The role of customization on avatar identification and performance in a kart racing game

THE ROLE OF CUSTOMIZATION ON AVATAR IDENTIFICATION AND PERFORMANCE IN A KART RACING GAME

Tianzuo Peng

Submitted in partial fulfillment of the requirements

for the degree of Master of Science in Game Science and Design

in the Graduate School of the College of Arts, Media and Design of

Northeastern University

May, 2021

Table of Contents

Please note that the table of content is prior to all other content in your thesis and is required.

Example:

Abstract 2

Acknowledgements 4

List of Figures 6

List of Tables 7

Introduction 8

Chapter 1: Background 10

Chapter 2: Methodology 26

Chapter 3: Game Design 38

Chapter 4: Results 56

Chapter 5: Discussion 77

Chapter 6: Conclusion 110

References 120

Appendix A 126

Appendix B 128

Appendix C 134

Acknowledgments

Optional.

Abstract

As the gaming industry has grown, many games allow players to customize their character.

Research shows that character customization enhances avatar identification, and avatar

identification affects many aspects of a player, including psychology, behavior, arousal, learning,

and self-building. Few studies focus on the interaction between avatar identification and

performance in racing games. The character types in racing games are different from those in

other games. In kart racing games, players can see their character and the kart at the same time.

When players can customize two different types of avatar and see them simultaneously in the

game, I wonder how avatar customization affects avatar identification and performance. A kart

racing game was designed and developed to study the influence of the combination of different

types of avatar on avatar identification and player performance. I compared four groups of

participants. In the first three groups, players customized different avatar types: (1) driver and

car, (2) driver only, (3) car only, and the control group with (4) the default character and kart.

The research collects avatar identification and player experience through a questionnaire, and

collects player performance through in-game data. The results showed that the Character Only

group had significantly higher avatar identification than the Kart Only, and the Kart Only group

performed significantly better than the Character Only group. There was no difference between

the other groups. Different types of customizations affect the avatar identification and

performance. Developers should pay attention to the type of customization, when designing kart

racing games.

List of Figures/Tables

Optional.

1. INTRODUCTION

As the gaming industry has grown, many games allow players to customize their character

characteristics. Players can enhance the avatar identification by customizing their own

characters, thus enhancing the flow experience (Li et al., 2018). There have been many studies

on avatar identification. Research has shown that avatar identification affects many aspects of a

player, including psychology, behavior, arousal, learning, and self-building (Biocca, 2014). In

massively multiplayer online games, allowing players to customize the appearance of their

characters improves avatar identification, which in turn increases players’ flow and game loyalty

(Mancini & Sibilla, 2017). Kao and Harrell’s findings also suggest that avatar type and

appearance improve character identity, player experience (Kao, 2019), and learning

performance (Kao & Harrell, 2015).

However, those studies take place in VR games, adventure games and jump games, and few

studies focus on the interaction between avatar identification and other factors in racing games.

The character types in racing games are different from those in other games. In common game

types, the avatar is usually humanoid, while in racing games, the avatar is usually a car, a

non-humanoid character. In kart racing games, players can see their character and the kart at

the same time. When players can customize two different types of avatar and see them

simultaneously in the game, I wonder how avatar customization affects avatar identification and

performance.

This paper aims to design and develop a kart racing game to study the influence of different

types of avatar on avatar identification and player performance. There will be four versions of

the game. In the first three versions, players are able to customize their avatars, and each group

has a different type of avatar to customize (driver and car, driver only, car only). The other

version serves as a control group, allowing players to use only the default avatar and kart. The

research collects avatar identification through a questionnaire, and collects player performance

through in-game data.

2. BACKGROUND

In order to design the game and analyze the effects of the game, it is necessary to learn from

relevant researches. Based on previous research, the following part explains three basic concepts

of this research: avatar identification, avatar customization, and racing game. The definition and

findings of each concept are explained.

2.1 Avatar Identification

The definition of avatar is varied from study to study. In this study, the definition of avatar is “a

stand-in for the person, which may be represented by a human character, another animal, a

vehicle, or any other image” (Hudson & Hurter, 2016). In this study, both the character and the

kart are created by the player, and the driving behavior is determined by the player, in this way,

the combination of the character and the kart serves as a stand-in for the player's activities in

the game world.

Avatar identification, according previous study, is “a temporary shift in self-perception whereby

the game role functions as a form of self-priming leading the player to imaginarily assimilate the

media character’s properties into their self-perception for the time of media exposure” (Looy et

al., 2012).

In this study, since the media is video games, the difference between video games and

traditional media should not be ignored. The definition of avatar identification refers to Li’s

model. The model includes four components: feelings during play, absorption during play,

positive attitudes toward the avatar, and importance of the avatar to one’s self identity. “Player

avatar identification is defined as the status when a player is absorbed in video games with

heightened feelings and adopts certain aspects of the in-game identity both emotionally and

cognitively” (Li et al., 2013).

Many researchers have studied the effects of avatar identification on players’ psychology and

behavior.

In terms of psychology, avatar identification affects players' internal motivation and game

experience. Avatar identification increases players' internal motivation. The longer the player

plays, the larger the avatar identification is translated into motivational behavior (Birk et al.,

2016). Avatar identification can enhance play experience, intrinsic motivation and self-efficacy

(Kao & Harrell, 2018).

In terms of behavior, avatar identification is associated with game addiction. In massively

multiplayer online games, the more ideal the avatar, the higher the level of identification. The

ideal avatar directly affects game addiction. A player who has customized an ideal avatar and has

a high identification with the avatar is more likely to become addicted to the game (Mancini et

al., 2019). In addition, avatar identification promotes the time that players spend on games, and

significantly improves the performance (the overall quality of player-designed game levels) (Kao

& Harrell, 2018).

Previous studies provide background knowledge and research methods, and shows the role of

avatar identification on players’ psychology and behavior. This study focuses on measuring the

avatar identification in a kart racing game, so the concept of avatar identification is important.

2.2 Avatar Customization

This study focuses on the avatar, and the concept of avatar customization also needs to be

explained. The purpose of avatar customization is to approach the players’ preference for the

appearance of their character in the game. Games often offer accessories or color schemes, and

players can customize avatars by choosing eye colors, hairstyles and body shapes (Liao et al.,

2019). Previous researches studied the effects of avatar customization on players’ psychology

and behavior.

In terms of psychology, customized avatars affect the player's identification. In the large

multiplayer online game, Lord of the rings online (LotRO), researchers found that game time

and customization of avatars have a positive impact on players' avatar identification (Turkay &

Kinzer, 2014). Avatar customization is also associated with game enjoyment and gender

stereotypes. Previous research found that customizing driving vehicles had a positive impact on

the enjoyment of a console driving game (Schmierbach et al., 2012). Avatar customization

aroused gender stereotypes and influenced the user's post-game behavior. Players who used

male avatars performed better in the post-game math task than those who used female avatars

among players who had customized their avatars with low embodiment (Ratan & Sah, 2015).

