference for moderate levels of task
demand, and this preference was greater for stressed as compared to bored participants.
However, the effect of this process on mood repair was not found. These findings are
discussed in further detail below
Task demand and selective exposure

The first hypotheses -— and perhaps the primary focus of Study 3 — were concerned
with the influence of task demand on individual’s selective exposure patterns. Based on
the pattern of post-game play affect scores from Study 1 it was predicted that participants
experiencing noxious mood states would prefer moderate levels of task demand as
compared to extremely high or low levels, and that stressed participants would show a

more distinct preference for moderate task demand levels than would bored participants.

61

Indeed, both of these predictions were supported. After being given a set amount of time
to develop learned expectations about a video game at three discrete levels of task
demand, both stressed and bored participants showed the greatest preference for moderate
task demand levels as compared to low or high levels of task demand, and stressed
individuals showed a greater bias for moderate task demand than did bored individuals.
This data is compelling for two reasons. First, it lends further empirical support to my
assertions regarding selective exposure processes in newer forms of interactive media. In
this case, the data give us a closer look at how we can understand the intervention
potential in terms of the demand a medium places on its user, and how this task demand
can be used as a predictor of video game choice related to two orthogonal and
commonly-experienced mood states. Second, this research provides empirical evidence
for selective exposure assumptions regarding learned expectations that have gone largely
untested.

Selective exposure theory as applied to entertainment media is based partly on the
dual propositions that (a) previous encounters with media lead to learned expectations
related to a medium’s ability to satisfy certain needs, and (b) these learned expectations
drive media choice. Yet, I know of no research that has experimentally varied learned
expectations to examine their influence on this part of the theoretical process. This may
be largely due to the fact that most prior selective exposure research has examined forms
of traditional entertainment media so ubiquitous that learned expectations are difficult to
vary, or assumed to be universal. For example, when conducting selective exposure
research on television and film, it makes sense to assume that individuals have well

learned expectations about standard offerings well before they enter an experimental

62

setting, making it difficult to manipulate and test learned expectations’ influence. At the
same time, it is important to note that while the experimental design used in Study 1 is
unique form other selective exposure and mood management research in that participants
are given an opportunity to practice and experience each of the three experimental video
game conditions prior to post—mood manipulation selection, the design did not include a
no-practice control group. As such, the different levels of learned expectation against a
no-expectation control cannot be measured or compared with the current design. To
rectify this potential oversight, future research might consider a Solomon four-group or
other related design in replications of this study.
Selective exposure and mood repair

Although the primary focus of Study 3 was to examine the effect of task demand
on selective exposure to video games, this study was also concerned with how actual task
demand choices would influence subsequent mood repair. It was expected that, regardless
of task demand choice, bored participants would experience greater mood repair than
stressed participants, and that participants who made task demand choices in line with the
selective exposure predictions would experience the greatest mood repair. Neither
prediction was supported by the data. Despite earlier research, mood repair was not
significantly greater for bored participants, and mood repair was not significantly affected
by task demand choices. In fact, bored participants in the study who chose moderate
levels of task demand experienced the least amount of mood repair of all participants in
the study. Especially in light data from Study 1 in which task demand was found to have
a generally positive effect on mood repair, these findings were unexpected and require

firrther investigation. One potential explanation lies in the nature of the mood induction,

63

specifically that the stress induction produced a significantly more noxious mood state
than the boredom induction in this study. This difference in the intensity of noxious mood
states might have introduced a ceiling effect and, as a result, restricted the ability of bored
individuals to improve their mood state. This being said, the post-mood manipulation
affect scores from Study 1 (M = 3.06, SD = 1.62) were not significantly different from the
post-manipulation affect scores in the current study (M = 3.41, SD = 2.12), t(43) = -.835,
ns, so the ceiling effect argument does not seem to be a likely explanation for the
observed data trends. A second potential alternative explanation for the non-support for
the mood repair hypotheses is related to the behavioral measure of task demand (the
response time task) used during the post-mood manipulation game play session. Although
this measure was included to more fully replicate the experimental design of Study 1, it
might also have inadvertently worked against mood repair because the unexpected nature
of the task, especially given that the task was not performed in any of the three practice
game play sessions. It seems likely that the introduction of any task that was not in line
with participant’s learned expectations associated with game play would have created
some amount of dissonance in the minds of those playing the game with the added tasks.
Given the potentially disrupting nature of the response time task in relation to video game
play, a negative effect on post-game play affect is plausible. Finally, a third alternative
explanation for the non-significant effect of task demand choice on mood repair would be
that participants were in fact choosing the task demand condition that would create what
they considered to be an optimal mood state. In contrast to my expectation that all
participants would require moderate levels of task demand in order to reach an optimal

mood state, the optimal levels required may have differed across participants. Perhaps the

64

levels of task demand actually selected by individuals were optimal in each case. Given
this, no difference in post game play affect scores would be expected across task demand
selections. Although the mood manipulations successfully induced similar levels noxious
mood in both bored and stressed participants, 1 do not have data to suggest what
individual participants would consider to be an optimal mood; thus, this assumption
cannot be tested with these data.
CONCLUSIONS

