THE EFFECTS OF MENTAL AND PHYSICAL LOAD IN EXERGAMES ON GAMEPLAY
EXPERIENCES
By
Irem Gokce Yildirim

A THESIS
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
Media and Information – Master of Arts
2017

ABSTRACT
THE EFFECTS OF MENTAL AND PHYSICAL LOAD IN EXERGAMES ON GAMEPLAY
EXPERIENCES
By
Irem Gokce Yildirim
The purpose of this study was to observe how the mental load and physical load affect players’
task load, enjoyment, flow, perceived competence, effort, performance and future play preference
in exergames. Moreover, the interaction between mental and physical load for their effects on
these variables was investigated. In a 3 (Mental Load: low-med-high) x 3 (Physical Load: lowmed-high) between subject experimental setting, participants played an exergame including a
“multiple object tracking" task. Mental load was manipulated by the number of target objects in
the gameplay screen to be tracked and physical load was manipulated by the tension level of the
exercise bike. Results showed that an increase in mental and physical load resulted in a significant
increase in perceived task load and decrease in performance while performance also mediated these
relationships. Moreover, a main effect of mental load on perceived effort was revealed. Although
there were no significant differences in motivational experiences, such as enjoyment, flow and
competence, between conditions with different levels of workloads, it was found that as mental
and physical load increases, players’ experience of enjoyment, flow and competence decreases.
As being the first experimental study applying dual-task approach into exergames using subjective
workload assessments, this study helps pave the way for future exergame research and design for
better gameplay experiences by explaining the interactions in exergames.

To my mother (Meziyet AYDIN) and father (Selamettin AYDIN)

iii

ACKNOWLEDGEMENTS

I would like to express my sincere gratitude to my advisor Taiwoo Park for his guidance, encouragement and support. I also would like to express my deepest appreciation to my professors
Constantinos K. Coursaris, Joshua Introne and our department chair Johannes M. Bauer for all their
tremendous support throughout my journey at Michigan State University. Many thanks to my thesis
committee members Wei Peng and Rabindra Ratan for their insightful comments and guidance on
this study.
Finally, my special thanks go to my family and friends for their endless love, support and
patience.

iv

TABLE OF CONTENTS

LIST OF TABLES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vii
LIST OF FIGURES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . viii
CHAPTER 1

INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1

CHAPTER 2 LITERATURE REVIEW
2.1 Mental Load in Games . . . . .
2.2 Physical Load in Exergames . .
2.3 Dual-Task in Exergames . . . . .

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

2
2
4
6

CHAPTER 3 METHODS . . . . . .
3.1 Participants . . . . . . . . .
3.2 Stimuli . . . . . . . . . . . .
3.3 Procedure . . . . . . . . . .
3.4 Measures . . . . . . . . . .
3.4.1 Manipulation Check
3.4.2 Dependent Variables
3.4.3 Mediators . . . . . .

.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.

9
9
9
11
15
15
16
17

CHAPTER 4 RESULTS . . . . . . . . . . . . . . .
4.1 Preliminary Analysis . . . . . . . . . . . . .
4.1.1 Manipulation Check . . . . . . . . .
4.2 Primary Analysis . . . . . . . . . . . . . . .
4.2.1 Main Effects on Dependent Variables
4.2.2 Mediation Analysis . . . . . . . . . .

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

18
18
18
21
21
22

CHAPTER 5 DISCUSSION AND CONCLUSION . . . . . . . .
5.1 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . .
5.2 Limitations . . . . . . . . . . . . . . . . . . . . . . . . .
5.2.1 Participants . . . . . . . . . . . . . . . . . . . . .
5.2.2 Exercising Characteristics and Cognitive Task Type
5.2.3 Exercising Device . . . . . . . . . . . . . . . . . .

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

25
25
27
27
28
28

APPENDICES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
APPENDIX A
SCREENING SURVEY . . . . . . . . . . . . . . . . .
APPENDIX B
CONSENT FORM . . . . . . . . . . . . . . . . . . .
APPENDIX C
NASA-TLX: TASK LOAD INDEX . . . . . . . . . . .
APPENDIX D
BORG RATING OF PERCEIVED EXERTION SCALE
APPENDIX E
GAME ENJOYMENT . . . . . . . . . . . . . . . . . .
APPENDIX F
GAME FLOW . . . . . . . . . . . . . . . . . . . . . .
APPENDIX G
PERCEIVED COMPETENCE . . . . . . . . . . . . .

.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.

29
30
35
39
40
41
42
43

.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.

v

APPENDIX H
APPENDIX I
APPENDIX J
APPENDIX K
APPENDIX L
APPENDIX M

PERCEIVED EFFORT . . . . . . . . . . . . . . . . . .
FUTURE PLAY PREFERENCES . . . . . . . . . . . .
PERFORMANCE . . . . . . . . . . . . . . . . . . . . .
OTHER GAMEPLAY METRICS . . . . . . . . . . . . .
CALIBRATION SESSION STATISTICS . . . . . . . .
MEAN SCORES FOR THE DEPENDENT VARIABLES

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

44
45
46
47
48
53

BIBLIOGRAPHY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59

vi

LIST OF TABLES

Table 3.1: Manipulations of Mental and Physical Load for Three Different Levels in the
Game . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
Table 4.1: One-Way ANOVA Results for Manipulation Check (with outliers) . . . . . . . . 19
Table 4.2: One-Way ANOVA Results for Manipulation Check (without outliers) . . . . . . 20
Table 4.3: Two-Way ANOVA and Two-Way ANCOVA Results Showing the Main Effects
of Mental and Physical Load on Dependent Variables . . . . . . . . . . . . . . . 22
Table 4.4: Pearson Correlations of Manipulation Check and Dependent Variables . . . . . . 23
Table 4.5: Test of Indirect Effects of Independent Variables (IVs) on Task Load (DV)
Through Performance (M) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24

vii

LIST OF FIGURES

Figure 2.1: Proposed research framework. . . . . . . . . . . . . . . . . . . . . . . . . . . .

8

Figure 3.1: Screenshot of the intro scene in the target game. . . . . . . . . . . . . . . . . .

9

Figure 3.2: Screenshot of the main gameplay scene. . . . . . . . . . . . . . . . . . . . . . 10
Figure 3.3: Screenshot of the gameplay scene in the cognitive calibration session. . . . . . . 13
Figure 3.4: Stationary exercise bike (with adjustable tension level) used in the study
connected to the computer. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
Figure 3.5: Screenshot of the gameplay scene in the physical calibration session. . . . . . . 14
Figure 4.1: Indirect effect of mental load on perceived task load through performance
(partial mediation). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
Figure 4.2: Indirect effect of physical load on perceived task load through performance
(full mediation). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
Figure L.1: Perceived mental load ratings across participants as mental load increases
during mental load calibration phase trials. . . . . . . . . . . . . . . . . . . . . 48
Figure L.2: Subjects’ last perceived mental load ratings and maximum mental load levels
identified at the end of mental load calibration phase. . . . . . . . . . . . . . . 49
Figure L.3: Subjects’ last perceived mental load ratings and maximum mental load level
statistics at the end of mental load calibration phase. . . . . . . . . . . . . . . . 50
Figure L.4: Subjects’ perceived exertion (BORG) ratings as physical load increases during
physical load calibration phase. . . . . . . . . . . . . . . . . . . . . . . . . . . 51
Figure L.5: Subjects’ last perceived exertion (BORG) ratings and maximum physical load
level statistics at the end of physical load calibration phase. . . . . . . . . . . . 51
Figure L.6: Subjects’ last perceived exertion ratings and maximum physical load level
statistics at the end of physical load calibration phase. . . . . . . . . . . . . . . 52
Figure M.1: Mean scores for NASA-TLX across mental and physical load conditions
respectively. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
Figure M.2: Mean scores for enjoyment across mental and physical load conditions respectively. 54
viii

Figure M.3: Mean scores for flow across mental and physical load conditions respectively. . 55
Figure M.4: Mean scores for competence across mental and physical load conditions
respectively. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
Figure M.5: Mean scores for effort across mental and physical load conditions respectively. . 57
Figure M.6: Mean scores for performance across mental and physical load conditions
respectively. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58

ix

CHAPTER 1
INTRODUCTION

With the increase in the level of capabilities and interactions provided by the technological advancements, people are now expected to perform complex tasks which require both mental and physical
resources. While subjective workload (or task load) assessments are generally used for mental or
physical activities separately in the domains of effective learning (Kyndt et al., 2011b; Kooiman and
Sheehan, 2014; Kyndt et al., 2011a; Pimenta et al., 2015; Rubio-Valdehita et al., 2014) and health
promotion (Jakobsen et al., 2014; Maillot et al., 2012; Sun, 2013; Staiano and Calvert, 2011), the
overall task load experience during a dual-task performance has not been focused in the previous
research (DiDomenico and Nussbaum, 2008).
The popularity of exercising devices as a solution for sedentary behavior in our daily lives has
made it essential to find engaging and effective ways of workouts (Owen et al., 2010). Exergames
have been proposed as an innovative solution to increase physical activity (Sween et al., 2014) by
adding the layer of physical demand to the design of video games in addition to the mental demand
they already require (Kalyuga and Plass, 2009). However, the composition of these demands may
turn a game into a complex activity for players. There is a lack of research on the effects of both
mental and physical demands on gameplay experiences. Since the level of these demands may
influence the overall experience and should be balanced to maximize outcomes (Kiili and Perttula,
2013, p. 140), it is very important to assess the subjective workload for mental and physical
demands to make exercising fun. To achieve this goal, the following questions are needed to be
answered for exergames. What are the contributions of mental and physical load to the overall
task load experience? How do mental and physical challenges influence enjoyment, flow, perceived
competence and effort? Which one is more effective in predicting performance? And lastly, is there
any interaction between mental and physical load for an optimum gameplay experience? Therefore,
the purpose of this study is to investigate the relationship between mental and physical load, and
their effects on taskload, enjoyment, flow, competence, effort and performance.
1