In terms of behavior, a previous study examined the effects of cooperative and competitive

games on spontaneous helping behavior. They found that cooperative participants who

customized their roles had more helping behaviors than those who competed (Dolgov et al.,

2014). In massively multiplayer online games, avatar customization is positively correlated with

avatar identification and online player loyalty (Liao et al., 2019). The type of avatar can trigger

stereotypes and affect students’ identification and learning performance. In a learning game,

players who use shape avatars perform better and have higher engagement than those who use

customized avatars (Kao & Harrell, 2015).

2.3 Driving Games

This study required the creation of an arcade style kart racing game, so it is necessary to clarify

the differences between the types of games. Kart racing is a sub-category of racing games.

Racing game is a type of video game in which players race by using different vehicles. The game

styles of racing games are divided into simulation and arcade. Simulation games focus on

realistic handling and physical feedback, while arcade styles have their own physics rules and

focus on the fun of racing.

In kart racing games, the avatar consists of a driver and a car, which are two different types of

avatar. Previous research proves the importance of virtual avatar type in the design of virtual

systems. The study compares three types of avatars in jump games: human (highly

anthropomorphic), blocky (low anthropomorphic), and robot (highly anthropomorphic). The

results showed that players who played robots had a better experience. Both robots and human

avatars lead to higher avatar identification. Avatar identification significantly improves player

experience and game time (Kao, 2019). In my kart racing game, the driver and kart is low-poly

style. The avatar types of the driver and the kart can be clearly distinguished by players.

3. METHODOLOGY

This study aims to design and develop a kart racing game to study the influence of different

types of avatar on avatar identification, player performance, and experience. When players can

customize two different types of avatar and see them simultaneously in the game, how does

customization of different types of avatars affect avatar identification and game performance?

3.1 Experimental Design

The independent variable for this experiment is the combination of the customizations. There

are four different combinations of customization: (1) customizing the character and the kart, (2)

customizing the character only, (3)customizing the kart only, and (4) no customization (driving

with default character and kart). The dependent variables are player performance and avatar

identification. The performance was measured by the average time to complete each lap and the

number of collisions per lap. Avatar identification is collected through the Player Inventory

Scale questionnaire (Looy et al., 2012). Player experience was also collected to see if the game

quality was good enough because the bias caused by game quality will affect the results of the

experiment. Player experience is collected through the Player Experience Inventory (Abeele et

al., 2020) and open-ended questions.

3.2 Game Design

For the experimental purposes mentioned above, my game was divided into a customization

section and a racing section. Players were randomly assigned to four groups, and each group had

different customizations in the customization section. In the racing section, all players will play

the same two tracks and the kart performance was not affected by customization. The goal of the

racing section is to drive as fast as possible without collision with obstacles (e.g., edges of the

tracks, stones).

3.2.1 Character and Kart Customization

The customization section of the first group includes the customization of the character and the

kart. After completing the section, players will control the customized character and the

customized kart during the racing section. The second group only includes character

customization, in the racing section, the player will control the customized character and the

default kart. The third group is only provided with kart customization, in the racing section, the

player will control the default character and customized kart. The fourth group is not provided

with customization, but players will see the default character and the default kart in the

customization section, and control them in the racing section.

Figure 1. Character customization. The default character is in the middle and on

the top left are aspects to choose from that can be customized (e.g., hair style,

pants).

Figure 1 shows the default character in the middle. In order to avoid racial or racial bias, the

character has pure white skin color and has no gender characteristics. The icons on the top left

represents eight aspects of character customization: skin color, hair style, face (e.g., eyebrows,

eyes, mouth), beard, head decoration (e.g., headphones, hat), upper body attire (e.g., T-shirt,

dress), lower body attire (e.g., pants, skirt), and shoes. Players can click on these icons to open

item menus of each aspect to change the corresponding part.

The player can also choose the color for each aspect (except face) through the color pickers,

which allows players to choose their favorite color by adjusting the hue, saturation, and value.

The color picker contains two colors, the main color for a large area of the aspect and the

secondary color for a small area of the aspect. The color pickers of each aspect will open with the

item menu.

Figure 2. Hair customization. On the top left is the hair style menu including main

hair and secondary hair, and on the top right are the color pickers to choose hair

colors.

Figure 2 shows an example of the hair style menu and color pickers. On the left side of the figure

is the options for main hair and secondary hair. The main hair style is a large area of hair,

secondary hair is a small part of the hair such as braids. Each character can have a

gender-specific hairstyle through a combination of two types of hair. In figure 2, the hairstyle is

a combination of the main and secondary hairstyles in the red circles on the hair style menu. On

the right side of the figure are the color pickers. In the figure, blue was selected as the main hair

color and red as the secondary hair color.

Figure 3. Kart customization. The default kart is in the middle and on the top left

are aspects to choose from that can be customized (e.g., kart body, tires.).

Figure 3 shows the default kart in the middle. The icon on the top left represents four aspects of

kart customization: kart body, tire, spoiler, and seat. Similar to the character customization,

players can click on these icons to change the corresponding aspect of their kart and change the

color through color picker on the top right. It should be noted that the kart customization

section did not offer many options. Players can only customize the colors of the kart body and

seat, but not the type. The type of the tires is customizable, but the color of the tires is not. The

color and type of the spoiler are customizable, but there are only two options.

Figure 4. Kart Spoiler customization. On the top left is the spoiler menu and on the

top right are the color pickers to choose spoiler colors.

Figure 4 shows an example of the spoiler menu and color pickers. On the left side of the figure is

the options for spoilers. The spoiler is the one in the red circles on the spoiler menu. On the right

side of the figure are the color pickers. Blue was selected as the main spoiler color and yellow as

the secondary spoiler color in the figure.

3.2.2 Mechanic

The goal of the racing section is to go as fast as possible without collision with obstacles. The

kart loses some speed when it hits obstacles. The player can control the kart through the

keyboard. The control consists of five functions: (1) acceleration, (2) deceleration/back, (3) left

steering, (4)right steering, and (5) drift.

Figure 5. Tutorial at the beginning of level 1. Displaying basic control and how to

drift to the left and to the right.

Figure 5 shows the tutorial part at the beginning of the first level. It shows the keyboard keys

corresponding to each function mentioned above, and the way of drifting. When the player is

turning, the kart will enter the drift state by pressing the space bar on the keyboard. During

drifting, the speed decreases at the beginning due to dynamic friction of the ground. The speed

eventually stops decreasing and reaches the balance, if the player is holding the acceleration

button. During drifting, the player can still control the steering and change the direction of the

kart. When the direction of the kart is consistent with the direction of the velocity, the drift will

end.