The first two studies focused exclusively on the role of task demand in the mood
repair process using logic from mood management theory. The third study focused I
specifically on the role of learned expectations regarding a medium’s task demand on
selective exposure behaviors. Secondary to the selective exposure hypotheses, Study 3
also examined how these ‘naturally-occurring’ selective exposure behaviors influenced
the mood repair process. In line with the post-game play affect scores from Study 1 in
which moderate task demand was found to result in the greatest mood repair, in Study 3
participants showed the greatest preference for moderate amounts of task demand; this
preference was stronger for stressed participants than for bored participants. Finally, no
evidence was provided to support predictions regarding the influence of ‘naturally-
occurring’ task demand choice on resultant mood repair. Thus, although Study 3 provides
support for assertions regarding the effect of task demand on selective exposure to
varying conditions of task demand in video games, it does not provide evidence that these
selections are actually effective in producing the desired mood repair. Nonetheless,
evidence in Study 3 demonstrating the role of learned expectations in the selective

exposure to video games adds to our understanding of selective exposure processes.

65

GENERAL DISCUSSION

The presented series of three studies was designed to examine the role of task
demand on selective exposure and mood management processes. These studies come at
an important impasse in media research. Although ample research exists on video game
uses and effects as well as on selective exposure to and mood management in traditional
media, no known research to date has combined the two. Many scholars have asserted
that video games are distinct from other forms of entertainment medium in that they
require near-constant feedback from the user, yet prior to this investigation this claim has
not been supported empirically. Study 1 varied the amount of task demand in a video
game, and found that increasing task demand results in greater mood repair to a point, at
which too much task demand becomes detrimental to the mood repair process; this
process was more effective for bored participants than for stressed participants. Study 2
used different behavioral measures of user engagement and found that, in general,
increased controller manipulation had a positive influence on mood repair for bored
individuals and a negative influence on mood repair for stressed individuals. Study 3
focused on selective exposure stemming from learned expectations following game play,
and found an overall preference for moderate levels of task demand; this preference was
more pronounced for stressed individuals. Combined, the three studies presented here
offer data to establish the role of task demand as a feature of intervention potential
relevant to understanding how selective exposure and mood management processes might
differ from traditional media (such as television) to more interactive media (such as video

games).

66

Limitations

Although limitations associated with each of the individual studies have been
discussed earlier in this manuscript, a few limitations common to the all studies are
discussed here. These limitations are related to the relatively short amount of time
participants spent playing video games in the studies, concerns with the simulated
television viewing condition, and the use of college-aged students. Each of these potential
limitations is discussed below.

For each of the three studies presented here, participants were given no more than
five minutes to play each version of the video game. Although this length of time seemed
to be sufficient enough to demonstrate both mood repair and selective exposure to task
demand conditions, the present series of studies does not investigate the potential for an
important relationship to exist between time spent playing video games and task demand.
For example, is it possible that as an individual spends more time with a video game and
become increasingly familiar with the game’s controls, the game might become less
demanding. In this case, we might expect task demand to be highest upon first exposure
with a video game, and reduced with each subsequent exposure (or, task demand may be
reduced simply as a function of time spent playing during any one exposure). In terms of
mood repair, we might expect the mood repair capacity of a video game to diminish as a
function of increase time spent playing the game; in terms of selective exposure, we
might expect that exposure to video games over time would create more stable learned
expectations regarding a video game’s mood repair capacity. At the same time, one might
also consider that video games are programmed to be more difficult as the player

progresses through the game (cf. Boyan & Bowman, 2007), thus maintaining a

67

heightened level of task demand as play is continued. Nonetheless, future research should
consider this relationship with more experimental rigor.

As discussed earlier, the low demand condition used in the current set of studies
was designed to simulate television viewing while keeping media content as constant as
possible. After weighing several options — including showing training footage of aircraft
landings from Boeing and Lockheed-Martin sales videos — the decision was made to have
participants in the low task demand conditions watch video game footage in I
demonstration mode rather than providing them with actual television programming to
watch. Showing demonstration footage from the video game was required inorder to
attribute any observed differences in selective exposure and mood repair to differences in
task demand. Controlling media content allows for the dismissal of claims that the
observed effects in this set of studies resulted from differences in the hedonic valence and
behavioral affinity of media content. At the same time, using the simulated television
viewing condition might reduce the generalizability of the reported results. Replications
of this study would benefit by incorporating an actual television viewing condition in
order to more completely capture nuanced differences between task demand in these
media forms. I

Finally, the set of studies presented here individually relied on a rather small
sampling of college-aged students. Although a priori power analyses justified the use of a
small sample in each study and the sampling frame represents a significant portion of
video game players (Jones, 2003), other populations may differ in their experience with

video games, which in turn might affect their selective exposure behaviors.

68

Future research directions

The studies presented here focus specifically on task demand as it relates to
intervention potential. Although the data presented from these studies provides empirical
support to bolster the claims of many authors who have argued that some unique quality
of video games, usually labeled interactivity, makes the technology well-adept at the
mood management process, left unanswered by this research are questions related to
other equally-important features of the theory, such as arousal regulation (the ability for a
medium to increase or decrease an individual’s felt arousal), behavioral affinity (the
similarity between message content and one’s current affective state) and hedonic
valence (the general pleasurable or unpleasurable tone of a message). The studies
presented here statistically controlled for self-reported arousal and experimentally
controlled for behavioral affinity and hedonic valence by keeping content constant across
task demand conditions. Left unaddressed by this research are questions related to how
these elements might operate differently. in video games or other interactive media as
opposed to traditional forms of media. Tangentially related to this point, the presented
studies focused on only a single type of video game (a flight simulator game) and only
two types of moods (boredom and stress). Additional research is needed to examine the
extent to which the findings here will generalize to other types of video games with
varying control schemes and levels of task demand, and whether the influence of task
demand seen here on boredom and stress will be observed also on other mood states such
as sadness (Kim & Oliver, 2007) and hostility and aggression (Bushman & Anderson,

2002; D111 & Dill, 1998; Griffiths, 1999) commonly studied in entertainment research.