CHAPTER 2
LITERATURE REVIEW

2.1

Mental Load in Games

Cognitive Load Theory asserts that cognitive load is the mental energy that is needed to process
a given information (as cited in Ang et al., 2007). Mental load is "the aspect of cognitive load that
originates from the interaction between task and subject characteristics" (Paas et al., 2003, p. 64).
According to Paas and Van Merriënboer (1994), mental load is based on individuals’ characteristics
and knowledge about the given task. Moreover, cognitive overload may occur when the amount of
resources to process the information exceeds one’s capacity (Kalyuga et al., 1999). Since “games
usually require simultaneous performances of several cognitive and motor activities” (Kalyuga and
Plass, 2009, p. 720), people may feel overloaded with the high level of mental demand that a game
requires. For example, in previous research, the cognitive resources in games were utilized (e.g.
levels of element interactivity, redundant information) for learning and performance by preventing
cognitive overload (Annetta et al., 2009; Huang and Johnson, 2009; O’Neil et al., 2005). Therefore,
three conditions with different levels of mental load manipulations (low-medium-high) were used
in the experiment to examine their effects on the outcomes for this study. To quantify the impact of
mental load on the subjects, mental workload, which is the "cost incurred by an individual, given
their capacities, while achieving a particular level of performance on a task with specific demands"
(as cited in Hart and Staveland, 1988, p. 977), is assessed using subjective ratings (Brunken
et al., 2003). Mental demand dimension of NASA Task Load Index (NASA-TLX) is one of these
subjective measurement instruments for perceived mental load (Hart and Staveland, 1988), and it
was used for this study.
Visual elements as sensory inputs require cognitive processing. Games require visuaspatial
skills along with other cognitive executive functions such as multiple object tracking, switching
or updating skills (as cited in Nankar, 2016, p. 90). Rooted from Pylyshyn (2004)’ theory of
2

multiple object tracking, Allen et al. (2006) expands this by stating that multiple object tracking
activity causes cognitive load. Franconeri et al. (2010) showed that the number of objects, the
speed of tracking, and closeness of the objects are some determinants of the level of mental load.
Similarly, Sweller (1994) claims that the number of elements is one of the main contributing factors
for cognitive load and Allen et al. (2006) found that as the number of targets increased, subjects’
performances decreased with increased cognitive load. Moreover, processing 7±2 chunks items is
suggested as the limited capacity of working memory (Miller, 1956), identical objects are harder
to track (Botterill et al., 2011), and “working memory is the key resource in multitasking” (as cited
in Ang et al., 2007, p. 170). Therefore, in this study, multiple object tracking task with identical
features is used to manipulate mental load as used in other previous studies (Pylyshyn, 2004; Hardy
et al., 2015).
Based on the previous research, mental load manipulation has an impact on the overall task
load ratings. DiDomenico and Nussbaum (2008)’s study showed that as the mental load increased
the overall task load (measured by NASA-TLX) ratings also increased. It is reasonable that since
the NASA-TLX measurement includes "mental demand" dimension, there is a contribution by
assessment of mental load to the overall task load (H1a). Moreover, since effort consists of mental
workload (Wickens, 2017) that needs to be exerted to perform a particular task in addition to
physical workload, perceived effort is expected to increase with higher levels of perceived mental
load (H1b). Research on the direct relationship between mental load and enjoyment while playing
video games is limited. Ang et al. (2007) suggests that with a decrease in performance due to
the cognitive overload, enjoyment may be retrieved leaving frustration in its place instead (H1c).
Moreover, it was found that decreasing performance (number of failures) may increase perceived
task load (Hancock, 1989). Therefore, there might be a mediation effect of performance between
mental load and overall task load experience (H2). As flow occurs when a player’s skills are
balanced with challenges, with an increasing mental load, flow decreases (Tozman et al., 2015;
Qin et al., 2009) (H1d). Mental load is also measures of task difficulty (Xie et al., 2016) and
competence is based on the optimal level of the game’s difficulty (Constant et al., 2017); therefore,

3

higher mental load is expected to decrease competence (H1e). Considering the positive relationship
between enjoyment and future play preference (Ryan et al., 2006), it is suggested that decreasing
enjoyment with an increase in mental load will result in decrease in future play preference (H1f).
It was also found that mental load can be assessed by performance (Paas and Van Merriënboer,
1994; Paas et al., 2003; Allen et al., 2006). Due to the negative relationship between mental load
and performance, an increase in mental workload may cause failures (Paas and Van Merriënboer,
1993; Ang et al., 2007) (H1g).
Based on these findings, the following hypotheses are proposed:
H1: Higher mental load will result in higher a) perceived task load, b) effort; whereas, it will result
in lower in c) enjoyment, d) flow, e) competence, f) future game preference and g) performance in
the exergame.
H2: Performance will mediate the relationship between mental load and overall task load in the
exergame.

2.2

Physical Load in Exergames

Physical load is "factors relating to biomechanical forces generated in the body" (Wahlström,
2005, p. 168). Perceived exertion, which is the "subjective intensity of effort, strain, discomfort,
and/or fatigue that is experienced during physical exercise" (Robertson and Noble, 1997, p. 407), is
one of the measurements of physical load in addition to the objective physiological determinants (e.g.
heart rate, energy expenditure based on oxygen expenditure) (Capodaglio, 2001). As Borg (1972)
found that the subjective ratings of perceived exertion are positively correlated with physiological
responses (as cited in Hutchinson and Tenenbaum, 2006, p. 466) and perceived physical and mental
load is complementary to the physiological measurements (Borg, 1990, p. 56), BORG’s Scale of
Perceived Exertion (Borg, 1998) as a subjective assessment tool was used to measure physical load
in this study.
Exergames, which are the video games requiring physical interactivity to function, are broadly
used by people from any ages. In these games, exercising activity is added as another layer to

4

the top of mental activity that a video game already demands and a strong correlation between
exergaming and increasing energy expenditure has been revealed in the previous studies (Sween
et al., 2014). Although exergames related studies are mainly focused on health and education
domains, the entertaining feature of these games is still prominent and crucial to promote physical
activities for people living in a sedentary lifestyle (Lai et al., 2012).
Although there is a little research in the literature about the relationship between players’
interaction with exergames and their experience of positive moods, some of the previous studies
showed that as physical workload increases, overall perceived taskload increases (DiDomenico
and Nussbaum, 2008) (H3a), and perceived effort increases since "workload is an indicator of the
mental and/or physical effort required to carry out one or more tasks at a specific performance level"
(as cited in Arroyo-Gómez et al., 2017) (H3b). Although there is no study which empirically tested
the effects of different levels of physical loads on enjoyment in an exergame, based on the balanced
level of challenges for an optimum experience approach in Flow theory (Sinclair et al., 2009; Qin
et al., 2009; Lai et al., 2012), high physical load is predicted to result in less enjoyment and flow
(H3c, H3d). In their study, Park et al. (2014) suggest that higher level of physical workload will
deteriorate perceived competence and performance (H3e, H3g). Due to the positive correlation
between enjoyment and future play preference (Ryan et al., 2006), it is predicted that an increase
in physical load will reduce the future play preference for the game (H3f).
Based on the evidences and interpretations above, we propose the following hypotheses:
H3: Higher physical load will result in higher a) perceived task load, b) effort; whereas, it will result
in lower in c) enjoyment, d) flow, e) competence, f) future game preference and g) performance in
the exergame.
H4: Performance will mediate the relationship between physical load and overall task load in the
exergame.
Park et al. (2013) found that lower perceived taskload has a positive effect on perceived competence and involvement. Since we used immersion subscale from the GameFlow measure, all the
items in our survey were related to involvement. Therefore, we argue that task load will also me-

5

diate the relationship between perceived mental and physical load and their effects on competence,
performance and flow experiences in our study (H5). Moreover, since "when in flow, a player does
not have to invest effort to keep his mind on the task" (Kiili et al., 2013, p. 202), an argument can
be made that flow will mediate the relationship between perceived effort and mental load as well
as physical load(H6).
H5: Perceived taskload will mediate the relationship between competence and mental and physical
load, respectively, as well as the relationship between flow and mental and physical load, respectively, in the exergame.
H6: Flow will mediate the relationship between mental and physical load separately with perceived
effort in the exergame.

2.3

Dual-Task in Exergames

According to dual-task methodology, cognitive resources are split for concurrent tasks and
attention switching occurs Damos (1991). For example, when a cognitive performance is accompanied by physical exercising, it was found that dual-task effect was significantly related to energy
demand by the task, and more attentional resources is necessary to control movements (Brisswalter et al., 2002, p. 555). Similarly, research on executive functions showed that cognitive skills
required to plan, monitor and execute actions are associated with physical performance (as cited
in Coppin et al., 2006). Although in some health promotion related studies dual-task interventions
have been used to ameliorate age related physical and cognitive declines (Nankar, 2016), they may
also result in cognitive overload as Strobach et al. (2012) stated that dual tasks require coordination
of actions and shifting between these actions. Shifting is one of executive functions, which refers
to “switching between attentional sets or task sets” (Van Muijden et al., 2012), and it may affect
dual-task performance (Miyake et al., 2000) and increase cognitive load (Ang et al., 2007).
In addition to the physical activity, exergames require cognitive functioning such as sensing
stimuli and paying attention (Larsen et al., 2013, p. 6). The mental and physical demand required
in exergames "make players experience divided attention and impose continuous demands on

6

the attention, potentially resulting in dual-task interference" (Park et al., 2014, p. 419). The
integration of physical and cognitive demands into exergames requires specific attention to their
implementations (Lyons, 2015). How some game mechanics (feedback, challenges and rewards)
are operationalized can affect physical and psychological reactions to exergames. The "interplay
between cognitive load and physical load" (Pisan et al., 2013) is necessary for expected outcomes.
Although the dynamics in the intervention of mental effort to physical exertion have been studied
in many previous research, there is no exergame focused study observing the interactions between
these in a concurrent setting using dual-task(multi-task) approach for gameplay experience related
outcomes. For example, the effect of physical exercise on cognitive (or mental) performance was
observed during or after (not a dual-task design) in many empirical studies and some contradictory
results were revealed (Brisswalter et al., 2002). While in some studies it was found that physical
exercise improved cognitive performance, some other showed that physical exertion deteriorated
the performance in cognitive tasks (as cited in Krausman et al., 2002, p. 4). In another study,
Davey (1973) found an inverted U relationship between the level of physical exertion and after
mental performance of short term memory. Therefore, in this study, mental and physical load
are manipulated in three different levels for a possible curvilinear relationship with gameplay
experiences. These results led researchers to highlight the importance of the type of the cognitive
task (e.g. visual perception, arithmetic, decision making), the duration and intensity of the physical
exertion on the performance results (Brisswalter et al., 2002; Krausman et al., 2002) and timing
of physical exertion and cognitive task (concurrent/dual-task or not) Brisswalter et al. (2002).
Moreover, Kiili and Perttula (2013) proposed a framework to balance these two loads for more
engaging exertion and learning experiences.
Based on dual-task methodology1, to contribute to the literature in the field of exergames,
in this study the interaction between mental and physical load is observed using the subjective
assessment tools during. There are only a few exergame related studies for observing gameplay
1 Although dual-task approach was included in this thesis to explain the dynamics in concurrent activities, the
exergame that was used in this study might not necessarily be defined as a dual-task but rather one multifaceted task
which requires both physical and cognitive resources.