Figure 6. Racing game play. The character and kart is in the middle. On the left top

is the lap, on the top right is time(best time, previous time and current time), and

on the bottom right is current speed.

The gameplay is a time trial, the player drives around a track for a specified number of laps as

quickly as possible. Figure 6 shows what the game looks like as it progresses. The top left of the

screen is the current lap over the total lap. The player should complete a total of three laps, and

now it’s the second lap in the figure. On the top right is the timer corresponding to the best time,

the previous lap time, and the current time. Players can see their current time and see if they are

going faster or slower than the previous lap. On the bottom right is the current speed of the kart.

The goal of the gameplay design is to create a strategic game play instead of holding the

acceleration button all the time. The game provided different ways to pass the corners, steering

or drift, and the player can control the speed of cornering to avoid collisions by releasing the

acceleration button or pressing the deceleration button. The design of the cornering speed also

serves the purpose of creating strategic game play. The optimal cornering speed is smaller than

the maximum speed of the kart, for example, in the game, the maximum speed of the car is 160

kilometers per hour, and the optimal speed for passing a typical 90 degree corner is around 100

kilometers per hour. The player can hit the wall while turning at maximum speed, or the player

can pass the corner without collision, if he or she slows down or drifts. The idea of the smaller

optimal cornering speed forces the player to control speed (slowing down or drifting) and to

choose better cornering lines, therefore it improves the strategy of the game.

3.2.3 Levels

Figure 7. Level 1 Desert (top) and Level 2 National park (bottom).

The game has two different themed levels, the first is a desert, the second is a national park.

Both levels are designed to ensure emotional differences alternating between relaxation and

tension. During driving, the field of vision also alternates between the open and narrow to

emphasize the emotional differences. For example, the top half of figure 7 is the top view and

oblique view of the first level. The player starts from the red circle, and drives anti-clockwise

along the gray track. The player encounters the ninety degrees of corner at the beginning and

feel a little nervous, then the player climbs a hill with a narrow field of vision, and after that the

player encounters a simple s-curve with a broad vision. The view of the bridge in the distance

relaxes the player’s mind.

The levels vary in difficulty, and the second level is more difficult than the first. The difficulty of

a level is reflected in four aspects: width of the track, length of the track, number of corners, and

angle of the corners. The width of the second level is narrower than that of the first level,

meaning the space of cornering is smaller. The second level is longer than the first. The player

takes longer to complete, and it is more prone to errors. The second level has more U-turns

which have larger angles and are harder to navigate. Larger angle of the corners requires more

precise control of speed and cornering line. The number of consecutive turns also changes the

optimal route.

3.3 Quantitative and Qualitative Measures

3.3.1 Player Inventory Scale

Player inventory scale (Looy et al., 2012) measures avatar identification. Originally it consisted

of 29 items including items of game (World of Warcraft) identification and group identification.

Since my game is a single player kart racing game, the items of game identification and group

identification are not applicable and removed. In this study, the questionnaire consists of 17

5-point Likert items ranging from strongly disagree to strongly agree. The 17 items measured

similarity identification, embodied presence, and wishful identification.

3.3.2 Player Experience

Player Experience Inventory (Abeele et al., 2020) measures player experience through

functional consequences and psychosocial consequences. Originally it consisted of 52 items

including items of functional consequences (ease of control, progress feedback, audiovisual

appeal, goals and rules, and challenge) and psychosocial consequences (mastery, curiosity,

immersion, autonomy). In this study, the questionnaire consists of ten 5-point Likert items

ranging from strongly disagree to strongly agree. First five items measured all the aspects of

psychosocial consequences, and the rest measured all the aspects of functional consequences.

3.3.3 Performance

Performance in this study refers to the accuracy of driving. I collected the in-game data

measuring the time of each lap and the number of hitting the obstacles of each lap. It’s the same

variables that Cassidy and Macdonald measured in their driving performance experiment.

Inaccuracy was measured by recording one mark each time the player collided with an obstacle

or a surrounding barrier as a measure of task vigilance and control (Cassidy & Macdonald,

2010).

3.4 Participants

Because of the coronavirus, participants participated in the experiment remotely. I recruited

participants through social platforms, and uploaded the recruitment poster to social platforms,

such as Discord, Slack, Twitter, WeChat, etc.

A total of 44 (30 male, 14 female) participants were recruited through WeChat and Discord.

Participants were between the ages of 20 and 50 years old, and most of them were from China.

More than half of the participants play video games everyday, but most participants did not

often play racing games (once or twice a year).

The subjects were randomly assigned into four groups of about 10 people each. All participants

will be treated in accordance with the IRB protocol. Experiments can only be carried out with

the consent of the participants. During the experiment, participants can quit at any time.

3.5 Research Procedure

The experiment consists of two parts: playing the game and filling out the questionnaire. There

were four versions of the game. In the first three versions, players are able to customize their

avatars, and each group has a different type of avatar to customize (character and kart, character

only, kart only). The other version serves as a control group, allowing players to use only the

default character and kart. Each player will drive three laps in each track. The game should last

about 10 - 15 minutes. After completing the game, the participants completed a long

questionnaire with four parts: (1) Player Inventory Scale (Looy et al., 2012), (2) Player

Experience Inventory (Abeele et al., 2020), (3) demographic information, (4) open-ended

questions about evaluations of the game . When the participants have finished answering the

questions, the experiment is over.

3.6 Data Collection and Analysis

I collected in-game data from all players to measure driving performance, including the average

time of a lap for each level, and the average number of collisions per lap for each level. I

calculated the Player Inventory Scale (Looy et al., 2012) average score for each player and the

Player Experience Inventory (Abeele et al., 2020) score for each Player. In addition, one way

ANOVA test was conducted for the average time per lap of each level, the number of collisions

per lap of each level, the average score of avatar identification and the average score of player

experience. When the results showed significant difference, I did post hoc analysis to determine

which group is significantly different. Four open-ended questions about evaluations and

motivations of the character customization and the kart customization are provided to the

players in the group that had such functions. An open-ended question about the evaluation of

the racing gameplay was provided to all the groups. I summarized the keywords of the open

question through coding, and analyzed the similarities and differences of the answers from both

general and inter-group perspectives.

4. RESULTS

4.1 Avatar Identification

Figure 8. Avatar Identification of Each Group. Group 1: full customization, Group

2: character customization only, Group3: kart customization only, Group 4: no

customization.