69

CONCLUDING REMARKS

The increasing popularity of video games as a form of media entertainment has
pushed scholars to re-conceptualize many of our more traditional theories of media uses 1
and effects. Although many scholars have made theoretical claims about the uniqueness
of video games as compared to non-interactive media, few scholars have attempted to
both clearly define the essential qualities of video games that make them unique and to
show the relevance of this uniqueness. The set of studies presented here define one
unique quality of video games over other forms of media — namely, their enhanced task
demand — and show how this task demand can influence selective exposure and mood
management processes, two long-established theories of media uses and effects. While
the data presented here add to a growing body of research attempting to expand the
theoretical scope of selective exposure and mood management theory to include
interactive entertainment media such as video games, the study also highlights several
areas in which future study is necessary. As the video game and interactive media
industry continues to grow, so must our understanding the reciprocal processes that
govern how mood states can affect game play choices, and how game play choices can in

turn affect mood states.

70

APPENDICES

71

Appendix A. Consent Form

RESEARCH PARTICPATION CONSENT FORM

Thank you for your interest in this research study, which is designed to look at video game
preferences. If you agree to participate in this study, you will be asked to play a set of video
games and answer some questions about your opinion of them. This is a research project for the
MSU Department of Communication. Your participation in this study will help researchers get a
better understanding of how video game opinions are formed.

This study will take no more than one hour to complete. For your participation, you will be
offered course credit from your instructors. If you are not a student at Michigan State University
eligible to receive credit for this study, you will receive a $10 payment upon completion of the
study. Additionally, you will be entered into a drawing for a $100 cash prize (chances in winning
are approximately 1 in 300).

You must be 18 years or older to participate. If you do choose to participate in this study, your
confidentiality will be protected to the maximum extent allowable by law. Only researchers
involved with this study will have access to this information.

Participation in this study is voluntary, and you may quit at any time. Also, you may choose to
skip any question(s) you feel uncomfortable answering. We will not collect any personal
identification information from you, unless you choose to be in the $100 cash prize drawing. If
you choose to be entered into this drawing, we will provide a separate form that will request your
full name and MSU e-mail; we will ask you to place this form into a locked ballot box before you
leave the laboratory. The only forms that will have your identity (i.e., name) marked are this
consent form and the $100 cash prize form, and both will be stored in separate locked file cabinet.
There will be no mechanism for us to identify any of your study data as yours.

If you have concerns or questions about this study, such as scientific issues, how to do any part of
it, or to report an injury, please contact the Study Coordinator, Nick Bowman, at 562 Com Arts,
East Lansing, MI 48824, bowmann5@msu.edu, or (517) 355 - 2170. You may also contact the
Primary Investigator, Dr. Ronald Tamborini, at 570 Com Arts, East Lansing, MI 48824,
tamborin@msu.edu, or (517) 355 - 0178.

If you have any questions orconcerns about your role and rights as a research participant, or
would like to register a complaint about this study, you may contact, anonymously if you wish,
the Michigan State University Human Research Protection Programs at (517) 355-2180, Fax
(517) 432-4503, or e-mail irb@msu.edu or regular mail at 202 Olds Hall, MSU, East Lansing, MI
48824.

Your signature below indicates your voluntary agreement to participate in this study.

 

Signature Printed Name Date

72

Appendix B. Video recording consent form.

RESEARCH PARTICPATION VIDEO CONSENT FORM

Thank you for your interest in this research study, which is designed to look at video game
preferences. If you agree to participate in this study, you will be asked to play a set of video
games and answer some questions about your opinion of them. This is a research project for the
MSU Department of Communication. Your participation in this study will help researchers get a
better understanding of how video game opinions are formed.

As part of this study, we would like your permission to video record your video game playing
session. Your face will not be recorded, only your hands while playing the video game. This
information will be used to measure your interactions with the video game environment. All
recordings will be kept confidential, and will not contain any information that could identify you
as a participant in the study.

If you have concerns or questions about this study, such as scientific issues, how to do any part of
it, or to report an injury, please contact the Study Coordinator, Nick Bowman, at 562 Com Arts,
East Lansing, MI 48824, bowmannSfiiemsucdu, or (517) 355 - 2170. You may also contact the
Primary Investigator, Dr. Ronald Tamborini, at 570 Com Arts, East Lansing, MI 48824,
tamborinfijmsuedu, or (517) 355 -- 0178.

 

If you have any questions or concerns about your role and rights as a research participant, or
would like to register a complaint about this study, you may contact, anonymously if you wish,
the Michigan State University Human Research Protection Programs at (517) 355-2180, Fax
(517) 432-4503, or e-mail irb@msu.edu or regular mail at 202 Olds Hall, MSU, East Lansing, MI
48824.

Your signature below indicates your voluntary agreement to allow us to use the video recorded
footage.