7

experiences using many different cognitive load manipulation tasks (i.e. choice reaction test, shortterm memory/retention test and arithmetic test) and physical load manipulation methods (e.g. target
speed) (as cited in Park et al., 2014). In their studies, Hardy et al. (2015) and Park et al. (2014)
utilized tension level to manipulate physical load and in this study we used the same method to
change physical load for three different levels in three conditions. Considering the contribution of
multiple-object tracking task to the perceived mental load as explained in the previous section, it is
used to manipulate mental load in three conditions with different levels.
Lastly, the following research question is proposed for this study to explore the question of
dual-task interactions in exergames:
RQ1: Will there be any interaction effect of mental and physical load on a) overall task load, b)
enjoyment, c) flow, d) competence, e) effort, f) future game preference and g) performance in the
exergame?
In this research, the goal is to explore the effects of different levels of mental and physical
load on perceived task load, flow, enjoyment, effort, competence, preference for future play and
performance, hoping to find an optimum levels of mental and physical load for an optimal gameplay
experience. This study may help researchers better understand how different levels of physical and
mental load in exergames affect users’ interactions. It will also help pave the way for future exergame
research around the concepts of personalized and adaptive exercising experiences (Figure 2.1).

Figure 2.1: Proposed research framework.

8

CHAPTER 3
METHODS

3.1

Participants

Undergraduate and graduate students from a large Midwest university in the United States were
asked to participate in the study. As the game used in the experiment requires physical activity, a
screening criteria of physical activity readiness was applied. In total, 114 participants (46 male; 68
female) were eligible to participate in the study. The average age of the participants was 21.

3.2

Stimuli

The target game for this study was an exergame entitled CogXer. A 27in iMac is used as
computer screen for the game and an FitDesk Pedal Desk exercise bike was connected to the
iMac through an Arduino sensor tracking the speed information. Audio was delivered via builtin speakers on the iMac (Figure 3.1). This exergame was designed following the feedback and
challenge design suggestions posit by Lyons (2015).

Figure 3.1: Screenshot of the intro scene in the target game.
The game was designed to provide two main challenges; mental and physical load. Multiple
object tracking task was applied in the game to promote cognitive challenge whereas the player
controls the avatar in the game by cycling as the physical challenge. This target game was tested
9

with one year of playtesting by undergraduate and graduate students from the target audience and
the design was improved based on the feedback collected.
In the game, multiple object tracking task was simulated as an implementation of visual tracking,
searching and recalling actions. In a zoo like environment, turtles are used as objects to track and
the goal in the game is feeding as many turtles as possible. Player has baits the number of which
is equal to the number of all turtles in the screen initially. However, the mental challenge about
feeding is that since all the turtles are identical and wandering around, player needs to remember and
visually track which turtle she or he already fed to reserve the rest of baits for the unfed ones. This
mental challenge was manipulated in the game by changing the "number" of target objects in the
game (i.e. decreasing or increasing the number of turtles). To support the feeling of competence,
timer mechanics is also implemented. For each turtle, the participant is given 10 seconds to feed
it (e.g. if there is 5 turtles in a round, player has 50 seconds to feed all the turtles). In the game,
player controls an avatar of an unicycle. The speed information was streamed from an exerbike
through an Arduino based connection to the game by mapping the speed data into the position of
the unicycle. The physical challenge about controlling of the unicycle is that based on the tension
level set on the exerbike, the difficulty level of cycling can be manipulated which in turns makes it
harder for player to control the avatar in the game (Figure 3.2).

Figure 3.2: Screenshot of the main gameplay scene.
10

3.3

Procedure

Before coming to the lab, participants completed a screening survey about their eligibility to
perform a physical activity and interest to play a cognitive exergame. Participants who indicated
that they had previous injury records or reasons not to do exercise using the Physical Activity
Readiness Questionnaire (PAR-Q) in the screening survey were not allowed to participate in the
study. Eligible participants scheduled their visit to the lab using the college’s participation pool
system.
Participants were randomly assigned to one of the nine conditions in this study: a 3 (mental load:
low-medium-high) x 3 (physical load: low-medium-high) between subjects design. Upon arrival to
the lab, participants were introduced to the procedure using a well-documented protocol script and
signed the consent form. There were three main sessions in the experiment. The first two session
were the ‘calibration’ sessions to identify participants’ maximum physical and mental performance
levels to gauge their baseline to set the difficulty level variables for the actual gameplay (Table
3.1). Calibration was needed because each individual may have different levels of endurance for
the mental and physical load presented in a given task. Since low-med-high values can be different
for each participant in this type of mental and physical challenge setting, instead of using a fixed
number of target game objects and tension level on bike for each condition, rather, we identified the
max values in calibration sessions and setting the number of objects and tension level for the med
and high conditions based on that max value to eliminate the individual differences in their respose
to different mental and physical demands (DiDomenico and Nussbaum, 2008, p. 978). Following
Ryu and Myung (2005)’s suggestion of "in a multitask condition, the task should be broken up
into its constituent elements and the demands of each element must be evaluated independently",
in these calibration sessions, a slightly different versions of the same target game, CogXer, were
used; one without physical challenge, one without mental challenge. The last session was the actual
game play which lasted around 5 minutes for each player (participants were let to stop playing if
there were tired or exhausted) and it was followed by an online post-test questionnaire.

11

Table 3.1: Manipulations of Mental and Physical Load for Three Different Levels in the Game
Levels
Low
Mental Load
Medium
High
Low
Physical Load Medium
High
IVs

Manipulation
Manipulation check item
# of turtles = Min (1)
# of turtles = Mean (1, max)
NASA-TLX1: How mentally demanding was the task?
# of turtles = Max (based on the mental calibration)
Tension level on bike = Min (1)
Tension level on bike = Mean (1, max)
BORG: Choose the number that best describes your level of exertion.
Tension level on bike = Max (based on the physical calibration)

• In the mental challenge calibration phase, participants played the game using keyboard
(without physical load - no exercise device), while gradually increasing the mental difficulty
of game by increasing the number of objects to track on the screen by 2 and giving 10
seconds for each object after each successful round (starting from 2) over time (Figure 3.3).
We provided three ’lives’ and three replay chance for every round (round refers to gameplay
with the same number of turtles), and reduced one for every ’UNSUCCESSFUL’ feeding,
including: double-feeding and mis-feeding. Therefore, successful round referred to finishing
a round with at least one life left (in other words, if there is N turtle in that round, finishing it
with at least N-2 successfully fed turtles is successful). Between each increment interval (no
rating between consecutive replay trials), participants were asked to rate perceived mental
demand to measure mental load (Kalyuga & Plass, 2009; Schrader; as cited in Schrader
& Bastiaens, 2012) using the 7-point Likert question from NASA-TLX popping up on the
window.
Maximum mental load for a participant was decided and so the mental calibration stops when
one of following conditions happens:
– If a player cleared the round but rate the round as ’7’. (SUBJECTIVE)
– Participant pressed Give Up button on the screen when s(he) felt it’s maximum limit
for him or her (after they press give up button, they were asked to rate for the last time.
calibration stops no matter they rate it with less than ’7’).(SUBJECTIVE)
– If a player failed three replay chance of the round, cognitive calibration stopped there
(no matter they rate it with less than ’7’). (OBJECTIVE, performance can be used as an
indicator of perceived mental load (Paas and Van Merriënboer, 1994; Qin et al., 2009))
12

When calibration stopped, the number of turtles in the last played round was noted as
‘max’ mental load for this participant. In this way, mental load measurement was mixed
with subjective rating and objective performance (Brunken, Plass, & Leutner, 2003) during
manipulation in this calibration phase.

Figure 3.3: Screenshot of the gameplay scene in the cognitive calibration session.
• In the physical challenge calibration phase, participants played the same game (without
mental load - no object tracking task), but they were asked to just cycle at certain speed
(e.g., 100 RPM). The tension level of the exerbike was gradually increased by one level in
every 30 seconds (levels are between 1-to-8) (Figure 3.4). Between each increment interval,
participants rated perceived physical demand using the BORG Rating of Perceived Exertion
(scores were between 6-20). Physical calibration was stopped when participants gave up and
identified a tension level as their "max"1. This level was noted as ‘max’ physical load for this
participant. Then, participants were asked to rest for 5 minutes before the actual gameplay
session (Figure 3.5).
1 During the data collection process in Spring 2017 term, if the participant rated 16 or more in the manipulation
check item (BORG scale), we stopped the calibration and took the tension level that he or she rated 16 or less as the
max value for that participant. Based on the pilot test results, to increase the difference in manipulation check item
between low and medium physical load conditions, we decided to let the participant go as far as they can for the further
data collection process which includes Fall 2017 term. Therefore, ’data set’ is used as a covariate in the further analysis
in this thesis.

13

Figure 3.4: Stationary exercise bike (with adjustable tension level) used in the study connected to
the computer.

Figure 3.5: Screenshot of the gameplay scene in the physical calibration session.
After the calibration sessions (see Appendix L), the actual gameplay session was conducted.
Based on the assigned condition and the maximum values marked during the calibration sessions,
mental load was manipulated by the number of target game object in the game to be tracked, and
physical load was manipulated by the resistance/tension of the exercise bike. Physical and mental
load for low-med-high conditions were set based on the min-mean-max values. "Min" values for

14

low conditions were always fixed as ’1’ for cognitive load and ’1’ for physical load. "Mean" and
"Max" are set based on the maximum value identified in the calibration phases (Table 3.1):
• If the participant is in high condition, we set the number of turtles and tension level variables
to the max value revealed in the calibration sessions,
• If the participant is in med condition, we set the number of turtles and tension level variables
to the mean of max value revealed in the calibration sessions and fixed min values,
• If the participant is in low condition, we set the turtles and tension level variables to fixed
min values which is ’1’ as number of target objects for mental load2 and ’1’ as tension level
for physical load.
In this way, participants in med and high conditions were assigned values based on their personal
med and max values that they could endure.
After setting the manipulation variables for the assigned condition, participant played the game
with the same number of turtles and same tension level over a series of rounds for about 5 minutes
and they were let to stop playing anytime due to fatigue. During this session, gameplay data was
recorded in a log file. After playing, participants completed a post survey including questions
about their gameplay experiences. The entire procedure took around 45 to 60 minutes to complete.
Participants received partial course credits for their participation.

3.4
3.4.1

Measures
Manipulation Check

Since the Mental Demand dimension from The National Aeronautical and Space Administration
Task Load Index (NASA-TLX) and BORG Perceived Exertion Scale scale were used during the
2 During the data collection process in Spring 2017 term, we used ’2’ as the number of turtles for low mental load
conditions. Based on the pilot test results, to increase the difference in manipulation check item between low and
medium mental load conditions, we decided to set it to ’1’ for the further data collection process which includes Fall
2017 term. Therefore, ’data set’ is used as a covariate in the further analysis in this thesis.

15

calibration phases to determine the maximum mental and physical load respectively for each
participant, they were used as manipulation check items in the post-survey.
Mental Demand dimension item from NASA-TLX ("How mentally demanding was the task?")
was used as an indicator whether the mental load manipulation was successful (see Appendix C).
It is an 7-point scale anchored by 1 (very low) and 7 (very high). In the mental load calibration
session, the question of "How much difficult was remembering, tracking and searching?" was used
to elaborate this item for participants.
BORG scale (Borg, 1998) was used as an indicator whether the physical load manipulation
was successful (see Appendix D). It is an numerical scale ranging from 6 to 20 with several short
descriptions provided to represent the level of exertion (e.g. "No exertion at all", "Extremely light",
"Very light", "Somewhat hard") (see Appendix D).