Figure 8 shows the distribution of avatar identification of four groups: (1) Group 1: full

customization, (2) Group 2: character customization only, (3) Group3: kart customization only,

(4) Group 4: no customization. I calculated the mean value and standard deviation of each

group. Table 1 below shows the mean and standard deviation of avatar identification of each

group. The results show that regarding avatar identification the Character Only condition (M =

4.02, SD = 0.92) has the highest score and the Kart Only (M = 2.71, SD = 0.93) the lowest. The

other two conditions, Full Customization and No Customization, are strikingly similar and fall

right in the middle of the Character Only and Kart Only scores.

Group

Full Customization

3.19

0.95

Character Only

4.02

0.92

Kart Only

2.71

0.93

No Customization

3.20

0.71

Table 1. Mean Avatar Identification of Each Group.

I used a one-way ANOVA test to check whether there was a significant difference in avatar

identification between the groups. The result indicates that there is a significant difference, F(3,

40) = 4.13, p = .012.

Figure 9. Differences in mean of avatar identification.

I conducted a post hoc analysis via the Turkey Test to find out which group is different. Figure 9

presents the differences in mean of avatar identification. Group 2 (character customization

only) and group 3 (kart customization only) were significantly different (p=.0007). Group 3 had

significantly lower avatar identification than group 2. There was no significant difference

between the other groups.

4.2 Performance

Figure 10. Average time and average hits of each group.

In-game data records the player performance through the average time of finishing each lap and

the average number of collisions per lap. Figure 10 shows the distribution of the time and

collisions of each level in four conditions. The results show that regarding average lap time of

level 1 the Kart Only condition (M = 54.47, SD = 6.63) has the lowest time and the Character

Only (M = 70.42, SD = 27.40) the highest. The average time of the Full Customization

condition(M = 59.79, SD = 10.89) is close to the Kart Only condition, and the average time of the

No customization condition(M = 67.90, SD = 30.67) is close to the Character Only condition.

Regarding the average lap time of level 2 Kart Only condition (M = 81.27, SD = 8.79) has the

lowest time and the No Customization (M = 105.49, SD = 51.02) the highest. The average time of

Full Customization condition(M = 84.89, SD = 10.29) is close to the Kart Only condition, and

the average time of the Character Only condition(M = 102.99, SD = 24.65) is close to the No

Customization condition.

Regarding the average number of collisions of level 1 No Customization condition (M = 8.58, SD

= 3.64) has the lowest and the Character Only (M = 18.45, SD = 7.68) the highest. The Kart Only

condition(M = 8.63, SD = 5.21) is close to the No Customization condition, and the Full

Customization condition(M = 11.21, SD = 12.51) is in the middle between the No Customization

and the Character Only collisions.

Regarding the average number of collisions of level 2 Kart Only condition (M = 13.89, SD =

9.45) has the lowest and the Character Only (M = 33.21, SD = 19.85) the highest. The Full

Customization and the No Customization is in the middle between the Kart Only and the

Character only collisions.

Table 2 below presents the mean and standard deviation of the average time for each group, and

the mean and standard deviation of the average number of collisions for each group.

Group

Level 1 Lap

time (s)

Level 2 Lap

time (s)

Level 1 Lap

Hits

Level 2 lap

Hits

Full Customization

59.79

84.89

11.21

19.24

10.89

10.29

12.51

16.50

Character Only

70.42

102.99

18.45

33.21

27.40

24.65

7.68

19.85

Kart Only

54.47

81.27

8.63

13.89

6.63

8.79

5.21

9.45

No Customization

67.90

105.49

8.58

16.33

30.67

51.02

3.64

13.79

Table 2. Mean Performance of Each Group.

One-way ANOVA tests were conducted to check whether there was a significant difference in

time and collisions between the groups. There was no significant difference in the average time

between the groups, (Level 1: F(3, 39) = 1.10, p = .36, Level 2: F(3, 39) = 1.73, p= .18), but there

was significant difference in the average collision between the groups (Level 1: F(3, 39) = 3.64, p

= .02, Level 2: F(3, 39) = 3.28, p= .03).

Figure 11. Difference in mean average collision of level 1.

Figure 12. Difference in mean average collision of level 2.

A post hoc analysis via the Turkey Test was conducted to find out which group is different.

Figure 11 and Figure 12 presents the differences in mean of average collision of level 1 and level

For the average collision in level 1, Group 2 (character customization only) and group 3 (kart

customization only) were significantly different (p=.04). Group 2 (character customization only)

and group 4 (no customization) were also significantly different (p=.02).Group 2 had

significantly more collisions than group 3 and group 4 in level 1. There was no significant

difference between the other groups.

For the average collision in level 2, Group 2 (character customization only) and group 3 (kart

customization only) were significantly different (p=.04). Group 2 had significantly more

collisions than group 3 in level 2. There was no significant difference between the other groups.

4.3 Player Experience

Figure 13. Player experience of each group.

Figure 13 shows the distribution of player experience of four groups. The results show that

regarding player experience the Kart Only condition (M = 3.67, SD = 0.58) has the lowest time

and the Character Only (M = 4.20, SD = 0.62) the highest. The Full Customization condition(M

= 4.04, SD = 0.41) is close to the Character Only condition, and the No customization

condition(M = 3.78, SD = 0.37) is close to the Kart Only condition.

Table 3 below shows the mean and standard deviation of player experience of each group.

Group

Full Customization

4.04

0.41

Character Only

4.20

0.62

Kart Only

3.67

0.58

No Customization

3.78

0.37

Table 3. Mean Player Experience of Each Group.

One-way ANOVA tests were conducted to check whether there was a significant difference in

player experience between the groups. The result (p=.07) suggests that there was no significant

difference in player experience between the groups, F(3, 40) = 2.55, p = .07.

4.4 Type of Player in Each Group

Figure 14. Game frequency of each group.

The questionnaire also collected how often players played the game and how often they played

racing games. The answers are numbers from 1 to 5 (1: never, 2: rarely, 3: sometimes, 4: often, 5:

frequently). Figure 14 shows how often each group played games.

In the group 1 (full customization), the half of players played games frequently (54.55%). The

proportion of players who often play games and the players who rarely play games are the same

(18.18%). The proportion of players who sometimes played games is small (9.09%).

In the group 2 (character customization only), The proportions of players who frequently play

games, players who often play games, players who sometimes play games are the same (27.27%).

The proportions of players who rarely played games and players who never played games are the

same (9.09%).

In the group 3 (kart customization only), the majority of players play games frequently (80%).

The rest are the players who often play games (10%) and players who sometimes play games

(10%).

In the group 4 (no customization), the proportions of players who frequently play games and the

players who sometimes play games are the same (33.33%). The proportions of players who often

played games and players who rarely play games are the same (16.67%).

Figure 15. Racing game frequency of each group.

Figure 15 shows how often each group played racing games.

In the group 1 (full customization), the majority of players rarely played racing games (63.64%).

27.27% of the players sometimes played racing games. Few players often played racing games

(9.09%).