 

Signature Printed Name Date

73

Appendix C. Pre-test questionnaire.

Background Information
Please answer the following questions about yourself:
What is your gender (circle one)? MALE ‘ FEMALE

What is your age (years, months)? ,

 

What is your major?

 

What is your year in school (Frosh, Soph., Junior, Senior, Grad. School)?

 

Below are statements about your ability to play video games. Please answer using the
scale provided below:

Strongly Strongly
Disagree 1 2 3 4 5 6 7 Agree

1 often win when playing videogames against other people.
I often win when playing videogames against the computer.
I am a good video game player.

I think about different video game strategies.

I can easily figure out how to play new games.

I have no problem handling the multiple buttons on currently popular game
controllers.

I can play games with complicated control systems well.
I have good video game playing skills.
1 am a better video game player than most of my friends.

I can finish video games quickly.

74

Appendix D. Study debriefing form.
RESEARCH PARTICIPANT DEBRIEFING FORM

Thanks again for participating in our experiment. This study was designed to examine how video
games can be used to alter mood states. We had you perform either a boring or stressful task for
20 minutes. After you performed this task, we had you play a video game that differed in the
amount of demand it required of the user to be played. We predicted that higher levels of task
demand would be better at repairing mood. Data from this study helps further our understanding
of how video games can affect mood.

Now that you have been debriefed as to the nature of our experiment, we ask that you please do
not discuss the details of the study until further notice. This is important because if people are
aware of the hypotheses in our study, they may accidentally or purposefully manipulate their
answers to survey questions. Be assured that all participants in this study will be debriefed in a
similar manner as you have been.

If you have concerns or questions about this study, please contact the Study Coordinator, Nick
Bowman, at 562 Com Arts, East Lansing, MI 48824, bowman115@msu.cdu, or (517) 432-3311.
You may also contact the Primary Investigator, Dr. Ronald Tamborini, at 570 Com Arts, East
Lansing, MI 48824, tamborinflmsuedu, or (517) 355-0178.

 

 

If you have any questions or concerns about your role and rights as a research participant, or
would like to register a complaint about this particular study, you may contact, anonymously if
you wish, the Director of MSU’s Human Research Protection Program, Dr. Peter Vasilenko, at
(517) 355-2180, Fax (517) 432-4503, or e-mail irb@msu.edu or regular mail at 202 Olds Hall,
MSU, East Lansing, MI 48824.

You make take this form with you for your records. Thanks again!

75

Appendix E. Questions from stress manipulation.

Intelligence Test

Below you will find sample questions taken from a variety of basic intelligence tests. The
questions represent a range of different knowledge bases. All of the questions use
information learned in a standard K-12 education, but college students tend to score
slightly higher than the general population.

Answer each question in the space provided; you may use the front and back of each
sheet of paper in your answer. For each question, be sure to show your work/proof of the
problem. Once you have moved to a new problem, please do not go back to a previous
one.

NOTE: Please provide an answer each question. If you cannot provide a complete
answer for each question, we might not be able to use your data in this study.

76

1. The Fly and the Bicycle

Mike and Sally were 10 km apart. They rode their bicycles at a constant 10 km/hr toward
each other. As they started off, a fly took off from one bicycle and flew to the other
bicycle at 20 km/hr, then it flew back to the other bicycle. The fly continued to fly back
and forth between the two bicycles until the bicycles collided, crushing the fly in
between. How far did the fly travel?

77

2. Throwing Dice

Jack is playing a dice game. He rolls five dice, and a computer calculates his score using
a particular scoring method. Jack’s last three rolls (and the scores resulting from those
rolls) are as follows:

 

 

 

 

 

Die 1 Die 2 Die 3 Die 4 Die 5 Score
5 4 6 3 5 10

l 3 3 4 2 4

1 6 2 6 4 0

 

 

 

 

 

 

What method is being used to score the dice?

78

 

3. Eight Loaves of Bread

Three travelers are sitting around a fire, and are about to eat a meal. One of them has five
small loaves of bread, and the second has three small loaves of bread. The third has no
food, but has eight coins. He offers to pay for some bread. They agree to share the eight
loaves equally among the three travelers, and the third traVeler will pay eight coins for his
share of the eight loaves. All loaves were the same size. The second traveler (who had
three loaves) suggests that he be paid three coins, and that the first traveler be paid five
coins. The first traveler says that he should get more than five coins. Is he right? How
should the money be divided up?

79

4. Magic number?

What is the smallest number that would meet all of the following criteria?

If divided by 2, the remainder is 1.
If divided by 3, the remainder is 2.
If divided by 4, the remainder is 3.
If divided by 5, the remainder is 4.
If divided by 6, the remainder is 5.
If divided by 7, the remainder is 6.
If divided by 8, the remainder is 7.
If divided by 9, the remainder is 8.
If divided by 10, the remainder is 9.

PWSP‘E‘PP’PT‘

80

5. The Missing Dollar

Three people are eating at a restaurant. The waiter gives them the bill, which totals up to
$30. The three people decide to share the expense equally, rather than figure out how
much each really owes. The waiter gives the bill and the $30 to the manager, who sees
that they have been overcharged; the real amount should be $25. He gives the waiter five
$1 bills to return to the customers, with the restaurant's apologies. But, the waiter is a
dishonest man. He puts $2 in his pocket, and returns $3 to the customers. Now, each of
the three customers has paid $9, for a total of $27. Add the $2 that the waiter has stolen,
and you get $29. But, the original bill was $30. What happened to the missing dollar?

