Flow to the Beat

32  Download (0)

Full text

(1)

Constantin Freiherr von Münchhausen

Student Number: 13292501 Entertainment Communication

constantin.freiherr.von.munchhausen@student.uva.nl

Graduate School of Communication University of Amsterdam Academic year 2021–2022

Master Thesis Entertainment Communication

Flow to the Beat

How Game Difficulty and Music–Movement Synchrony in Virtual Reality Gaming influence Game Flow

Amsterdam, July 1st, 2022 Word Count: 7.497

(2)

Abstract

Gaming in VR has implications that are beneficial for the optimal experience and enjoyment, described as game flow due to higher immersion, presence, and prominent

physical movement. The current study investigates how difficulty in relation to a player’s skill influences game flow, and which role rhythmically synchronous or asynchronous movements to music plays in that context. An experiment was conducted among 202 university students playing the rhythm game Beat Saber in too easy, matching, or too hard difficulties with either synchrony or asynchrony between music and movements. Results showed that matching skill and difficulty influences flow as expected. Matching difficulty and players’ skills led to higher levels of flow, whereas unmatched conditions led to a decrease in flow. Higher states of flow were in turn cause for higher enjoyment and lower levels of boredom. Synchrony however did not prove to be an integral part of the manipulation of flow, asynchrony between music and movement proved to be more beneficial to flow. The highest state of flow could be reached in an optimal condition, with a combination of synchronous movements and

matching difficulty.

(3)

Introduction

Gaming in Virtual Reality (VR) is a current gaming trend, with VR Headset owners around the globe purchasing devices primarily for gaming purposes due to gaming

experiences that are more enjoyable. In comparison to traditional, non–VR gaming settings, VR games increase immersion and presence due to physical movements, interaction, and simulated alternate environments (Yoo et al., 2018). This benefits an increase in immersion and presence which makes gaming experiences more enjoyable for players (Sweetser &

Wyeth, 2005). Apart from that, there are multiple identified factors contributing to a positive gaming experience, such as convincing graphics, intuitive controls, matching skill and difficulty levels as well as the presence of music (Faric et al., 2019a). The latter two contributors are in line with the flow theory of Csikszentmihalyi (2009), since both are important factors to enjoyment.

Flow describes an optimal psychological state in which a user is fully immersed in an activity and draws a high level of focus, interest, and enjoyment from it (Bodzin et al., 2021).

Users find themselves in a situation of deep concentration and control when entering the flow state (Rutrecht et al., 2021). Flow is a vital component of gaming research (Fang et al., 2013).

Transferring the concept of flow to video games leads to game flow, the state of flow induced by the affordances of video games (considering the similarities between flow and game flow the terms will be used interchangeably throughout this paper) (Sweetser & Wyeth, 2005).

Game flow is a compound construct describing the contributors to an enjoyable gaming experience, which include concentration, control, clear goals, feedback, immersion as well as challenge and skills. The relation between the latter two is particularly important to evoke the state of game flow (Pallavicini & Pepe, 2019; Schmierbach et al., 2014).

Matching game difficulty levels with the players' skillset influences the players'

gaming experience (Faric et al., 2019a). According to game flow theory, the difficulty level of a game should match the individual capabilities of the player and develop as the skills

(4)

develop, to produce the highest amount of flow (Bodzin et al., 2021). When difficulty level and player skills are in balance and above average, flow is more likely to occur. In turn, low difficulty might lead to boredom, whereas high difficulty might lead to anxiety (Sweetser &

Wyeth, 2005). In line with this, research on flow in traditional, non–VR games has shown that too high or too low levels of difficulty subsequently lead to a decrease in flow, whereas matching difficulty to skill positively influences flow (Jin, 2012; Schmierbach et al., 2014).