3.4.2

Dependent Variables

Overall task load was measured by 7-point scale NASA-TLX (see Appendix C). This scale has six
dimensions-mental demand, physical demand, temporal demand, performance, effort and frustration (Hart and Staveland, 1988). Although physical demand dimension is included in this measure,
this dimension is not intended to be used to assess perceived physical load. Cronbach’s Alpha of
the scale was above .7 and reliability was secured (α = .71).
Game enjoyment was measured using the scale developed by Song et al. (2011) (see Appendix
E). It is an 7-point scale on some adjectives to describe the game including "boring"(R), "exciting",
"enjoyable", "entertaining", "fun", "interesting", "pleasant" (α = .90).
To measure flow, Immersion dimension and two items from Challenge dimension of game flow
scale developed by Fu et al. (2009) based on Sweetser and Wyeth (2005)’s framework were used
(see Appendix F). Since the immersion dimension has the most factor loading, all the items from
this dimension were included in the scale. Sample items includes "I forget about time passing
while playing the game","I can become involved in the game" . Moreover, as the balance between
skills and challenges is the key to experience flow (Csikszentmihalyi, 1990) and players feel bored
16

or frustrated when they are out of the flow zone (Sweetser and Wyeth, 2005), two items from
Challenge dimension "I enjoy the game without feeling bored or anxious" and "The challenge is
adequate, neither too difficult nor too easy" were also included to measure flow in this study (α =
.84).
Competence and effort were measured using the Perceived Competence and Effort measures
(McAuley et al., 1989) from Intrinsic Motivation Inventory (IMI) (α = .88 and α = .77, respectively)
(see Appendix G and Appendix H). Although effort dimension is included in NASA-TLX measure,
that dimension is not intended to be used to assess perceived effort.
Future play preference measure included two items from Ryan et al. (2006) and additional
two items including "Given the chance I would play this game in my free time" and "I would be
interested in having my own personal copy of this game" (α = .97) (see Appendix I).

3.4.3

Mediators

Performance was proposed as potential mediator in this study. Using the gameplay log file recorded
during actual gameplay session, performance was measured as the success in the visual tracking
task which is successful feeding of the turtles in the game (see Appendix J). Since successful
feeding requires a certain amount of physical effort (to control the position of the avatar in the
game), we can assert that the performance measure is a representation of a mixture of mental and
physical effort exerted by the player.
In addition, total game play time (active and resting), average time spent per round, average
performance per round, average speed, average speed per round and minimum speed violation per
round were also recorded in that log file (see Appendix K).

Gender, weekly gaming hours and weekly exercising hours were reported by the participants in
the prescreening survey considering these as potential covariates (see Appendix A).

17

CHAPTER 4
RESULTS

4.1

Preliminary Analysis

We examined the relationship between gender, weekly gaming hours and weekly exercising
hours with dependent variables to control in the subsequent analysis as potential covariates for
homogeneity of regression. Gender was found to have main effects on only perceived exertion,
competence and performance. Males perceived less exertion [F(1,112) = 5.21, p < .05], perceived
more competence [F(1,112) = 7.22, p < .01] and performed better [F(1,106) = 4.54, p < .05] than
did females. A Pearson correlation coefficient was computed to assess the relationship between
weekly gaming hours and weekly exercising hours and dependent variables. There was significant
negative correlation between weekly gaming hours and perceived exertion [r = -0.331, n = 106,
p = 0.001] and it had a main effect on perceived exertion [F(15, 90) = 1.96, p < .05]. Moreover,
a significant positive correlation was found between weekly gaming hours and competence [r =
0.227, n = 106, p = 0.019]. Weekly exercising hours variable was only negatively correlated with
performance [r = -0.218, n = 106, p = 0.029].

In the subsequent analysis, gender, weekly gaming hours, weekly exercising hours and data set
(Spring 2017 vs Fall 2017, see Section 3.3) variables were controlled. All the data were examined
for homogeneity of variance using Levene Tests.

4.1.1

Manipulation Check

Two sets of one-way analysis of variance (ANOVA) tests were conducted to check the manipulation
of the independent variables, mental load and physical load.
Shapiro-Wilk values as well as were obtained to observe the distributions within nine conditions for manipulation check items (NASA-TLX1:Perceived Mental Demand for mental load
18

manipulation and BORG: Perceived Exertion for physical load manipulation). The distributions for
perceived mental demand in the data of mental load low - physical load medium and mental load
medium - physical load high conditions failed the assumption of normality. Similarly, the data in
mental load high - physical load medium and mental load high - physical load high conditions failed
the assumption of normality for perceived exertion. After applying log10 transformation for the
corresponding manipulation item data, normality still failed in these conditions. NASA-TLX1 and
BORG contained six outliers in total. To determine whether the outliers influenced manipulation
results, the inferential statistical tests (ANOVA) were conducted twice for independent variables
respectively, with outliers and without outliers.
With outliers, there was a significant effect of mental load on perceived mental demand for the
three conditions [F (2, 111) = 10.39, p < 0.001]. A Bonferroni post hoc test revealed that compared to the high mental load condition (M = 4.98, SD = 1.52) the perceived mental demand was
significantly lower in low (M = 3.41, SD = 1.59, p < 0.001) and medium mental load conditions
(M = 4.11, SD = 1.49, p = 0.034). However, the mean scores were not significantly different between low and medium mental load conditions (p = 0.158). One-way ANOVA results for physical
load showed that the means of the levels of the perceived exertion significantly different between
three physical load conditions [F (2, 111) = 37.64, p < 0.001] (Low: M = 12.66, SD = 2.35;
Medium: M = 14.72, SD = 2.25; High: M = 16.95, SD = 2.87) suggesting that the manipulation
was successful (Table 4.1).

Table 4.1: One-Way ANOVA Results for Manipulation Check (with outliers)
Mental Load
NASA-TLX1 (Perceived Mental Demand)
- How mentally demanding was the task?
(i.e. How much difficult was remembering and tracking?)
Physical Load
BORG (Perceived Exertion)
- Please choose the number from below
that best describes your level of exertion.

Low (n = 37) Medium (n = 35) High (n = 42)
F (2, 111)
M (SD)
M (SD)
M (SD)
3.41 (1.59)

4.11 (1.49)

4.98 (1.52)

10.39**

Low (n = 38) Medium (n = 39) High (n = 37)
F (2, 111)
M (SD)
M (SD)
M (SD)
12.66 (2.35)

N ote. ∗ p < .05, ∗∗ p < .01, ∗∗∗ p < .001.

19

14.72 (2.25)

16.95 (2.87)

37.64**

η2
0.16

η2
0.37

When we excluded the outliers from the analysis, there was a still significant effect of mental load on perceived mental demand for the three conditions [F (2, 105) = 10.72, p < 0.001].
However, Bonferroni post hoc test revealed that there was only a significant difference in perceived mental demand between low and high mental load conditions. Although, the difference
approached significance between low-medium (p = 0.099) with the exclusion of outliers, it was
non-significant between medium-high mental load conditions (p = 0.051). Without outliers,
one-way ANOVA results for perceived exertion were nearly identical between three physical load
conditions [F (2, 105) = 28.43, p < 0.001]. Pairwise Bonferroni post hoc comparisons still revealed significant differences for all three physical load conditions (Table 4.2).

Table 4.2: One-Way ANOVA Results for Manipulation Check (without outliers)
Mental Load
NASA-TLX1 (Perceived Mental Demand)
- How mentally demanding was the task?
(i.e. How much difficult was remembering and tracking?)
Physical Load
BORG (Perceived Exertion)
- Please choose the number from below
that best describes your level of exertion.

Low (n = 33) Medium (n = 35) High (n = 40)
F (2, 105)
M (SD)
M (SD)
M (SD)
3.33 (1.41)

4.11 (1.49)

4.95 (1.62)

10.72**

Low (n = 35) Medium (n = 39) High (n = 34)
F (2, 105)
M (SD)
M (SD)
M (SD)
12.86 (2.23)

14.72 (2.25)

17.00 (2.39)

28.43**

η2
0.17

η2
0.35

N ote. ∗ p < .05, ∗∗ p < .01, ∗∗∗ p < .001.

Both one-way ANOVA analyses with and without outliers concluded there was no significant
difference between average perceived mental demand scores for participants who were in low mental
load condition and participants who were in medium mental load condition. Therefore, we decided
to collapse low and medium mental load conditions into one "low" condition without removing the
outliers (M = 3.75, SD = 1.57) and keep the high condition as it was for the subsequent analysis
and applied t-test analysis on this new low and high conditions; t(112) = 4.06, p < 0.001. As a
result, the design changed to 2 (mental load: low-high) x 3 (physical load: low-medium-high) from
3 x 3.

20

4.2
4.2.1

Primary Analysis
Main Effects on Dependent Variables

Three sets of two-way analysis of covariance (ANCOVA) were conducted to examine the effects
of mental load and physical load on competence (controlling only for data set, after dropping other
covariates due to their non-significant effects), effort (controlling only for data set, after dropping
other covariates due to their non-significant effects) and performance (controlling only for weekly
exercising hours, after dropping other covariates due to their non-significant effects). Considering
the non-significance in their effects on the other dependent variables, all the covariates were dropped
and a series of two-way analysis of variance (ANOVA) were conducted to examine the main effects
of mental load (low-high) and physical load (low-med-high) on the other dependent variables and
to test whether there were interactions.
Both mental load and physical load had statistically significant main effects on NASA-TLX
task load ratings. As the mental load increases across low-high conditions, perceived task load
increased, F (1, 108) = 12.99, p < .001. Similarly, higher physical load resulted in higher perceived
task load F (2, 108) = 5.34, p < .01. The main effect of mental load on effort was also statistically
significant F (1, 107) = 4.24, p < .05. Participants in high mental load condition reported higher
effort than those in low mental load condition. Lastly, due to the unequal number of subjects
across conditions and the violation of Levene’s Test of homogeneity of variances revealed in the
two-way ANCOVA analysis for performance dependent variable within each of the mental and
physical load conditions, two sets of one-way between subjects ANOVA were conducted for each
of the independent variables using Welch’s test (with Games-Howell post-hoc comparison for three
physical load conditions). Results showed that subjects in high mental load condition performed
significantly better F (1, 106) = 20.48, p < .001 and subjects low physical load condition performed
significantly better than those in high physical load condition F (2, 105) = 8.12, p < .01. However,
no significance was found in the performance difference between low-medium and medium-high
conditions (Table 4.3). Although the means are supporting the arguments in H1 and H3, only H1a,

21

Table 4.3: Two-Way ANOVA and Two-Way ANCOVA Results Showing the Main Effects of Mental
and Physical Load on Dependent Variables
Low (n = 72)
M (SD)
4.23 (0.96)
4.81 (1.02)
4.64 (1.07)
3.99 (1.25)
4.84 (0.96)
3.49 (1.77)
0.59 (0.20)

Dependent Variables
NASA-TLXtwo−way−anova
Enjoymenttwo−way−anova
Flowtwo−way−anova
Competencetwo−way−ancova
Efforttwo−way−ancova
Future Play Preferencetwo−way−anova

Performanceone−way−anova b1

Dependent Variables
NASA-TLXtwo−way−anova
Enjoymenttwo−way−anova
Flowtwo−way−anova
Competencetwo−way−ancova
Efforttwo−way−ancova
Future Play Preferencetwo−way−anova
Performanceone−way−anova b2
N ote.∗ p < .05, ∗∗ p < .01.
two−way−anova Two-way between subjects ANOVA analysis was conducted.
two−way−ancova Two-way between subjects ANCOVA analysis was conducted.
one−way−anova One-way between subjects ANOVA analysis was conducted.
b1 Low(n = 67), High(n = 41), F (1, 106), 6 data were missing due to a technical error of gameplay logging.
b2 Low(n = 35), Medium(n = 39), High(n = 34), F (2, 105), 6 data were missing due to a technical error of gameplay logging.