In the group 2 (character customization only), The proportions of players who sometimes play

racing games, players who rarely play racing games are the same (36.36%). Some players often

played racing games (18.18%). Few players frequently played racing games (9.09%).

In the group 3 (kart customization only), the majority of players rarely play racing games (70%).

The rest are players who sometimes play games (20%) and the players who often play games

(10%).

In the group 4 (no customization), the half of players rarely play racing games (50%). Some

players sometimes played racing games (33.33%). The proportions of players who often play and

the players who never play are the same (8.33%).

4.5 Open-ended questions

There are a total of five open-ended questions. These questions asked participants (1) their

thoughts about the character customization, (2) thoughts about the car customization, (3)

motivations of customizing their characters and (4) motivations of customizing karts, and (5)

thoughts about gameplay. If a group has no content about character customization or car

customization, the related questions were hidden. The group 1(full customization) was asked all

the questions. The group 2 (character customization only) had no questions about car

customization, and the group 3 (kart customization only) had no questions about character

customization. The group 4 (no customization) had no questions about customization. Based on

the responses of the players, I summarized the following results.

4.5.1 Thoughts about Character Customization

Players in group 1 (full customization) and group 2 (character customization only) were asked

about the character customization. Feedbacks about character customization can be divided into

positive and negative.

In the positive feedback, players in both groups mentioned that character customization was fun

and provided enough freedom to create their favorite characters. One of the participants wrote,

“I like this customization part very much, because it includes the customization of each part of

the role, and I can modify the color of each part in detail, which makes me feel very free, and I

can make the role I like” (p10, group 1). Also a participant in group 2 mentioned, “amazing. It

provides so many different faces and clothes” (p4, group 2).

In negative thoughts, players in both groups mentioned the need for more details to reflect the

character's characteristics. “It's fine, but it annoys me that only part of the color of the cloth is

allowed to change”(p3, group 1). “I suggest a more varied style of clothes” (p6, group 2).

In group 1 (full customization), there are 7 positive feedbacks and 4 negative feedbacks. Players

(n=3) complained that the automatic rotation of the character was inconvenient and they

needed to control the rotation of the character themselves. "It is good, but if I can control the

perspective, It can be better"(p11, group1).

In group 2 (character customization only), there are 5 positive feedbacks and 5 negative

feedbacks. All the negative feedback was about the need for more details to reflect the

character's characteristics, such as different body shapes, more colors, and more clothes.

4.5.2 Motivation of Character Customization

Players in group 1 (full customization) and group 2 (character customization only) were also

asked about the motivation of customizing their character.

In general, players in both groups mentioned the motivation was for fun. Other motivations are

to create a character that resembles themselves, or a character that mimics another game or

animation. There is one participant in each group answered with no motivation. “No particular

reason, just the game asking me to do it” (p6, group 1).

In group 1 (full customization), two of the players’ motivation is creating an ideal character and

being cool. “My motivation was to create an ideal image of myself: a ponytail with a cheeky

expression, a blue and white sweatshirt, a blue and white dress, cool shoes, and headphones like

Beats. All in all, I want to make myself look cool in the game” (p4, group 1).

In group 2 (character customization only), players (n=3) mentioned the motivation to express

their mood and feelings. “Customize the character based on my mood and feelings at the time”

(p7, group 2).

4.5.3 Thoughts about Kart Customization of Group 1 and Group 3

Players in group 1 (full customization) and group 3 (kart customization only) were asked about

the kart customization. Feedbacks about kart customization can be divided into positive and

negative.

In the positive feedback, players in both groups mentioned that the kart customization is good.

63.64% (n=7) of players in group 1 (full customization) mentioned think it’s good. One player

said, “Great! I like choosing my own Kart” (p7, g1). 70% (n=7) of players in group 3 (kart

customization only) also mentioned this. “Providing the color customization rather than

switching outlook is a smart move. It will reduce the art work but make everyone unique"(p6,

g3).

In negative thoughts, players in both groups mentioned that the kart customization is limited.

63.64% (n=7) of players in group 1 (full customization) mentioned this. "Good and cool. If it can

be more varied, it will be better" (p5, g1).

20% (n=2) of players in group 3 (kart customization only) mentioned this. “I hope there are

more kinds of vehicle appearance, so far I can only modify its color, which is not comprehensive

enough"(p10, g1).

In addition, 9.09% (n=1) of players in group 1 (full customization) also mentioned the need to

add more game features, “Not bad, but it seems to be just a cosmetic change, hopefully these

accessories will bring changes in things like speed, handling” (p3,g1).

In group 3 (kart customization only), 10% (n=1) of players also mentioned that karting

customization allowed him to express himself, "I like that part. It gives me some space to use my

art talent" (p7, g3). Another player mentioned that he was not interested in kart customization.

"It is cool to have but as a player I don't care too much about aesthetics at least in this game"

(p3, g3).

4.5.4 Motivation of Kart Customization of Group 1 and Group 3

Players in group 1 (full customization) and group 3 (kart customization only) were also asked

about the motivation of customizing their kart.

Overall, there were two motivations mentioned most by both groups. The first motivation is that

players want to make their karts cool and recognizable. 36.36% (n=4) of group 1 (full

customization) and 50% (n=5) of group 3 (kart customization only) mentioned this. One player

mentioned that "Want it to be cool and distinctive" (p2, g1), and another player mentioned that

"I wanted it to look cool" (p6, g3). The second motivation is to express one's preferences.

18.18%(n=2) of the first group and 40% (n=4) of the third group mentioned this. "I want to

customize to find my favorite kart by changing color, accessories, etc." (p10, g1). Player in group

3 mentioned that he wants to "make it unique and express my color preference" (p5, g3).

In addition, 18.18% (n=2) of players in group 1 (full customization) mentioned that they are

trying to build the karts in other games or animations. One player said that his motivation is “to

respect the Mario” (p9,g1), and one player’s motivation is for fun (p8, g1).

In group 3 (kart customization only), one of the players does not show a motivation. “I did not

really spend much time on the customization. Just picked some random option if they look

good” (p3, g3).

4.5.5 Thoughts about the Game-play of each Group

Players in each group were asked for the feedback on the gameplay, in order to see whether the

gameplay caused bias.

In the positive feedback, players in all groups mentioned that the driving control is easy. 18.18%

(n=2) of players in group 1 (full customization) mentioned the driving is easy. One player said,

“The controls are easy to understand and the drift feels very similar to the racing games I've

played before” (p10, g1). 27.27% (n=3) of players in group 2 (character customization only) also

mentioned this. “simple and easy to control"(p5, g2). 10% (n=1) of players in group 3 (kart

customization only) mentioned this. “It is fluent and intuitive like other racing games. I like the

track design" (p3, g3). 33.33% (n=4) of players in group 4 (no customization) mentioned this.