Difficulty in VR games is connected to frequency and intensity of physical actions. This physical difficulty is determined by either a rhythmic component that indicates movement speed or a frequency component describing the familiarity with movement patterns in a gaming interval (Caronongan & Marcos, 2021). Hence, rhythm can be indicating difficulty via a variety in movement speed and rhythmic intensity (Costello, 2020), and therefore influence the sensation of flow.

A common way to integrate rhythm in gaming is via giving music a prominent role in the game (Georgiou et al., 2018). The presence of music in games is linked to a higher enjoyment and performance of gaming (Wechselberger, 2016). Especially when music is an integral part of the gameplay (e.g. music games, rhythm games), it increases the sense of spatial and physical presence for the user (Jin, 2012). Following the idea of game flow, the increase in fun combined with the presence of music is described to be positively influential on flow and subsequently enjoyment in games (Sites & Potter, 2018). The relation between music and the experience of flow in VR games is yet underexamined (Faric et al., 2019a).

Music in VR games is beneficial for gaming performance, which in turn increases fun while gaming and even increases the mood after the gaming experience (Rutrecht et al., 2021). To fully examine its influence, music should be attributed a central role in a VR gaming context, which implies gaming with rhythmic elements. Therefore, a rhythm game was selected for this study.

(5)

Rhythm games rely on music as a central component by indicating the actions a player has to perform based on the beat of a specific piece of music (Mueller & Isbister, 2014). This type of gaming requires tight synchrony between auditive, visual, and haptic cues (Georgiou et al., 2018). The rhythmic combination of music and movement in games has positive physiological effects as well as positive effects on enjoyment (Mueller & Isbister, 2014). The inclusion of music and synchronous movement in a gaming context has been subject to some research, mainly describing implications for eliciting flow (Mueller & Isbister, 2014; Walther

& Larsen, 2020), although it is unclear if these findings transfer to VR gaming. Combining music–movement synchrony with a matching level of physical difficulty has shown to cause higher levels of enjoyment in rhythmic games (Bronner et al., 2013), although it is up for examination whether this combination might not be overwhelming in VR.

Gaming in VR has implications for the creation of flow. Players feel more present and immersed in the actual game due to fewer (real–world) distractions in comparison to

traditional, non–VR gaming settings, which in turn leads to higher attentiveness and flow (Pallavicini & Pepe, 2019; Rutrecht et al., 2021). Higher levels of immersion through VR elicit more positive emotions and enjoyment compared to traditional, non–VR gaming

settings (Pallavicini & Pepe, 2019). Indeed, VR gaming experiences can induce a higher flow state among participants compared to traditional gaming settings (Bodzin et al., 2021). On the other hand, the potentially arousing character of VR might diminish flow due to visual fatigue and cognitive overload, when the amount of incoming prompts exceeds the processing

capacities (Souchet et al., 2022).

The implications of the inclusion of matching difficulty and synchrony in rhythmic VR gaming for the sensation of flow are topic of this study. The aim of the current study is twofold: a) to examine what effect matching rhythmic difficulty to the players' skills has on game flow in VR and b) what effect music–movement synchrony has on the players'

(6)

perceived game flow in VR. Subsequently, this study tries to answer the following research question:

How do rhythmic gaming difficulty and the synchrony between music and movements influence the experience of game flow in VR games?

Theoretical Framework Game Flow

The flow theory describes the optimal psychological state in which a person is fully immersed in an activity, feeling deep concentration, focus, and enjoyment. Within his theory, Csikszentmihalyi (1993) identifies eight determinants of flow: (1) a challenging activity that requires skills, (2) the merging of action and awareness, (3) clear goals and feedback, (4) concentration on the task at hand, (5) the paradox of control, (6) the loss of self–

consciousness, (7) an autotelic experience, and (8) the transformation of time. It describes a state where the experiencing entity enters a state of effortless attention, in a sense that high levels of attention require less effort than usual (Csikszentmihalyi & Nakamura, 2010).

Hence, the sensation of flow is primarily seen as a psychological state closely linked to emotions and enjoyment (Schmierbach et al., 2014).