81

6. Seeing Stars

Typical "stars" are drawn in connected, but not repeated, line segments. For example, a 5-
point star is drawn as such - line segments AC, CE, EB, BD, DA. The segments must
always alternate a constant number of points (in the above case, skipping 1 point in
between). Example:

 

Given that there is only 1 way to draw a 5-point star, and that there is no way to draw a 6-
point star (in continuous lines, that is), and there are 2 ways to draw a 7-point star, how
many different ways are there to draw a lOOO-point star?

82

7. Bowling Playoffs

Four bowling teams (named A, B, C, and D) will be participating in a round robin (every
team plays every other team once, no team is eliminated) playoff. A playoff is called a
roll-off in bowling. Each team willplay three games. For each game, two teams will bowl
each other on lanes 3 and 4, while the other two teams bowl each other on lanes 5 and 6.
One of the organizers draws up a complete schedule, showing which teams play which
teams, and who bowls on which alley:

 

 

 

 

r I Lanes 3 and 4 I Lanes 5 and 6
IGame #1 I A vs. B I c vs. D
IGame #2 I A vs. C I B vs. D
IGame #3 I A vs. D I B vs. C

 

 

The coach of one team then complains that one of the teams (team A) will get to bowl on
the same pair of lanes all three games. In general, repeatedly bowling on the same pair of
lanes should be an advantage. Using only the two pairs of lanes mentioned, can you
suggest a schedule which will have no team bowling on the same pair of lanes for all
three games? *

83

8. The New ABCs

During the first day of a college algebra class, a professor writes the following equation
on the board:

ABCD x E = DCBA
There is only one solution.

He firrther mentions that each letter represents a different digit, and none of the digits is
zero. The professor mentions that “anyone with a basic understanding of high school
mathematics should be able to solve this problem,” implying that there are no tricks to the
solution. Solve the equation.

84

9. The County Fair

Four children (including the Anders child) entered paintings in the local County Fair.
Based on the information provided below, determine who painted what, and what ribbon
each won (lst=blue, 2nd=white, 3rd=red, 4th=yellow):

1.
2.

3.

Bill's painting was not the still life.

The four paintings were (in some order) Cindy's, the Clark child's, the still life,
and the one that took 2nd place.

The Brown child's painting finished ahead of Alice's, which finished ahead of the
abstract painting.

The portrait finished ahead of Doug's painting (which was not the abstract).

Bill's painting finished just ahead of the landscape, which finished just ahead of
the Davis child's painting.

85

10. Archimedes’ Cattle

The sun god had a herd of cattle consisting of bulls and cows, one part of which was
white, a second black, a third spotted, and a fourth brown. Among the bulls, the number
of white ones was one half plus one third the number of the black greater than the brown;
the number of the black, one quarter plus one fifth the number of the spotted greater than
the brown; the number of the spotted, one sixth and one seventh the number of the white
greater than the brown. Among the cows, the number of white ones was one third plus
one quarter of the total black cattle; the number of the black, one quarter plus one fifth
the total of the spotted cattle; the number of spotted, one fifth plus one sixth the total of
the brown cattle; the number of the brown, one sixth plus one seventh the total of the
white cattle. What was the color composition of the herd?

86

11. Fries with that?

For lunch Friday, the five employees on duty at Gert's Gas & Go sent out to Big Bill's
Burgers & Fries for lunch, with each ordering a different hamburger and a different kind
of flies from the restaurant; no two spent the same amount of money on the meal. Given
the menu items and prices below and the clues that follow, can you solve this difficult
Logic Problem by finding how much each employee spent on lunch, his or her job at

Gert's Gas & Go, and the hamburger and flies he or she ordered?

The Works Burger 2.85
Carolina PitBurger 2. 75
Tucson T acoBurger 2. 45
Double CheeseBurger 2.25
Whaler F ishBurger 1. 95

1. Tom's meal cost $1.20 more than the car wash operator's lunch, which was more

expensive than Sharon's.

.U‘PP’!“

87

Big Fries
Gravy Fries
Curly Fries
Cheese Fries
Cajun Fries

Wayne spent $1.20 more on his combo than the cashier did.
The mechanic's hamburger cost 50¢ more than Russ's hamburger cost.

The Gas & Go stock clerk, who isn't Vicky, didn't have the Curly Fries.
The store manager isn't the one who ordered the Big Fries.

1.95
1.75

' 1.65

1.35
1.25

12. Time to burn

There are two lengths of rope. Each one can burn in exactly one hour. They are not
necessarily of the same length or width as each other. They also are not of uniform width
(may be wider in middle than on the end), thus burning half of the rope is not necessarily
1/2 hour. By burning the ropes, how do you measure exactly 45 minutes worth of time?

88

l3. Cereal

There is a free gift in my breakfast cereal. The manufacturers say that the gift comes in
four different colors, and encourage one to collect all four (thus eating lots of cereal).
Assuming there is an equal chance of getting any one of the colors, what is the expected
number of boxes I must consume to get all four?

89

14. The Amoebae

A jar begins with one amoeba. Every minute, every amoeba turns into 0, 1, 2, or 3
amoebae with a probability of 25% for each of the following: dies, does nothing, splits

into 2, or splits into 3. What is the probability that the entire amoeba population
eventually dies out?