“Overall it is very easy to control, the graphic is distinctive, Gameplay: 8/10"(p1, g4).

In the negative feedback, players in all groups mentioned that the kart is difficult to control.

18.18% (n=2) of players in group 1 (full customization) mentioned the driving is hard. One

player said, “It is hard to control the change of direction, since the power of the directing is

weak” (p9, g1). 18.18% (n=2) of players in group 2 (character customization only) also

mentioned that they did not know how to drift properly. “Gruelling, because I have played some

similar racing game several years ago, I still have a brief understanding of how to use drift. But I

can't know how to use drift properly, if I press the space button, it will keep drifting for a long

time and hit the wall ultimately"(p4, g2). 30% (n=3) of players in group 3 (kart customization

only) mentioned that the steering is heavy. “It is hard to control without drifting" (p6, g3). 25%

(n=3) of players in group 4 (no customization) mentioned this. “Kind of fun but hard to control,

it is not as easy as driving in reality. "(p11, g4).

In addition, 63.64% (n=7) of players in group 1 (full customization) mentioned the game play is

cool, “Cool, it is the feel of Mario Kart” (p3, g1). One player mentioned that the camera effect

should be polished to show the change of the speed. (p8, g1)

In group 2 (character customization only), 27.27% (n=3) of players mentioned that the game

play is good, "quite good, heavy understeer, which makes drifting meaningful" (p2, g2). 27.27%

(n=3) of players also mentioned that the game needed more challenge, "I hope there are some

obstacles on the track" (p9, g2).

In group 3 (kart customization only), 20% (n=2) of players mentioned that the game is good in

aspects like environmental color and music, "I like the music of the game" (p7, g3). 10% (n=1) of

players mentioned adding multiplayer futures, " If i can race with others, it would be more

interesting" (p10, g3).

In group 4 (no customization), 8% (n=1) of players mentioned that the collision physics is not

real enough, "the feeling of hitting border is not realistic enough: when you hit border, one side

of the car model immediately sides the racing track, hence a fast recovery and sometimes

reversed directions. Overall the gameplay is great though, smooth and no bugs. " (p7, g4). 25%

(n=3) of players also mentioned that the gameplay should be more varied, "the gameplay can be

varied and competitive" (p3, g4).

4.6 Summary

In terms of the avatar identification, group 2 (character customization only) had a significantly

higher score than group 3 (kart customization only). There were no significant differences

between group 1(full customization), group 3 (kart customization only) and group 4 (no

customization).

In terms of performance, The average time per lap of the four groups did not differ significantly

in all the levels. There was no significant difference in the number of collisions between group 1

(full customization), group 3 (kart customization only) and group 4 (no customization) in all the

levels. In the first level, the number of collisions in group 2 (character customization only) was

significantly higher than that in groups 3(kart customization only) and group 4 (no

customization). In the second level, the number of collisions in group 2 (character

customization only) was significantly higher than group 3 (kart customization only).

In terms of experience, there was no significant difference between the four groups.

5. DISCUSSION

The type of customizations is indicated by the abbreviation in this section:

1. Group 1 (FC) – full customization

2. Group 2 (CO) – character customization only

3. Group 3 (KO) – kart customization only

4. Group 4 (NC) – no customization

5.1 The Impact of Customization on Avatar Identification

There was a significant difference in avatar identification between the group 2 (CO) and the

group 3 (KO). Based on previous experiments, the possible reason is that players generally have

low avatar identification with low likeness avatars. In previous educational games with likeness

avatar and shape avatar, player identification with shapes is low. Players detailing their shape

avatars are more likely to use impersonal pronouns (e.g., it, it’s, those) and articles (e.g., a, an,

the), and less likely to use first person singular (e.g., I, me, my). In particular, likeness avatars

scored higher on those dimensions that relate to player avatar identification (Kao & Harrell,

2015). In the previous platform jumping game, human (high anthropomorphism) and Robot

(high anthropomorphism) participants have higher avatar identification than Block-like (low

anthropomorphism) participants (Kao, 2019). The results of this experiment were consistent

with those of previous experiments. Players in the group 3 (KO) did not identify with their karts

as strongly as players in the group 2 (CO) did with their characters.

The reason the difference in avatar identification was not significant between the other groups

could be that the game had not been played for long enough to generate enough identity

differences. In a customization part of the previous experiment, the researchers wrote that,

players spent considerable time customizing their character in the first session and their

reported engagement was strongly related to their identification in the last session. This result

implies that as players spend more time in the game with their customized avatars, their

identification increases (Turkay & Kinzer, 2014). Players need more time to customize and play

with their characters to create a greater sense of identity. The total time of this game in this

study is 10-15 minutes, which may not be long enough to make a significant difference, but on

average player identification, group 2 (CO) had a higher mean than group 1 (FC) and group 4

(NC), and group 3 (KO) had a lower mean value of avatar identification than group 1 (FC) and

group 4 (NC).

The reason why group 1 (FC) and group 4 (NC) had close mean values might be that the two

types of customization were combined in group 1, and the diversity difference between the two

customizations affected the results. Compared with the 63.64% (n = 7) players in the group 1

(FC), only 20% (n = 2) of players in group 3 (KO) mentioned that the kart customization lacks

diversity. Players in group 3 (KO) did not care as much as group 1 (FC) players how many

customization options there were. It suggested that the diversity of the two types of

customization contrasts in the group 1 (FC), when there are more options in character

customization than kart customization. The feeling of the lack of diversity had an impact on the

avatar identification of group 1 (FC), so it is on the same level of group 4 (NC).

5.2 The Impact of Customization on Performance

The group 2 (CO) had significantly more collisions than the group 3 (KO). This result may be

influenced by the avatar identification and diversity of customization. Group 2 (CO) had

significantly higher avatar identification than group 3 (KO). In my game, character

customization has more aspects and options to customize, and players spend more time on

customization. The kart takes less time to customize with fewer aspects and options. According

to previous educational experiments, high avatar identification has a negative impact on

performance. Bowman et. al suggest that avatars more like “objects” cause players to focus more

on in-game mechanics and challenges (“pleasures of control”). Failure in the game (which is

almost guaranteed, the mean number of attempts in the first level was 8.4), may be especially

thwarting when the character failing is you (Kao & Harrell, 2015). This can be used to explain

the results here. When players hit the obstacles on the track, the higher identification of

characters brought greater frustration. Players of group 2 (CO) had a higher avatar identification

to their characters, and their performance was affected by greater frustration. Players in group 3

(KO) used the default character and had a short customization time, so they could focus more on

the mechanics and challenges of the game. Simple default characters and less customization

features of the kart improved the player's performance.