The relation between flow and video games is described with the concept of game flow, stating that a game has to require concentration and attention from the player to evoke flow and absorb the user (Sweetser & Wyeth, 2005). Flow in a video game context describes an optimal experience where nothing else matters combined with the sensation of influencing the activities in the virtual world (Pallavicini & Pepe, 2019). Elements of game flow are closely connected to classic flow elements and include concentration, challenge, skills, control, clear goals, feedback, and immersion (Sweetser & Wyeth, 2005). Video games are specifically prone to evoke a state of flow since they have concrete rules, are adjustable to

(7)

players' skills, provide information on users' performance, and have multimodal information that facilitates concentration (Jin, 2012). More immersive games are able to create higher levels of positive emotions, enjoyment, and game flow due to fewer environmental distractions (Pallavicini & Pepe, 2019).

Game Difficulty and Flow

The main requirement regarding the sensation of flow in games is the relation between the difficulty of a game and the players' skillset. Players should unconsciously have the feeling that challenge and skills are in balance, to keep enjoyment on a high level

(Schmierbach et al., 2014). Difficulty should be variable and slowly increasing to provide a sensation of mastery, with an initially minimal challenge increasing over the course of the game (Sweetser & Wyeth, 2005). This balance in difficulty creates a feeling of competency, which is a necessary requirement for flow. If players are under the impression that they are in control of an environment and the tasks are manageable and of adequate challenge, the sensation of flow is more likely (Bodzin et al., 2021). Too difficult tasks would lead to frustration, whereas too easy tasks hamper concentration and lead to boredom. Both

subsequently result in lower levels of flow (Rutrecht et al., 2021). Additionally, players need to experience a continuous balance between their increasing, developing skills and the requirements of the game (Jin, 2012). In rhythmic games difficulty refers to the frequency of physical actions, implying that the frequency of actions needs to match the coordinative skills.

Either too low or too high of a challenge might result in less flow (Jin, 2012). Hence, the following hypothesis is stated:

H1: Matching rhythmic game difficulty to the players' coordinative skills results in higher levels of game flow compared to a too low difficulty or a too high difficulty.

(8)

Music, Rhythm, and Flow in Games

Music increasingly becomes a topic of interest in game studies (Sites & Potter, 2018).

Not only can music decrease heart rate (HR) or distract players from fatigue (Dillman Carpentier & Potter, 2007; Mueller & Isbister, 2014), it has also proven to be motivational and performance–enhancing (Bronner et al., 2013). In addition, diegetic music increases spatial presence and the feeling of players being able to manipulate surroundings, increasing the sense of immersion (Jin, 2012). Compared to silent games, a strong presence of music is more likely to elicit positive emotions while gaming (Lim et al., 2014). Enjoyment of the included music leads to less awareness of real–world surroundings and hence a higher possibility of immersion (Sanders & Cairns, 2010).

One way of giving music a central, gameplay–relevant role in video games can lead to rhythm games. Music itself is organized in regular patterns with an underlying pulse or beat, called rhythm (Bacon, 2012). In a gaming context rhythms occur in combination with visual and auditive elements, but also as part of physical movements. The patterns of these

components can vary in speed, flow, energy, tempo, or duration (Costello, 2020). From the players’ perception, rhythm is not only a cause–consequence relationship but also touches upon the need for competence. Not only does rhythm determine actions, but in the player’s perception actions in a certain rhythm can also affect the gameplay (Walther & Larsen, 2020).

Including music in the game also heightens interest and lowers auditive distractions, which is beneficial for concentration and hence for flow (Sites & Potter, 2018).