Mental Load

High (n = 42)
M (SD)
4.87 (0.90)
4.74 (1.12)
4.55 (0.95)
3.46 (1.19)
5.26 (0.81)
3.40 (1.73)
0.43 (0.11)

Physical Load
Low (n = 38) Medium (n = 39) High (n = 37)
M (SD)
M (SD)
M (SD)
4.15 (0.95)
4.46 (0.85)
4.81 (1.04)
4.97 (1.05)
4.79 (0.99)
4.60 (1.11)
4.89 (0.68)
4.66 (1.04)
4.26 (1.21)
3.97 (1.28)
3.82 (1.22)
3.59 (1.26)
4.87 (0.99)
5.02 (0.93)
5.10 (0.87)
3.59 (1.64)
3.60 (1.82)
3.18 (1.78)
0.58 (0.14)
0.58 (0.15)
0.43 (0.24)

F (1, 108)
12.99∗∗

0.05
0.129
2.90
4.24∗
0.09
20.48∗∗

F (2, 108)
5.34∗∗
0.88
2.07
1.25
0.91
0.39
8.12∗∗

η 2p

p

0.11 0.000
0.000 0.82
0.001 0.72
0.03 0.09
0.05 0.042
0.001 0.76
0.16 0.000
η 2p

p

0.09
0.01
0.02
0.42
0.04
0.13
0.003 0.30
0.01 0.41
0.01
0.68
0.15 0.001

H1b, H1g and H3a and H3g are supported with a statistical significance.
No interaction effects were found (RQ1 is answered).

4.2.2

Mediation Analysis

Before testing the mediation, a Pearson correlation was run to gauge the relationship between
manipulation check variables and other dependent variables (Table 4.4). There was a statistically
significant, negative correlation between performance and perceived task load (r = -.37, n = 108,
p < .001), whereas there was a positive correlation between performance and competence (r = .29,
n = 108, p = .002). Moreover, perceived task load was positively correlated with effort (r = .40, n
= 114, p < .001) while was negatively correlated with competence (r = .-29, n = 114, p = .002).
Interestingly, it was revealed that there was a significant positive correlation between perceived
mental load and enjoyment and flow (r = .19, n = 114, p = .048 and r = .20, n = 114, p = .031,
respectively).
Multiple mediation tests were conducted after taking the correlations and main effects in the
previous analysis into account to examine whether the proposed mediators had effects on the other
dependent variables. Since there were no main effects of mental and physical load on competence
and flow, we did not conduct further mediation analysis for these variables (H5 and H6 are not

22

Table 4.4: Pearson Correlations of Manipulation Check and Dependent Variables
Perceived
Perceived Exertion NASA-TLX Enjoyment
Mental Demand
M = 4.20
M = 14.75
M = 4.47
M = 4.79
SD = 1.66
SD = 2.87
SD = 0.98 SD = 1.05
Perceived Mental Demand
Perceived Exertion
.25**
NASA-TLX
.62**
.62**
Enjoyment
.19*
.03
.15
Flow
.20*
-.02
.15
.55**
Competence
-.01
-.32**
-.29**
.31**
Effort
.20*
.34**
.40**
.48**
Future Play Preference
.03
-.02
.05
.52**
Performancea
-.29**
-.40**
-.37**
.08
N ote. n = 114.
∗ p < .05, ∗∗ p < .01.
a n = 108, 6 data were missing due to a technical error of gameplay logging.

Flow

Competence

M = 4.61
SD = 1.02

M = 3.80
SD = 1.25

.30**
.37**
.47**
.19

-.05
.16
.29**

Future Play
Performance
Preference
M = 5.00
M = 3.46
M = 0.53
SD = 0.93 SD = 1.75
SD = .19
Effort

.27**
-.05

.05

-

supported).
An SPSS macro entitled PROCESS developed by Hayes (2012)1 was run and bootstrapping
method was used to determine the significance of indirect effects (resample was set to 1,000). Three
conditions should be established to determine whether mediation has occurred. First criteria is that
the total effect of IV on DV (path c) should be statistically significant. Second, IV should predict
the mediator M (path a). Lastly, the indirect effect of IV and M together should be statistically
significant (confirming M significantly predicts DV (path b) and a drop in the significance (nonsignificance for full mediation) of direct effect (path c’)). For example, in the first analysis of
performance as a mediator between mental load and perceived task load by controlling gender; the
total effect of mental load on task load was significant .71 (SE = .18), p < 0.001 and when the
mediator included in the model, the significance direct effect reduced to .49 (SE = .20), p = 0.015
(partial mediation). The biased corrected confidence interval of the indirect effect (.009 and .50),
which is the difference between total and direct effect, did not contain zero which showed that the
indirect effect was significant. The same procedure was used to test the other mediation hypothesis
about physical load. Results showed that performance mediated the relationship between perceived
mental and physical load and task load, therefore H2 and H4 were supported (Table 4.5).

1 The PROCESS macro for SPSS: http://www.processmacro.org/

23

Table 4.5: Test of Indirect Effects of Independent Variables (IVs) on Task Load (DV) Through
Performance (M)
IV (Mental Load), M (Performance), DV (Perceived Taskload) p
IV to M (path a)
M to DV (path b)
Total effect of IV on DV (path c)
Direct effect of IV on DV (path c’)
Indirect effects of IV on DV through M (path a*b)
Biased corrected 95% CI: .0.009 - .50
IV (Physical Load), M (Performance), DV (Perceived Taskload)
IV to M (path a)
M to DV (path b)
Total effect of IV on DV (path c)
Direct effect of IV on DV (path c’)
Indirect effects of IV on DV through M (path a*b)
Biased corrected 95% CI: 0.02 - 0.22
N ote.n = 114.p = partial mediation. b∗ = standardized regression coefficient, SE = standard error, CI = confidence interval

b∗
-0.17
-1.28
0.71
0.49
0.22

SE
0.03
0.51
0.18
0.20
0.13

p
0.000
0.014
0.002
0.015

b∗
-0.07
-1.60
0.29
0.18
0.12

SE
0.02
0.49
0.12
0.12
0.05

p
0.001
0.002
0.013
0.13

Figure 4.1: Indirect effect of mental load on perceived task load through performance (partial
mediation).

Figure 4.2: Indirect effect of physical load on perceived task load through performance (full
mediation).

24

CHAPTER 5
DISCUSSION AND CONCLUSION

The purpose of this study was to observe how the perceived physical load and mental load affect
players’ gameplay experiences in exergames in an experimental setting. Using dual-task methodology approach, this research is the first attempt to explain the contributions and interactions between
mental and physical demand required by exergames for gameplay experiences including enjoyment,
flow, competence, effort, future play preferences and performance.

5.1

Discussion

Supporting the previous findings about the positive relationship between mental and physical
workload and overall taskload (DiDomenico and Nussbaum, 2008; Hancock, 1989), we found
that as perceived mental load and physical load increased in an exergame, players’ perceived task
load also increased significantly. As DiDomenico and Nussbaum (2008) showed and Park et al.
(2013) suggested, higher mental and physical load resulted in lower performance in this study (Paas
and Van Merriënboer, 1994; Paas et al., 2003). The mediating effects of performance between
mental (partial) and physical (full) load and perceived task load were found which suggests that
performance can also explain the impacts of these load on how much we perceive the overall task
load. Moreover, subjects rated more perceived effort with an increase in mental load as it was
proposed by Wickens (2017).
Although there was no significance in the results regarding other dependent variables of enjoyment, flow, competence and future play preference, the changes in the patterns with higher mental
and physical load (see Appendix M) are supported by other research. As mental and physical load
increased, enjoyment and flow decreased as proposed in the previous studies (Ang et al., 2007;
Sinclair et al., 2009; Tozman et al., 2015; Qin et al., 2009). When mental and physical load are
observed individually, the results highlight that players enjoy playing the exergame when the physical load and mental load are low. However, when we consider the combinations of these two load,
25

such an approach can mislead us since the interactions are ignored (Kiili et al., 2013, pp. 203-204).
One important aspect in the results is that following the perspective of dual-task interference (Park
et al., 2013), the combinations of different levels of mental and physical load in exergame design
may result in different gameplay experiences. For example, in mental and physical load medium
condition, subjects rated their enjoyment as highest, whereas in mental load medium and physical
load high condition, subjects rated their enjoyment as lowest across all conditions.
The main challenge in this study was to implement dual-task methodology by combining the
physical and mental challenge effectively. In exergames, "players frequently perform multiple tasks
simultaneously" (Park et al., 2013, p. 419). Since, in exergame design, there is a continuous demand
by two channels of tasks (physical performance and mental attention), it becomes crucial to find
a way of manipulating these two for an optimum gameplay experience. This might underline the
reason why there were no significant differences between different levels of loads in their effects on
motivational outcomes such as enjoyment and flow. More specifically, as physical performance also
requires a certain level of cognitive effort (e.g. precision, endurance, balance) (Lyons, 2015), the
target game in this study might be more mental load focused hindering the physical load. Moreover,
"exercise intensity impacts the focus of attention" (as cited in Kiili et al., 2013), and "during high
physical workload it is hard to concentrate on problem-solving and game stimuli" (Kiili et al.,
2013, p. 204). This increase in "allocation of attention" during dual-task performance (Brisswalter
et al., 2002; DiDomenico and Nussbaum, 2011) to coordinate physical and perceptional resources
might intensify the level of mental load required by the game and perceived by the subjects. Thus,
trying to implement and manipulate physical and mental challenges seperately in the game design
might be useless. Considering these interactions, the dual-flow model for exergames highlights
that "optimal exergaming experience can be achieved when both the psychological (cognitive) and
physiological challenges are in balance" (Sinclair et al., 2009). Therefore, assessing exergames
as dual-task interplay, the challenges in the game should be designed in a way that the game
mechanics requiring cognitive and physiological aspects should be adjusted and utilized for a
continuous challenge-skill balance (Martin-Niedecken and Götz, 2017; Kiili et al., 2013; Sinclair

26

et al., 2009; Velazquez et al., 2013). One approach is suggested by Kiili et al. (2013, p. 205) is
"sequence the cognitive gameplay and physical gameplay" such that the tasks are alternated over
time (i.e. one challenge is introduced while controlling the other). Therefore, instead of using
a design which requires concurrent mental and physical effort, a game which implements this
approach (requiring either mental or physical effort at a time) could be tested in future research. In
addition, deviating from dual-task methodology, since physical and cognitive activities are serving
for the same goal of "playing the game", exergames can be evaluated as one combined task of
physical and cognitive activities rather than multiple tasks performed simultaneously. Further
research is needed to reveal the interplay of these two components in exergame design.