90

15. Pirates

Five pirates have come across a treasure of 1000 gold coins. According to The Pirate
Code, the pirate of highest rank must make a suggestion on how to divide the money; if
the majority agrees to his suggestion, then it is to be followed by all the pirates. However,
if the suggestion does not get a majority approval then the suggesting pirate is thrown
overboard, after which time the remaining pirate of highest rank then makes a suggestion
under the same rules. This process repeats, if necessary, until only the pirate of lowest
rank is left, in which case he would get everything. Any pirate may make suggestions,
and rank does not guarantee getting more coins than anybody else.

Assume that all pirates are infinitely greedy, infinitely logical, and infinitely bloodthirsty,
and that each pirate knows this to be true of every other pirate.

The highest priority of each pirate is to get as much money for themselves as possible.
The second highest priority is to throw overboard the other pirates. A pirate will vote to
throw another one over even if they have no monetary gain by doing so, and even if it
would cost them their own life, but would not if throwing them over would cost even 1
coin. How should the first pirate suggest dividing the money to maximize his coins and
save his own life?

91

16. Bill & George

George is driving 100 ft/sec toward an intersection. He looks to his right, and sees Bill,
driving 30 ft/sec toward the same intersection. George foolishly slams on his brakes. If he
had kept going 100 ft/sec, he would have been through the intersection long before Bill
got there. At the instant that he slams on his brakes, the center of George's car is 125 ft
from the intersection, and the center of Bill's car is 150 ft from the intersection. George's
brakes give his car an acceleration of -30 ft/sec/sec. Bill never changes his speed. Each
car is 13 ft long and 7 ft wide.

Will there be a collision?

92

17. Coins

We have two identical coins. And we roll the one on the left halfway around the other
coin, so it rotates without slipping against the other coin, so that it ends up on the right of
the other coin. It has rolled over a length of only half its circumference, and yet it has
made one complete rotation. If it started right side up, then it ended right side up. Does
that make sense?

93

18. More Coins

In Rengefall, copper coins are minted with the portrait of the Queen on one side and the
portrait of the King on the other side.

One day, a half-crazed executioner gives the captured Talin one chance to avoid
execution. The executioner brings Talin into an unlit room. He tells Talin that scattered
on the table in front of him are one hundred copper coins of which twenty have the
Queen side facing up while the rest have the King side facing up. If Talin can separate the
coins into two piles, each with the same number of Queens facing up, he will release

Talin. One other constraint is that Talin must accomplish this task in 5 minutes. If Talin
fails, he will be beheaded.

It is impossible for Talin to see which side the coins are facing up in the darkness, and the
contours of the portraits are too similar to decipher by touch. Nevertheless, Talin
managed to separate the coins into two piles with the same number of Queens facing up
in the time allotted.

How did he accomplish this?

94

19. The Red Hat Club

Suppose there are 4 people, two of which are wearing red hats, and two of which are
wearing blue hats. One person is behind a wall, and the other three are standing in a
uniform line, only able to see the person directly in front of them. Which person knows
exactly which color hat he's wearing, and why? '

95

20. Power Shift

At Widgett, Inc., the pecking order of executives under CEO Norm Bates is easily
determined: each senior executive has an office on a different floor 5th-9th of company
headquarters, below the CEO's 10th floor spaces; and the closer to the 10th floor an
executive works, the more in favor he or she is. Last week, the offices of the five senior
executives, including the VP for Marketing, were swapped, signaling a power shift at the
company. Given the data below, can you determine who was on each floor before the
move and who is there now: each executive's full name (one first name is Richard, one
surname Dubois) and his or her position at Widgett, Inc.?

1.

Three of the five senior executives saw their power diminish as they moved down
in the building, one dropping three floors; the two who gained power moved up
three and two floors.

F eldman didn't occupy the 7th floor office before the power shift.

The Chief Information Officer moved up in the building to the suite vacated by
Blocker.

The VP for sales moved into the Chief Financial Officer's old spaces.

Hazlett was moved down one floor in the shakeup, with Linda taking over
Hazlett's office.

Alice got Grayson's former suite, while Grayson moved into the office vacated by
the VP for Human Resources. Both Alice and Grayson have seen their power
diminish as they are assigned to lower floors than before.

Michael left a pen set behind as a gift for Larry, who moved into Michael's old
office.

The executive who had the 9th floor office before the power shift didn't end up on
the 8th floor afterwards.

96

Appendix F. NASA-TLX (Task Load Index).
Mission Feedback

We would now like to ask you about the landing mission you just completed. Think about
the many aspects of the landing mission — including the physical and mental demands —
and answer the questions below.

Please read each of the questions below. Answer each question by marking an “X” in the
appropriate scale box.

MW: How much mental and perceptual activity was required (e.g., thinking, deciding,
calculating, remembering, looking, searching, etc)? Was the mission easy or demanding, simple or
complex, exacting or forgiving?

LowllIIJJIIJIIIIIIIIIIJIHigh

W: How much physical activity was required (e.g., pushing, pulling, turning,
controlling, activating, etc.)? Was the mission easy or demanding, slow or brisk, slack or strenuous,
restful or laborious? '

LowllIlllllllllllllIlIIIHigh

Wad: How much time pressure did you feel due to the rate or pace at which the
mission occurred? Was the pace slow and leisurely or rapid and frantic?