Combining a VR gaming setting with the presence of music and rhythm implies more physical movement compared to sedentary gaming settings. Games can use physical activity in order to make the whole experience more exciting and to increase enjoyment (Pallavicini &

Pepe, 2019). Adding this physical component to gaming (e.g. via rhythm games) is beneficial to the sensation of game flow, since players feel their whole body immersed and present in the gaming environment (Faric et al., 2019b). Due to the physiological activity and the

(9)

immersive nature of the medium in combination with music and rhythm, VR rhythm games can increase the players' overall gaming experience and positively influence flow (Bronner et al., 2013).

Music and the inherent rhythms determine players’ actions, moving rhythmically is strongly related to joy and a feeling of competence (Costello, 2020). Aligning movements and rhythm results in synchrony, the state in which the time lag between an indicating entity and a performed action is close to zero (Feniger-Schaal et al., 2021). This means there should be no deviation between the rhythmic indication of an action to perform and the actually performed action. Hence, creating a state of synchrony lies in the hand of the player, as it is him or her who aligns action to music (Sites & Potter, 2018). The state of synchrony elicits positive emotions and a sense of connection (Feniger-Schaal et al., 2021). The ability to (often subconsciously) synchronize motoric tasks is inherent to most people, as can be observed in dancing or finger tapping to the beat of a song (Bacon, 2012). Music can serve as a strong incentive to synchronize behavior, as it necessarily serves a specific rhythm (Llobera et al., 2016). Players tend to physically react to rhythms, for example by moving in tact of music during intense gaming intervals (Costello, 2020). Music helps to identify a beat, to visualize upcoming movements, or to identify rhythms (Mueller & Isbister, 2014). Thus, not the mere presence of music but specifically synchrony between movements and music is found to be able to make gaming experiences more pleasurable compared to the same activities in silence (Lim et al., 2014). Aligning actions and rhythm not only favors physical endurance (Bacon, 2012) but creates a psychological experience in which the player experiences a “back and forth” between following and determining rhythm (Costello, 2020). Moving synchronously combined with the sense of agency should hence benefit the sensation of flow and enjoyment (Lim et al., 2014). Extending this idea by the concept of difficulty creates two types of matches: Synchrony between music and movement and between skill and difficulty.

Following flow theory, the state of flow should be the highest when not only music and

(10)

movements are synchronous, but the rhythmic difficulty and the players’ skills are also in line with each other (Harmat et al., 2015). This leads to the following hypotheses:

H2: Synchrony between music and movement elicits higher levels of game flow than asynchrony between music and movements.

H3: Matching game difficulty to the players' skills result in higher levels of game flow when music and movements are synchronous compared to when music and movements are asynchronous.

Enjoyment and Boredom

Higher levels of perceived game flow are a prerequisite to higher states of enjoyment, whereas a lower sensation of flow is linked to boredom or frustration (Csikszentmihalyi, 2009; Rutrecht et al., 2021). The sensation of flow and enjoyment is also a physiological experience. Every physical activity is potentially enjoyable if there is an experience of control and harmony between capabilities and actions (Csikszentmihalyi, 2009). The experience of flow is linked to increased activity of the sympathetic nervous system, increased respiratory depth, and enhanced muscle relaxation, which in turn also favors enjoyment (Harmat et al., 2015). Hence, flow and enjoyment as excitative responses determine the degree of

physiological arousal.Enjoyment in games can be differently valent and intense in the sense that high enjoyment might lead to excitement and a stronger physiological response and a higher HR, whereas lower enjoyment might be connected to weak physiological responses, boredom, and lower HR (Bevilacqua et al., 2018; Harmat et al., 2015). Especially boredom and physical inactivity are connected to notably less enjoyment and lower HR (Bronner et al., 2013). Additionally, the combination with music influences physiological arousal by

activating the sympathetic nervous system (Dillman Carpentier & Potter, 2007). Hence, it is to be clarified whether perceived boredom in a rhythmic, movement–intense gaming setting also leads to lower physiological arousal.

(11)

H4: Higher levels of boredom lead to lower average levels of physiological arousal.

In turn, a higher gaming difficulty should lead to higher levels physiological arousal due to increased movement, which is connected to higher levels of enjoyment (Bevilacqua et al., 2018).