5.2

Limitations

5.2.1

Participants

Since only participants enrolled in mainly undergraduate studies at a major Midwest university
participated in the study, this may impact generalizability. Therefore, whether our the findings
in this study hold for other populations remains to be determined. Moreover, when we consider
the characteristics of participants like gameplay and exercising experiences, a wider range of
participants with broader skill levels should be included and these prior experiences can be measured
with well-established instruments from the literature. Additionally, there might be a role of "interest"
towards what the exergame is promising for participants on their overall experiences. According
to Chen and Wang (2017), interest in learning outcomes offered by an exergame may motivate
students to engage in physical activity. Similarly, the outcomes offered by the exergame (e.g. energy
expenditure, in-game rewards) might have an impact on subjects’ motivational experience during
gameplay. In the future research, the ratings for such an inclination (e.g. interest in exercising,
interest in competition) might be also observed for its effect on overall gameplay experiences.

27

5.2.2

Exercising Characteristics and Cognitive Task Type

Exercising duration and increased fatigue are some determinants of the effects of physical exertion
on cognitive performance (Brisswalter et al., 2002, p.559). For example, with a moderate level of
physical exercise, performance can increase and with a more intense exercising activity cognitive
performance may start to decrease with an Inverted-U effect (as cited in Brisswalter et al., 2002. In
this study, due to the length of procedure in the experimental setting with the calibration sessions,
we limited the time of the gameplay to 5 minutes. Since longer gameplay activity may result in
higher levels of physical demand, its contributions to the overall gameplay experiences might be
different. Moreover, although mental load was manipulated using multiple-object tracking task in
the study, other types of cognitive tasks (e.g. requiring complex decision making, reaction time)
should be integrated in the exergame design (using different game genres) and tested for future
research.

5.2.3

Exercising Device

Lastly, exercising device was another limitation in our study. Since we tested our hypotheses on
one exercising medium, gameplay experiences might be different on when using another device.
Independently from the game design itself, different exercising settings may require different levels
of physical load (and probably cognitive attention) which may affect gameplay experience in turn. A
comparison between different exercising devices may also contribute to the research in this direction.

Focusing on the motivational outcomes, this study is promising for the future research to observe
the dynamic and continuous interaction between mental and physical load in exergames. A design
guideline can be provided for optimized gameplay experiences through the lens of mental-physical
load balance in exergames after conducting future research with more significant results on game
enjoyment and flow. Considering the limitations listed above, future work can help designers to
create better exergames for a permanent interest in exercising and continuous interest in gameplay.

28

APPENDICES

29

APPENDIX A
SCREENING SURVEY

A.1

Demographics

1. What is your gender?
 Male
 Female
 Transgender
2. What is your age?
 16
 17
 18
 19
 20
 21
 22
 23
 24
 25
 over 25
3. How would you describe yourself?
 American Indian or Alaska Native
30

 Asian
 Black or African American
 Hispanic or Latino
 Native Hawaiian or Other Pacific Islander
 White or Caucasian
 Mixed Race

A.2

Physical Activity Readiness (PAR)

Thomas, S., Reading, J., & Shephard, R. J. (1992). Revision of the Physical Activity Readiness
Questionnaire (PAR-Q). Canadian journal of sport sciences.
1. Has your doctor ever said that you have a heart condition and that you should only do physical
activity recommended by a doctor?
 Yes
 No
2. Do you feel pain in your chest when doing physical activity?
 Yes
 No
3. In the past month, have you had chest pain when you were not doing physical activity?
 Yes
 No
4. Do you lose your balance because of dizziness or do you ever lose consciousness?
 Yes
31

 No
5. Do you have a bone or joint problem that could be made worse by a change in your physical
activity?
 Yes
 No
6. Is your doctor currently prescribing medications for your blood pressure or heart condition?
 Yes
 No
7. Do you know of any other reason why you should not do physical activity?
 Yes
 No

A.3

Previous Gaming Experience

1. How many years have you been playing video games?

(e.g. 10)

2. How many hours in a week do you play video games?

(e.g. 10)

3. Which genre of video games do you play? (Check all that apply)
 Action
 Action-Adventure
 Simulation
 Strategy
 Role Playing
 Sports
32

 Casual
 Other (Please specify):
4. Are you comfortable playing a casual game including a cognitive task?
 Yes
 No
5. Are you comfortable tracking task of multiple objects in the screen?
 Yes
 No
6. Are you comfortable using exercise bike while playing a casual game including a cognitive
task?
 Yes
 No

A.4

Previous Exercising Experience

1. Are you comfortable doing exercise?
 Yes
 No
2. How many hours do you exercise in a week?
3. Which exercise devices do you use? (Check all that apply)
 Treadmill
 Exercise Bike
 Elliptical
33

(e.g. 6)

 Weight Benches
 Home Gyms
 Rowers
 Other (Please specify):
4. What type of exercises do you do? (Check all that apply)
 Walking
 Running
 Swimming
 Cycling
 Hiking
 Fitness
 Yoga / Plates
 Strength & Weight Training
 Other (Please specify):
5. Are you familiar with exercise bikes?
 Yes
 No

34

APPENDIX B
CONSENT FORM

INVESTIGATORS
Irem Gokce Yildirim, yildiri4@msu.edu Kuo-Ting Huang, huangku1@msu.edu Tom Day,
daythoma@msu.edu Dr. Taiwoo Park, twp@msu.edu, 517-353-2198. 423 Communication Arts
& Sciences, MSU Dr. Wei Peng, pengwei@msu.edu, 517-432-8235, 429 Communication Arts &
Sciences, MSU

GENERAL INFORMATION ABOUT RESEARCH STUDIES
You are being asked to take part in a research study. To join the study is completely voluntary.
You may refuse to join, or you may withdraw your consent to be in the study at any time, for any
reason, without penalty or loss of any benefits to which you are entitled. Details about this study
are discussed below. It is important that you understand this information so that you can make an
informed choice about being in this research study.

PURPOSE OF RESEARCH
You are being asked to participate in a research study on the perception of physical and cognitive
loads in exergames. You have been selected as a possible participant in this study because of your
involvement with the MSU SONA system. From this study, the researchers hope to understand
experiences of exergames.

LENGTH OF STUDY
Participants will complete the study only once and the procedure will take about 60 minutes.

35

PROCEDURE OF STUDY
The study is part of an academic research project to examine how people experience exergames.
You completed an online screening survey before coming into the lab and we will use this information for research only if you agree to participate in this study.
In the lab, the first session will be the individual based ‘calibration’ session to gauge your
baseline physical and cognitive performance levels in two separate phases.
• In the cognitive challenge calibration phase, you will be asked to play the game (without
physical load - no exercise device), while gradually increasing the difficulty of game (e.g., by
increasing the number of objects to track on the screen) over time. Anytime when you do not
want to play the game, you are allowed to stop the process. At the end of cognitive challenge
calibration phase, you will be given 1 minutes to rest.
• In the physical challenge calibration phase, you will be asked to cycle at certain speed (e.g.,
100 RPM) on the exercise bike, while gradually increasing the tension level over time, to
identify the maximum physical challenge for you. Anytime when you feel it becomes either
too hard or uncomfortable for you to cycle, you are allowed to stop the process. At the end
of physical challenge calibration phase, you will be given 5 minutes to rest.
After the calibration session, the gameplay session will be conducted. During this session, some
gameplay data will be logged. Only the researchers of this study will have access to this data. You
will be asked to play the video game using exercise bike for about 5 minutes, and after playing,
you will be asked to complete some basic questions about your gameplay experiences, including
physical and cognitive aspects. The entire procedure should take no more than 50 to 60 minutes to
complete.

POTENTIAL RISKS
While the game used in this study is age-appropriate for teens and adults and there is an
individual based adaptation for cognitive challenge and physical challenge on the exercise bike,
36

there is the possibility that you might feel uncomfortable when playing the game and cycling. It
is also possible that you may feel fatigue or experience some muscle soreness after performing
the physical workout. This will not be more than one you experience in your exercise routine.
Therefore, there will be no special problems or risks during or after the study. If you have some
physical injury, performing the tasks may worse it. If you have such injury, you may not participate
in this study. Should you feel any discomfort as a result of this (or any other) reasons resulting from
your participation you may withdraw your consent and data from the study by stopping playing at
any time. If you withdraw from the study before its completion, we will not use any of the data you
provide us.
If you encounter any injury as a result of your participation in the study, Michigan State
University will assist you in obtaining emergency care. If you have insurance for medical care, the
bills of treatment will be charged to your insurance carrier and any costs that are not covered by
your insurance will be your responsibility.

PRIVACY AND CONFIDENTIALITY
The data for this project are being collected anonymously using your SONA ID. Neither the
researchers nor anyone else will be able to link data to you. Your survey answers for the study
will be collected and stored on a separate secure website. Information that is printed out will be
stored in a locked filing cabinet in the locked office of the principal investigator and destroyed when
analysis is complete. Data that is downloaded will be stored on the investigators’ laptop computer
and will have a double password protected log-on and the files will be password protected as well.
Additionally, the IRB will have access to the data. Research records will be kept for at least 3 years
after the close of the study. The only people with access to the data will be Dr. Taiwoo Park, Dr.
Wei Peng, Irem Gokce Yildirim, Tim Huang, Tom Day and the MSU HRPP. Your confidentiality
will be protected to the maximum extent allowable by law.
Participants will not be identified in any report or publication about this study.
If you have questions or concerns about your role and rights as a research participant, would like

37

to obtain information or offer input, or would like to register a complaint about this study, you may
contact, anonymously if you wish, the Michigan State University’s Human Research Protection
Program at 517-355-2180, Fax 517-432-4503, or e-mail irb@msu.edu or regular mail at Olds Hall,
408 West Circle Drive #207, MSU, East Lansing, MI 48824.

COSTS AND COMPENSATION FOR BEING IN THE STUDY
In the SONA system, 1 hour of research participation is worth 1 SONA credit and this credit is
pro-rated in 15-minute increments. The duration of this study is approximately 60 minutes. Hence,
participants who complete this study will receive 1 SONA credits.