Another explanation is that players with Action Video Game (AVG) experience performed better

in driving games. Although 70% of Group 3 (KO) players rarely play racing games, only 36.36%

of Group 2 (CO) players rarely play racing games, and the rest of Group 2 (CO) players play

racing game sometimes (36.36%), often (18.18%), and frequently (9.09%), 80% of Group 3 (KO)

players play video games frequently. Only 27.27% of players in the group 2 (CO) played games

frequently. Based on previous experiments of AVG experience and driving performance,

long-term 3D game players performed better by driving less out of the lane. Participants with

higher AVG experience were better drivers in the simulation across all drives, which indicated

evidence of task related transfer (Rupp et al., 2016). Players in group 3 (KO) were likely to have

more 3D gaming experience on a regular basis, leading to improved performance and fewer

collisions.

Gender differences may also play a role here. The group 1 (FC) and group 4 (NC) were nearly

50% male and 50% female, but 63.64% of the group 2 (CO) was women (n=7, N=11) and all

players in the group 3 (KO) were male. According to previous experiments, male workers are

better at spatial rotation than female workers in multitasking situations. In two experiments,

participants completed a multitasking session with four gender-fair monitoring tasks and

separate tasks measuring executive functioning (working memory updating) and spatial ability

(mental rotation). In both experiments, males outperformed females in monitoring accuracy

(Mäntylä, 2013). Whether or not my game should be considered multitasking is up in the air.

More research is needed on the effects of gender differences.

5.3 The Impact of Customization on Player Experience

As can be seen from the above data, the mean value of player experience of group 1 (FC) and

group 2 (CO) is higher than that of group 3 (KO) and group 4 (NC), which may be caused by the

similarity of the avatar. Group 1 (FC) and group 2 (CO) have character customization, so the

character similarity is higher. In a previous experiment, Kao found that avatars with higher

anthropomorphism led to higher player experience (Kao, 2019). The distribution of average

player experience is consistent with the previous experiment. However, differences in player

experience can be influenced by the amount of time played (Turkay & Kinzer, 2014). The players

in this study might not have been playing long enough to make a significant difference in player

experience.

5.2 Limitation

This experiment has the following limitations. The distinction between avatar and character is

still controversial. This study used the definition of Klimmt et al., is a temporary alteration of

media users’ self-concept through adoption of perceived characteristics of a media person

(Klimmt et al., 2009).

The overall playing time is short, which probably is not enough for players to develop a sense of

identity with the characters. The lack of diversity for character and kart customization leads to

short time spent in the customization interface. This can have a negative impact on identity and

experience.

In the epidemic environment, the online experiment environment can not be strictly controlled.

There was player feedback mentioned that he did not hear the background music, in fact the

player did not turn on the volume. Previous studies have shown that music affects arousal and

thus driving performance (Ünal et al., 2013). Online experiments were unable to determine how

many players heard the music, which could lead to errors.

The game frequency and gender of players were not guaranteed when randomly assigned. Those

were two factors that may have influenced my experimental results. In terms of the frequency of

playing video games, group 3 (KO) is generally higher than group 2 (CO). This might affect the

performance. In addition, the group 2(CO) had seven or 11 women, while the group 3 (KO) was

all men. There might be a potential gender difference that could lead to an error.

5.3 Future work

I should provide more customization details, extend customization time, and provide more

levels so that players have more time to show their identity with their characters. Also I should

provide more customization content, such as the character characteristics, the diversity of the

kart customization.

The offline experiment environment can strictly control the test conditions, and minimize the

impact of the differences in the experiment environment. I should make sure that the players’

experiment conditions are similar to each other in order to get more accurate results.

Finally, I want to make sure the game frequency and gender are close when randomly assigned

to reduce the possibility of bias caused by gender or game frequency.

6. CONCLUSION

This experiment fills in the blank of the research on the type of customization in kart racing

games and provides a reference for the development of racing games.

In my game, the difference in avatar identification between group 1 (FC) and group 4 (NC) was

not significant, but the character customization only group was significantly higher than the kart

customization only. Assuming that the result is correct, this means that in a kart racing game,

customization of the kart decreases avatar identification, and customization of the character

increases avatar identification. The combination of customization of the character and the kart

offsets this effect, and avatar identification in group 1 (FC) becomes similar to the situation

without customization in group 4 (NC). In kart racing games, customizing a character similar to

the player improves the player's experience, but more similar characters transmit greater

frustration when the player meets failures (e.g., hits obstacles), thus reducing the driving

performance.

What this means is that in future kart racing games, developers can focus more on character

customization than on kart customization when developing customization features. Character

customization increases the player's character identity, which in turn improves the player's

experience. At the same time, in order to avoid more frustration caused by character similarity,

the player should be given less anthropomorphic feedback when they are frustrated (e.g.,

collisions), such as cartoonish representation of hitting.

REFERENCES

Looy, J. V., Courtois, C., Vocht, M. D., & Marez, L. D. (2012). Player Identification in Online Games:

Validation of a Scale for Measuring Identification in MMOGs. Media Psychology, 15(2),

197–221. https://doi.org/10.1080/15213269.2012.674917

Kao, D., & Harrell, D. F. (2018). The Effects of Badges and Avatar Identification on Play and Making

in Educational Games. Proceedings of the 2018 CHI Conference on Human Factors in

Computing Systems, 1–19. https://doi.org/10.1145/3173574.3174174

Birk, M. V., Atkins, C., Bowey, J. T., & Mandryk, R. L. (2016). Fostering Intrinsic Motivation through

Avatar Identification in Digital Games. Proceedings of the 2016 CHI Conference on Human

Factors in Computing Systems, 2982–2995. https://doi.org/10.1145/2858036.2858062

Mancini, T., Imperato, C., & Sibilla, F. (2019). Does avatar’s character and emotional bond expose to

gaming addiction? Two studies on virtual self-discrepancy, avatar identification and gaming

addiction in massively multiplayer online role-playing game players. Computers in Human

Behavior, 92, 297–305. https://doi.org/10.1016/j.chb.2018.11.007

Li, K., Nguyen, H. V., Cheng, T. C. E., & Teng, C.-I. (2018). How do avatar characteristics affect avatar

friendliness and online gamer loyalty? Perspective of the theory of embodied cognition. Internet

Research, 28(4), 1103–1121. https://doi.org/10.1108/IntR-06-2017-0246

Biocca, F. (2014). Connected to My Avatar: In G. Meiselwitz (Ed.), Social Computing and Social

Media (pp. 421–429). Springer International Publishing.

https://doi.org/10.1007/978-3-319-07632-4_40

Mancini, T., & Sibilla, F. (2017). Offline personality and avatar customisation. Discrepancy profiles

and avatar identification in a sample of MMORPG players. Computers in Human Behavior, 69,