LowIlllllIlIlllllIllllllHigh

W: How successful do you think you were in accomplishing the goals of the mission? How
satisfied were you with your performance in accomplishing these goals?

LowLJILIIIllll'ililIIIIIIHigh

m: How hard did you have to work (mentally and physically) to accomplish your level of
performance?

LowIlIIIJLIIIllllllllllllHigh

mm: How discouraged, stressed, irritated, and annoyed versus gratified, relaxed, content,
and complacent did you feel during your mission? '

LowlllllllllLIlllIIllllIHigh

97

BIBLIOGRAPHY

98

BIBLIOGRAPHY

Atkin, C. (1985). Informational utility and selective exposure to entertainment media. In
D. Zillmann & J. Bryant (Eds), Selective exposure to communication. (pp. 63-
92). Hillsdale, NJ: LEA.

Augustine, A. A., I& Hemenover, S. H. (2008). On the relative effectiveness of affect
regulation strategies: A meta-analysis. Cognition & Emotion, 22(1), 1-40.

Boyan, A. C., & Bowman, N. D. (2007, November). A semiotic analysis of video games.
Paper presented at the Annual Meeting of the National Communication
Association, Chicago.

Bracken, C. C. & Skalski, P. (2006, August). Presence and video games: The impact of
image quality and skill level. Proceedings of the Ninth Annual International
Workshop on Presence. Cleveland, OH: Cleveland State University.

Bryant, J ., & Davies, J. (2006). Selective exposure to video games. In P. Vorderer and D.

Zillmann (Eds.) Playing video games: Motives, responses, and consequences (pp.
181-194). Hillsdale, NJ: LEA.

Bushman, B. J ., & Anderson, C. A. (2002). Violent video games and hostile expectations:
A test of the General Aggression Model. Personality and Social Psychology
Bulletin, 28(12), 1679-1686.

Bryant, J ., & Zillmann, D. (1984). Using television to alleviate boredom and stress:
Selective exposure as a function of induced excitational states. Journal of
Broadcasting, 28, 1-20. I

Chang, F. K., (2008, August 15). Russia resurgent. The New York Times. Retrieved
April 20, 2009 from: http://topics.bIogs.nytimes.com/tag/south-ossetia/page/3/

Chen, Y., & Raney, A. A. (2009, November). Mood management and highly interactive
video games: An experimental examination of Wii playing on mood change and
enjoyment. Paper presented at the annual meeting of the International
Communication Association, Marriott, Chicago.

De Petrillo, L., & Winner, E. (2005). Does art improve mood? A test of a key assumption
underlying art therapy. Art therapy: Journal of the American Art Therapy
Association, 22(4), 205-212.

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

99

Eich, E., & MaCaulay, D. (2000). Fundamental factors in mood-dependent memory. In J.

P. F orgas (Ed.), Feeling and thinking: The role of affect in social cognition. New
York: Cambridge.

Freeman, J. D., Siegrist, E. N., Wise, W. , Maxian, W. A., Brown, K. R., Nutting, B. H.
& Bradley, S. D. (2008, May). Attending to Television: Linking Cardiac
Deceleration and Controlled Attention. Paper presented at the annual meeting of
the International Communication Association, Montreal.

Gametap. (2009). Lock-on: Air combat simulator [Web site]. Retrieved April 17, 2009
from: wwwgametapcom.

 

Green, C. S., & Bavelier, D. (2003). Action video game modifies visual selective
attention. Nature, 423, 534-537.

Griffiths, M. (1999). Violent video games and aggression: A review of the literature.
Aggression and Violent Behavior, 4(2), 203-212.

Grodal, T. (2000). Video games and the pleasures of control. In D. Zillmann & P.
Vorderer (Eds), Media entertainment: The psychology of its appeal (pp. 197-
212). Mahwah, NJ: LEA.

Jones, S. (2003). Let the games begin: Gaming technology and entertainment among
college students. Pew Internet and American Life Project. Retrieved May 5, 2008
from http://wwwpewinternetorg/gdfs/PIP College_Gaming_Reportzy3df

Kahana, A., Gopher, D., Grunwald, A., Iani, C., & Lavie, P. (2004, March).
Measurement of workload by Peripheral Arterial Tone (PA T). Paper presented at
Minerva Center for Cognitive Processes and Human Performance, University of
Haifa, Israel.

Killgore, W. D. (1998). The affect grid: A moderately valid, nonspecific measure of
pleasure and arousal. Psychology Reports, 83(2), 639-642.

Kim, J ., & Oliver, M. B. (2007). Do we improve, disrupt, or embrace sadness? Exploring
sadness-based media choice and its anticipated effects on coping. Paper presented
at the annual meeting of the International Communication Association, Montreal,
Canada.

Klimmt, C., & Hartmann, T. (2006). Effectance, self-efficacy, and the motivation to play
video games. In P. Vorderer and D. Zillmann (Eds), Playing video games:
Motives, responses, and consequences (pp. 133-146). Hillsdale, NJ: LEA.

Knobloch, S., Hastall, M., Zillmann, D., & Callison, C. (2003). Imagery effects on the

selective reading of Internet newsmagazines: A cross-cultural examination.
Communication Research, 30(1), 3-29.

100

Knobloch, S., & Zillmann, D. (2002). Mood management via the digital jukebox. Journal
of Communication, 52 (2), 351-366.