QUESTIONS ABOUT THIS 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 researcher (Irem Gokce Yildirim, 430 Communication Arts and Sciences Building, yildiri4@msu.edu, 517-348-3866; Kuo-Ting (Tim) Huang,
huangku1@msu.edu, 850-345-3805; Tom Day, daythoma@msu.edu, 586-453-1832; Dr. Taiwoo
Park, 517-353-2198; Dr. Wei Peng, 517-432-8235).

STATEMENT OF CONSENT
Your signature below means that you voluntarily agree to participate in this research
study.

Signature

Date

38

APPENDIX C
NASA-TLX: TASK LOAD INDEX

Hart, S. G. (2006, October). NASA-task load index (NASA-TLX); 20 years later. In Proceedings
of the human factors and ergonomics society annual meeting (Vol. 50, No. 9, pp. 904-908). Sage
CA: Los Angeles, CA: Sage Publications. Retrieved from: https://humansystems.arc.nasa.
gov/groups/tlx/downloads/TLXScale.pdf

Please rate the work load you experienced during the gameplay in the following questions, using
the following scale.

1

2

3

4

5

6

Very low

7
Very high

1. How mentally demanding was the task? (MENTAL DEMAND)
2. How physically demanding was the task? (PHYSICAL DEMAND)
3. How hurried or rushed was the pace of the task? (TEMPORAL DEMAND)
4. How successful were you in accomplishing what you were asked to do? (R) (PERFORMANCE)
5. How hard did you have to work (mentally and physically) to accomplish your level of
performance? (EFFORT)
6. How insecure, discouraged, irritated, stressed, and annoyed were you? (FRUSTRATION
LEVEL)

39

APPENDIX D
BORG RATING OF PERCEIVED EXERTION SCALE

Borg, G. A. (1982). Psychophysical bases of perceived exertion. Med sci sports exerc, 14(5),
377-381.

Please choose the number from below that best describes your level of exertion.
 6 - No exertion at all
 7 - Extremely light
 8
 9 - Very light (easy walking slowly at a comfortable pace)
 10
 11 - Light
 12
 13 - Somewhat hard (It is quite an effort; you feel tired but can continue)
 14
 15 - Hard (heavy)
 16
 17 - Very hard (very strenuous, and you are very fatigued)
 18
 19 - Extremely hard (You can not continue for long at this pace)
 20 - Maximal exertion
40

APPENDIX E
GAME ENJOYMENT

Song, H., Peng, W., & Lee, K. M. (2011). Promoting exercise self-efficacy with an exergame.
Journal of health communication, 16(2), 148-162.

Please rate how well the following adjectives describe your game playing experience, using the
following scale.

1

2

3

4

Very poor

5

6

7
Very well

1. boring (R)
2. exciting
3. enjoyable
4. entertaining
5. fun
6. interesting
7. pleasant

41

APPENDIX F
GAME FLOW

Fu, F. L., Su, R. C., & Yu, S. C. (2009). EGameFlow: A scale to measure learners’ enjoyment of
e-learning games. Computers & Education, 52(1), 101-112.

Please indicate your agreement with the following statements, using the following scale.

1
Strongly

2

3

4

5

6

7
Strongly Agree

Disagree

1. I forgot about time passing while playing the game.
2. I became unaware of my surroundings while playing the game.
3. I temporarily forgot worries about everyday life while playing the game.
4. I experienced an altered sense of time.
5. I could become involved in the game.
6. I felt emotionally involved in the game.
7. I felt viscerally involved in the game.
8. I enjoyed the game without feeling bored or anxious.
9. The challenge was adequate, neither too difficult nor too easy.

42

APPENDIX G
PERCEIVED COMPETENCE

McAuley, E., Duncan, T., & Tammen, V. V. (1989). Psychometric properties of the Intrinsic Motivation Inventory in a competitive sport setting: A confirmatory factor analysis. Research quarterly
for exercise and sport, 60(1), 48-58. Retrieved from http://selfdeterminationtheory.org/
intrinsic-motivation-inventory/

For each of the following statements, please indicate how true it is for you, using the following
scale.

1

2

3

Not at all true

4

5

6

Somewhat true

1. I think I am pretty good at this game.
2. I think I did pretty well at this game, compared to other players.
3. After playing this game for awhile, I felt pretty competent.
4. I am satisfied with my performance at this game.
5. I could become involved in the game.
6. I was pretty skilled at this game.
7. This was a game that I couldn’t do very well. (R)

43

7
Very true

APPENDIX H
PERCEIVED EFFORT

McAuley, E., Duncan, T., & Tammen, V. V. (1989). Psychometric properties of the Intrinsic Motivation Inventory in a competitive sport setting: A confirmatory factor analysis. Research quarterly
for exercise and sport, 60(1), 48-58. Retrieved from http://selfdeterminationtheory.org/
intrinsic-motivation-inventory/

For each of the following statements, please indicate how true it is for you, using the following
scale.

1

2

3

Not at all true

4
Somewhat true

1. I put a lot of effort into this game.
2. I didn’t try very hard to do well at this game. (R)
3. I tried very hard on this game.
4. It was important to me to do well at this game.
5. I didn’t put much energy into this game. (R)

44

5

6

7
Very true

APPENDIX I
FUTURE PLAY PREFERENCES

Ryan, R. M., Rigby, C. S., & Przybylski, A. (2006). The motivational pull of video games: A
self-determination theory approach. Motivation and emotion, 30(4), 344-360.

For each of the following statements, please indicate how true it is for you, using the following
scale.

1
Not at all true

2

3

4

5

Somewhat true

6

7
Very true

1. Given the chance I would play this game in my free time.
2. I would be interested in having my own personal copy of this game.
3. I would like to spend more time playing this game.
4. I am excited to play this game in the future.

45

APPENDIX J
PERFORMANCE

Performance was calculated as the ratio of the average of the successful feeding number across all
of the rounds played during the main gameplay session to the total number of turtles set for the
gameplay session. The data was collected during gameplay in a log file.

46

APPENDIX K
OTHER GAMEPLAY METRICS

Average speed and total gameplay duration were other gameplay log metrics we collected.
• Average Speed: Calculated based on the non-zero speed information during gameplay.
• Total Gameplay Time: Total time spent on gameplay during 5 min of gameplay duration.
• Total Resting Time: Total time spent on resting without cycling (speed=0) during the 5 min
of gameplay time.

47

APPENDIX L
CALIBRATION SESSION STATISTICS

L.1

Mental Load Calibration Phase Data

Figure L.1: Perceived mental load ratings across participants as mental load increases during
mental load calibration phase trials.

Figure L.2: Subjects’ last perceived mental load ratings and maximum mental load levels
identified at the end of mental load calibration phase.
(Mmaximum_mental_load = 10.44, Mlast_mental_load_rating = 5.61)

48

Figure L.3: Subjects’ last perceived mental load ratings and maximum mental load level statistics
at the end of mental load calibration phase.

L.2

Physical Load Calibration Phase Data

49

Figure L.4: Subjects’ perceived exertion (BORG) ratings as physical load increases during
physical load calibration phase.

Figure L.5: Subjects’ last perceived exertion (BORG) ratings and maximum physical load level
statistics at the end of physical load calibration phase.
(Mmaximum_physical_load = 7.08, Mlast_physical_load_rating = 18.50)

50

Figure L.6: Subjects’ last perceived exertion ratings and maximum physical load level statistics at
the end of physical load calibration phase.

51

APPENDIX M
MEAN SCORES FOR THE DEPENDENT VARIABLES

Figure M.1: Mean scores for NASA-TLX across mental and physical load conditions respectively.

52

Figure M.2: Mean scores for enjoyment across mental and physical load conditions respectively.

53

Figure M.3: Mean scores for flow across mental and physical load conditions respectively.

54

Figure M.4: Mean scores for competence across mental and physical load conditions respectively.

55

Figure M.5: Mean scores for effort across mental and physical load conditions respectively.

56

Figure M.6: Mean scores for performance across mental and physical load conditions
respectively.

57

BIBLIOGRAPHY

58

BIBLIOGRAPHY

Allen, R., Mcgeorge, P., Pearson, D. G., and Milne, A. (2006). Multiple-target tracking: A role for
working memory? The Quarterly journal of experimental psychology, 59(6):1101–1116.
Ang, C. S., Zaphiris, P., and Mahmood, S. (2007). A model of cognitive loads in massively
multiplayer online role playing games. Interacting with computers, 19(2):167–179.
Annetta, L. A., Minogue, J., Holmes, S. Y., and Cheng, M.-T. (2009). Investigating the impact of
video games on high school students’ engagement and learning about genetics. Computers &
Education, 53(1):74–85.
Arroyo-Gómez, N., Laparra-Hernández, J., Soler-Valero, A., Medina, E., and de Rosario, H. (2017).
Physiological model to classify physical and cognitive workload during gaming activities. In
International Conference on Applied Human Factors and Ergonomics, pages 246–254. Springer.
Borg, G. (1990). Psychophysical scaling with applications in physical work and the perception of
exertion. Scandinavian journal of work, environment & health, pages 55–58.
Borg, G. (1998). Borg’s perceived exertion and pain scales. Human kinetics.
Botterill, K., Allen, R., and McGeorge, P. (2011). Multiple-object tracking. Experimental psychology.
Brisswalter, J., Collardeau, M., and René, A. (2002). Effects of acute physical exercise characteristics on cognitive performance. Sports medicine, 32(9):555–566.
Brunken, R., Plass, J. L., and Leutner, D. (2003). Direct measurement of cognitive load in
multimedia learning. Educational psychologist, 38(1):53–61.
Capodaglio, P. (2001). The use of subjective rating of exertion in ergonomics. Giornale italiano
di medicina del lavoro ed ergonomia, 24(1):84–89.
Chen, A. and Wang, Y. (2017). The role of interest in physical education: A review of research
evidence. Journal of Teaching in Physical Education, 36(3):313–322.
Constant, T., Levieux, G., Buendia, A., and Natkin, S. (2017). From objective to subjective difficulty
evaluation in video games. In IFIP Conference on Human-Computer Interaction, pages 107–127.
Springer.
Coppin, A. K., Shumway-Cook, A., Saczynski, J. S., Patel, K. V., Ble, A., Ferrucci, L., and
Guralnik, J. M. (2006). Association of executive function and performance of dual-task physical
tests among older adults: analyses from the inchianti study. Age and ageing, 35(6):619–624.
Csikszentmihalyi, M. (1990). Flow: the psychology of optimal experience. Harper Perennial
modern classics.
Damos, D. (1991). Multiple task performance. CRC Press.
59