275–283. https://doi.org/10.1016/j.chb.2016.12.031

Kao, D., & Harrell, D. F. (2015). Toward avatar models to enhance performance and engagement in

educational games. 2015 IEEE Conference on Computational Intelligence and Games (CIG),

246–253. https://doi.org/10.1109/CIG.2015.7317959

Kao, D. (2019). The effects of anthropomorphic avatars vs. Non-anthropomorphic avatars in a

jumping game. Proceedings of the 14th International Conference on the Foundations of Digital

Games, 1–5. https://doi.org/10.1145/3337722.3341829

Liao, G.-Y., Cheng, T. C. E., & Teng, C.-I. (2019). How do avatar attractiveness and customization

impact online gamers’ flow and loyalty? Internet Research, 29(2), 349–366.

https://doi.org/10.1108/IntR-11-2017-0463

Turkay, S., & Kinzer, C. K. (2014, January 1). The Effects of Avatar-Based Customization on Player

Identification. International Journal of Gaming and Computer-Mediated Simulations

(IJGCMS). www.igi-global.com/article/the-effects-of-avatar/115575

Schmierbach, M., Limperos, A. M., & Woolley, J. K. (2012). Feeling the Need for (Personalized)

Speed: How Natural Controls and Customization Contribute to Enjoyment of a Racing Game

Through Enhanced Immersion. Cyberpsychology, Behavior, and Social Networking, 15(7),

364–369. https://doi.org/10.1089/cyber.2012.0025

Ratan, R., & Sah, Y. J. (2015). Leveling up on stereotype threat: The role of avatar customization and

avatar embodiment. Computers in Human Behavior, 50, 367–374.

https://doi.org/10.1016/j.chb.2015.04.010

Dolgov, I., Graves, W. J., Nearents, M. R., Schwark, J. D., & Brooks Volkman, C. (2014). Effects of

cooperative gaming and avatar customization on subsequent spontaneous helping behavior.

Computers in Human Behavior, 33, 49–55. https://doi.org/10.1016/j.chb.2013.12.028

Li, D. D., Liau, A. K., & Khoo, A. (2013). Player–Avatar Identification in video gaming:

Concept and measurement. Computers in Human Behavior, 29(1), 257–263.

https://doi.org/10.1016/j.chb.2012.09.002

Hudson, I., & Hurter, J. (2016). Avatar Types Matter: Review of Avatar Literature for Performance

Purposes. In S. Lackey & R. Shumaker (Eds.), Virtual, Augmented and Mixed Reality (pp.

14–21). Springer International Publishing. https://doi.org/10.1007/978-3-319-39907-2_2

Abeele, V. V., Spiel, K., Nacke, L., Johnson, D., & Gerling, K. (2020). Development and

validation of the player experience inventory: A scale to measure player experiences at the

level of functional and psychosocial consequences. International Journal of

Human-Computer Studies, 135, 102370. https://doi.org/10.1016/j.ijhcs.2019.102370

Rupp, M. A., McConnell, D. S., & Smither, J. A. (2016). Examining the Relationship

Between Action Video Game Experience and Performance in a Distracted Driving Task.

Current Psychology, 35(4), 527–539. https://doi.org/10.1007/s12144-015-9318-x

Mäntylä, T. (2013). Gender Differences in Multitasking Reflect Spatial Ability. Psychological

Science, 24(4), 514–520. https://doi.org/10.1177/0956797612459660

Ünal, A. B., de Waard, D., Epstude, K., & Steg, L. (2013). Driving with music: Effects on

arousal and performance. Transportation Research Part F: Traffic Psychology and

Behaviour, 21, 52–65. https://doi.org/10.1016/j.trf.2013.09.004

Klimmt, C., Hefner, D., & Vorderer, P. (2009). The Video Game Experience as “True”

Identification: A Theory of Enjoyable Alterations of Players’ Self-Perception.

Communication Theory, 19(4), 351–373.

https://doi.org/10.1111/j.1468-2885.2009.01347.x

Appendix A

Part 1: Avatar Identification

Player Inventory Scale(PIS)

5-point Likert items ranging from strongly disagree to strongly agree which consists of three

second order factors:

1) Similarity identification (e.g., My character is similar to me);

2) Embodied identification (e.g., In the game, it is as if I become one with my character);

3) Wishful identification (e.g., I would like to be more like my character).

Item codes

1. My character is similar to me

2. I resemble my character

3. My character resembles me

4. I identify with my character

5. My character is like me in many ways

6. My character is an extension of myself

7. In the game, it is as if I become one with my character

8. I feel like I am inside my character when playing

9. When I am playing, it feels as if I am my character

10. When I am playing I am transported into my character

11. When playing, it feels as if my character's body becomes my own

12. In the game, it is as if I act directly through my character

13. I would like to be more like my character

14. If I could become like my character, I would

15. My character is an example to me

16. My character is a better me

17. My character has characteristics that I would like to have

Note: Adapted from (Looy et al., 2012)

Part 2: Player Experience

Evaluate the Player Experience. You can sum/average all the items to get a total score on

Player Experience.

1. Playing the game was meaningful to me. [MEANING]

2. I felt I was good at playing this game. [MASTERY]

3. I was fully focused on the game. [IMMERSION]

4. I felt free to play the game in my own way. [AUTONOMY]

5. I wanted to explore how the game evolved. [CURIOSITY]

6. I thought the game was easy to control. [EASE OF CONTROL]

7. The challenges in the game were at the right level of difficulty for me. [CHALLENGE]

8. The game informed me of my progress in the game. [PROGRESS FEEDBACK]

9. I liked the look and feel of the game. [AUDIOVISUAL APPEAL]

10. The goals of the game were clear to me. [GOALS AND RULES]

Part 3: Demographics

including gameplay preferences (as to know whether they play racing games)

What is your gender?

● woman

● man

● non-binary

● "prefer not to answer"

● "prefer to self-describe"

What race / ethnicity do you most identify with?

● American Indian or Alaska Native

● Asian

● South Asian

● Black or African American

● Hispanic, Latino or Spanish Origin

● Middle Eastern or North African

● Native Hawaiian or Other Pacific Islander

● White

● Some other race, ethnicity or origin, please specify:

● Prefer not answer

What is your age in years?

● please specify

● prefer not to answer

How frequently do you play video games for enjoyment?

1- Never

2- Rarely (one or two times a year)

3- Sometimes (one or two times a month)

4- Often (one or two times a week)

5- Frequently (daily)

(If last question's answer is not 1)

How frequently do you play racing games for enjoyment?

1- Never

2- Rarely (one or two times a year)

3- Sometimes (one or two times a month)

4- Often (one or two times a week)

5- Frequently (daily)

Open-ended questions

What do you think of the character customization?

What do you think of the kart customization?

What was your motivation behind customizing your character?

What was your motivation behind customizing your kart?

What do you think of the racing gameplay?