Landis, J. R., & Koch, G. G. (1977). The measurement of observer agreement for
categorical data. Biometrics, 33, 159-174.

Lang, A., Bolls, P., Potter, R. F ., Kawahara, K. (1999). The effects of production pacing
and arousing content on the information processing of television messages.
Journal of Broadcasting & Electronic Media, 43(4), 451-475.

Lang, A., Bradley, S. D., Park, B., Shin, M., & Chung, Y. (2006). Parsing the resource
pie: Using STRTs to measure attention to mediated messages. Media Psychology,
8, 369-394.

Malmstron, M. A., Opton, E., & Lazarus, R. S. (1965). Heart rate measurement and the
correlation of indices of arousal. Psychosomatic Medicine, 27(6), 546-556.

Mastro, D. E., Eastin, M. S., & Tamborini, R. (2002). Internet search behaviors and mood
alterations: A selective exposure approach. Media Psychology, 4, 157-172.

Moroney, W. F ., Reising, J ., Biers, D. W., Eggemeier, F. T. (1993). The effect of
previous levels of workload on the NAA task load index (NASA-TLX) in a
simulated flight task. Proceedings of the 7m International Symposium on Aviation
Psychology. Columbus, OH: Ohio State University.

Nunes, L., Recarte, M. A. (2002). Cognitive demands of hands-free phone conversation
while driving. Transportation Research Part F: Traffic Psychology and Behavior,
5(2), 133-144.

NASA-TLX (n.d.), NASA TLX: Task Load Index. Retrieved August 1, 2009, from:
http://humansvstemsarc.nasagov/groups/tlx.

Nunes, L., Recarte, M. A. (2002). Cognitive demands of hands-free phone conversation
while driving. Transportation Research Part F: Traffic Psychology and Behavior,
5(2), 133-144.

Parkinson, B., & Totterdall, P. (1999). Classifying affect regulation strategies. Cognition
and Emotion, 13, 277-303.

Raney, A. A., Smith, J. K., & Baker, K. (2006). Adolescents and the Appeal of Video
Games. In P. Vorderer and D. Zillmann (Eds), Playing video games: Motives,
responses, and consequences (pp. 165-108). Hillsdale, NJ: LEA.

Riddle, K. E. (2004, May). Bi g-screen TV and arousal: An exploration of the effects of

television screen size. Paper presented at the annual meeting of the International
Communication Association, New Orleans.

101

Rueb, J ., Vidulich, M. A., & Hassoun, J. (1992). Establishing workload acceptability: An
evaluation of a proposed KC- 1 35 cockpit redesign. Proceedings of the Human
Factors and Ergonomics Society, 3 6” Annual Meeting. Santa Monica, CA: HFES.

Russell, J. A., Weiss, A., & Mendelsohn, G. A. (1989). Affect grid: A single-item scale
of pleasure and arousal. Journal of Personality and Social Psychology, 5 7(3),
493-502.

Strayer, D. L., & Johnston, W. A. (2001). Driven to distraction: Dual-task studies of

simulated driving and conversing on a cellular telephone. Psychological Science,
12(6), 462-466. »

Schweingruber, J. (1.999). Analysis and evaluation of cockpit concept and crew concept
of a two-seater combat aircraft. Proceedings of the Human Factors and
Ergonomics Society, 43rdAnnual Meeting. Santa Monica, CA: HF ES.

Snook-Luther, D. C. (1990). Female nudity, arousal, and ad response: An experimental
investigation. Journal of Advertising, 19(4), 51-62.

Strayer, D. L., & Johnston, W. A. (2001). Driven to distraction: Dual-task studies of

simulated driving and conversing on a cellular telephone. Psychological Science,
12(6), 462—466.

Swindells, C., MacLean, K. E., Booth, K. S., & Meitner, M. (2007). Exploring affective
' design for physical controls. Proceedings of Conference on Human Factors in
Computing Systems, CHI Letters, 9(1).

Tamborini, R. & Bowman, N. D. (2010). Presence in Video Games. In C. Bracken & P.
Skalski (Eds) Immersed in Media: T elepresence in Everyday Life (pp. 87-110).
New York: Routledge.

Tamborini, R., Eastin, M., Lachlan, K., Skalski, P., Fediuk, T., & Brady, R. (2004).
Violent Virtual Video Games and Hostile Thoughts. Journal of Broadcasting &
Electronic Media, 48, (3), 335-357.

Vorderer, P. (2000). Interactive entertainment and beyond. In D. Zillmann & P. Vorderer
(Eds), Media entertainment: The psychology of its appeal (pp. 21-36). Mahwah,
NJ: LEA.

Wolf, M. J. P., & Perron, B. (2003). The video game theory reader. New York:
Routledge.

Yates, F. (1934). Contingency table involving small numbers and the 312 test. Supplement
to the Journal of the Royal Statistical Society 1(2), 217-235.

102

Zillmann, D., & Bryant, J. (1985). Affect, mood, and emotion as determinants of
selective exposure. In D. Zillmann & J. Bryant (Eds), Selective exposure to
communication. (pp. 157-190). Hillsdale, NJ: LEA.

Zillmann, D., Knobloch, S., & Yu, H. (2001). Effects of photographs on the selective
. reading of news reports. Media Psychology, 3(4), 301-324.

103