Davey, C. (1973). Physical exertion and mental performance. Ergonomics, 16(5):595–599.
DiDomenico, A. and Nussbaum, M. A. (2008). Interactive effects of physical and mental workload
on subjective workload assessment. International journal of industrial ergonomics, 38(11):977–
983.
DiDomenico, A. and Nussbaum, M. A. (2011). Effects of different physical workload parameters on
mental workload and performance. International Journal of Industrial Ergonomics, 41(3):255–
260.
Franconeri, S., Jonathan, S., and Scimeca, J. (2010). Tracking multiple objects is limited only by
object spacing, not by speed, time, or capacity. Psychological science, 21(7):920–925.
Fu, F.-L., Su, R.-C., and Yu, S.-C. (2009). Egameflow: A scale to measure learners’ enjoyment of
e-learning games. Computers & Education, 52(1):101–112.
Hancock, P. (1989). The effect of performance failure and task demand on the perception of mental
workload. Applied Ergonomics, 20(3):197–205.
Hardy, S., Dutz, T., Wiemeyer, J., Göbel, S., and Steinmetz, R. (2015). Framework for personalized
and adaptive game-based training programs in health sport. Multimedia Tools and Applications,
74(14):5289–5311.
Hart, S. G. and Staveland, L. E. (1988). Development of nasa-tlx (task load index): Results of
empirical and theoretical research. Advances in psychology, 52:139–183.
Huang, W. D. and Johnson, T. (2009). Instructional game design using cognitive load theory. In
Handbook of research on effective electronic gaming in education, pages 1143–1165. IGI Global.
Hutchinson, J. C. and Tenenbaum, G. (2006). Perceived effort—can it be considered gestalt?
Psychology of Sport and Exercise, 7(5):463–476.
Jakobsen, M. D., Sundstrup, E., Persson, R., Andersen, C. H., and Andersen, L. L. (2014). Is borg’s
perceived exertion scale a useful indicator of muscular and cardiovascular load in blue-collar
workers with lifting tasks? a cross-sectional workplace study. European journal of applied
physiology, 114(2):425–434.
Kalyuga, S., Chandler, P., and Sweller, J. (1999). Managing split-attention and redundancy in
multimedia instruction. Applied cognitive psychology, 13(4):351–371.
Kalyuga, S. and Plass, J. L. (2009). Evaluating and managing cognitive load in games. In Handbook
of research on effective electronic gaming in education, pages 719–737. IGI Global.
Kiili, K. and Perttula, A. (2013). A design framework for educational exergames. New pedagogical
approaches in game enhanced learning: Curriculum integration, pages 136–158.
Kiili, K., Perttula, A., Arnab, S., and Suominen, M. (2013). Flow experience as a quality measure
in evaluating physically activating serious games. In International Conference on Games and
Learning Alliance, pages 200–212. Springer.

60

Kooiman, B. J. and Sheehan, D. P. (2014). The efficacy of exergames played proximally and over the
internet on cognitive functioning for online physical education. American Journal of Distance
Education, 28(4):280–291.
Krausman, A. S., Crowell III, H. P., and Wilson, R. M. (2002). The effects of physical exertion on
cognitive performance. Technical report, ARMY RESEARCH LAB ABERDEEN PROVING
GROUND MD.
Kyndt, E., Dochy, F., Struyven, K., and Cascallar, E. (2011a). The direct and indirect effect of
motivation for learning on students’ approaches to learning through the perceptions of workload
and task complexity. Higher Education Research & Development, 30(2):135–150.
Kyndt, E., Dochy, F., Struyven, K., and Cascallar, E. (2011b). The perception of workload and task
complexity and its influence on students’ approaches to learning: A study in higher education.
European journal of psychology of education, 26(3):393–415.
Lai, Y.-C., Wang, S.-T., and Yang, J.-C. (2012). An investigation of the exergames experience with
flow state, enjoyment, and physical fitness. In Advanced Learning Technologies (ICALT), 2012
IEEE 12th International Conference on, pages 58–60. IEEE.
Larsen, L. H., Schou, L., Lund, H. H., and Langberg, H. (2013). The physical effect of exergames
in healthy elderly—a systematic review. GAMES FOR HEALTH: Research, Development, and
Clinical Applications, 2(4):205–212.
Lyons, E. J. (2015). Cultivating engagement and enjoyment in exergames using feedback, challenge,
and rewards. Games for health journal, 4(1):12–18.
Maillot, P., Perrot, A., and Hartley, A. (2012). Effects of interactive physical-activity video-game
training on physical and cognitive function in older adults. Psychology and aging, 27(3):589.
Martin-Niedecken, A. L. and Götz, U. (2017). Go with the dual flow: Evaluating the psychophysiological adaptive fitness game environment “plunder planet”. In Joint International Conference
on Serious Games, pages 32–43. Springer.
McAuley, E., Duncan, T., and Tammen, V. V. (1989). Psychometric properties of the intrinsic
motivation inventory in a competitive sport setting: A confirmatory factor analysis. Research
quarterly for exercise and sport, 60(1):48–58.
Miller, G. A. (1956). The magical number seven, plus or minus two: some limits on our capacity
for processing information. Psychological review, 63(2):81.
Miyake, A., Friedman, N. P., Emerson, M. J., Witzki, A. H., Howerter, A., and Wager, T. D. (2000).
The unity and diversity of executive functions and their contributions to complex “frontal lobe”
tasks: A latent variable analysis. Cognitive psychology, 41(1):49–100.
Nankar, M. (2016). The influence of different mental processes (cognitive loads) on gait: A study
of dual task function. Master’s thesis, University of Manitoba.
O’Neil, H. F., Wainess, R., and Baker, E. L. (2005). Classification of learning outcomes: Evidence
from the computer games literature. The Cirriculum Journal, 16(4):455–474.
61

Owen, N., Sparling, P. B., Healy, G. N., Dunstan, D. W., and Matthews, C. E. (2010). Sedentary
behavior: emerging evidence for a new health risk. In Mayo Clinic Proceedings, volume 85,
page 1138. Mayo Foundation.
Paas, F., Tuovinen, J. E., Tabbers, H., and Van Gerven, P. W. (2003). Cognitive load measurement
as a means to advance cognitive load theory. Educational psychologist, 38(1):63–71.
Paas, F. G. and Van Merriënboer, J. J. (1993). The efficiency of instructional conditions: An
approach to combine mental effort and performance measures. Human factors, 35(4):737–743.
Paas, F. G. and Van Merriënboer, J. J. (1994). Instructional control of cognitive load in the training
of complex cognitive tasks. Educational psychology review, 6(4):351–371.
Park, T., Lee, U., Lee, B., Lee, H., Son, S., Song, S., and Song, J. (2013). Exersync: facilitating
interpersonal synchrony in social exergames. In Proceedings of the 2013 conference on Computer
supported cooperative work, pages 409–422. ACM.
Park, T., Lee, U., MacKenzie, S., Moon, M., Hwang, I., and Song, J. (2014). Human factors of
speed-based exergame controllers. In Proceedings of the SIGCHI Conference on Human Factors
in Computing Systems, pages 1865–1874. ACM.
Pimenta, A., Gonçalves, S., Carneiro, D., Fde-Riverola, F., Neves, J., and Novais, P. (2015). Mental
workload management as a tool in e-learning scenarios. In Pervasive and Embedded Computing
and Communication Systems (PECCS), 2015 International Conference on, pages 25–32. IEEE.
Pisan, Y., Marin, J. G., and Navarro, K. F. (2013). Improving lives: using microsoft kinect to predict
the loss of balance for elderly users under cognitive load. In Proceedings of The 9th Australasian
Conference on Interactive Entertainment: Matters of Life and Death, page 29. ACM.
Pylyshyn, Z. (2004). Some puzzling findings in multiple object tracking: I. tracking without
keeping track of object identities. Visual cognition, 11(7):801–822.
Qin, H., Rau, P.-L. P., and Salvendy, G. (2009). Effects of different scenarios of game difficulty on
player immersion. Interacting with Computers, 22(3):230–239.
Robertson, R. J. and Noble, B. J. (1997). 15 perception of physical exertion: Methods, mediators,
and applications. Exercise and sport sciences reviews, 25(1):407–452.
Rubio-Valdehita, S., López-Higes, R., and Díaz-Ramiro, E. (2014). Academic context and perceived
mental workload of psychology students. The Spanish journal of psychology, 17.
Ryan, R. M., Rigby, C. S., and Przybylski, A. (2006). The motivational pull of video games: A
self-determination theory approach. Motivation and emotion, 30(4):344–360.
Ryu, K. and Myung, R. (2005). Evaluation of mental workload with a combined measure based on
physiological indices during a dual task of tracking and mental arithmetic. International Journal
of Industrial Ergonomics, 35(11):991–1009.

62

Sinclair, J., Hingston, P., and Masek, M. (2009). Exergame development using the dual flow model.
In Proceedings of the Sixth Australasian Conference on Interactive Entertainment, page 11.
ACM.
Song, H., Peng, W., and Lee, K. M. (2011). Promoting exercise self-efficacy with an exergame.
Journal of health communication, 16(2):148–162.
Staiano, A. E. and Calvert, S. L. (2011). The promise of exergames as tools to measure physical
health. Entertainment computing, 2(1):17–21.
Strobach, T., Frensch, P. A., and Schubert, T. (2012). Video game practice optimizes executive
control skills in dual-task and task switching situations. Acta psychologica, 140(1):13–24.
Sun, H. (2013). Impact of exergames on physical activity and motivation in elementary school
students: A follow-up study. Journal of Sport and Health Science, 2(3):138–145.
Sween, J., Wallington, S. F., Sheppard, V., Taylor, T., Llanos, A. A., and Adams-Campbell, L. L.
(2014). The role of exergaming in improving physical activity: a review. Journal of Physical
Activity and Health, 11(4):864–870.
Sweetser, P. and Wyeth, P. (2005). Gameflow: a model for evaluating player enjoyment in games.
Computers in Entertainment (CIE), 3(3):3–3.
Sweller, J. (1994). Cognitive load theory, learning difficulty, and instructional design. Learning
and instruction, 4(4):295–312.
Tozman, T., Magdas, E. S., MacDougall, H. G., and Vollmeyer, R. (2015). Understanding the
psychophysiology of flow: A driving simulator experiment to investigate the relationship between
flow and heart rate variability. Computers in Human Behavior, 52:408–418.
Van Muijden, J., Band, G. P., and Hommel, B. (2012). Online games training aging brains: limited
transfer to cognitive control functions. Frontiers in human neuroscience, 6.
Velazquez, A., Martinez-Garcia, A. I., Favela, J., Hernandez, A., and Ochoa, S. F. (2013). Design
of exergames with the collaborative participation of older adults. In Computer Supported Cooperative Work in Design (CSCWD), 2013 IEEE 17th International Conference on, pages 521–526.
IEEE.
Wahlström, J. (2005). Ergonomics, musculoskeletal disorders and computer work. Occupational
Medicine, 55(3):168–176.
Wickens, C. D. (2017). Mental workload: Assessment, prediction and consequences. In International Symposium on Human Mental Workload: Models and Applications, pages 18–29.
Springer.
Xie, J., Xu, G., Wang, J., Li, M., Han, C., and Jia, Y. (2016). Effects of mental load and fatigue
on steady-state evoked potential based brain computer interface tasks: a comparison of periodic
flickering and motion-reversal based visual attention. PloS one, 11(9):e0163426.

63