H5: Higher difficulties lead to increased physiological arousal, which in turn positively predicts enjoyment.

Method Sample

To appropriately address the hypotheses, an experiment among 202 participants recruited from the University of Amsterdam was conducted. Research credit points or a financial reward were offered as compensation for taking part. Participants were between 18 and 60 years old and gave their active consent for participation. The mean age of participants was 21.31 years (SD = 3.89). Apart from that, there were no requirements for participation, no experience with VR or games was required. The majority of participants were female (n = 150; 74.26%), 48 participants were male (23.76%), three participants were non–binary (1.49%), one participant did not indicate gender (0.49%).

Procedure

The procedure started with a short introduction and actively given consent.

Participants were asked to put on a Polar H10–chest strap that measured HR in combination with the Elite HRV app. Afterwards, participants received an introduction and safety

instructions for wearing a VR Headset. A short tutorial explaining the gameplay had to be completed. To manipulate flow experience participants were assigned to one condition of music–movement synchrony and one of three difficulty stages. Based on their performance in

(12)

the baseline measure, participants would either continue in a condition not matching their skill level (too high difficulty or too low difficulty) or start in a condition matching their skills, with slightly increasing difficulty based on their increasing skill to keep the conditions for game flow optimal. Participants played three sets of two songs each, equaling a total of six songs. Each gaming interval of two songs was 5 to 7 minutes long. After every two songs average values for HR, the level score, and the players’ estimation of playtime duration were noted down. After the experimental condition participants were offered a snack as well as a drink. Headset and heart rate monitor were taken off at this point, the second measurement phase via a questionnaire began. Participants answered questions about game flow,

enjoyment, extertion, demographics as well as about familiarity with the included songs. The experiment had a total duration of 40 minutes, with 25 minutes spent in VR. Participants were thanked and debriefed afterwards.

Two manipulation checks tested whether manipulations had the intended effect. First, an independent samples T–Test showed that those playing synchronously perceived

movements and music more in line with each other (M = 4.07, SD = 0.95) than those playing asynchronously (M = 3.74, SD = 1.07). The mean difference of 0.33 is statistically significant, t (194) = 4.13, p < .001, 95% CI [0.31, 0.88], the manipulation had the intended effect. The second manipulation check tested whether participants perceived difficulty levels as too hard, too easy, or matching. The effect is statistically significant, χ2 (6) = 59.30, p < .001, however not only as intended to be. 88.20% of participants indicating that gameplay was too easy played a too easy condition. Similarly, the majority of participants indicating a too high challenge were assigned to a too hard condition (94.40%). However, the wording of items might have led to some confusion, which is why only 43.60% of participants indicating slowly increasing difficulty played a condition matching their skill. Participants might have also had the feeling of increasing difficulty in too easy or too hard conditions.

(13)

Experimental Manipulation

The experiment followed a 2x3 mixed design, with music–movement synchrony (synchronous, asynchronous) and game difficulty (matching, too low, too high) as independent variables, and game flow as the main dependent variable. Difficulty levels differed in the speed and direction of movements that needed to be performed. Synchronous refers to the actual rhythm of the song and the movements that were to perform. In the asynchronous condition, the rhythm of movements did not match the rhythm of the song. In both synchronous and asynchronous conditions, the frequency of actions to perform

determines movement speed and thereby difficulty. Table 1 provides an overview of the operationalization over the three difficulty conditions.

Table 1

Research Design and participants

Low Difficulty Matching Difficulty High Difficulty

Movement Asynchronous to Rhythm

Slow (easy), asynchronous movements

n = 33 Participants

Matching asynchronous movements (to skill) n = 33 Participants

Fast (difficult) asynchronous movements

n = 33 Participants

Movement Synchronous to Rhythm

Slow (easy), synchronous movements

n = 34 Participants

Matching synchronous

movements (to skill) n = 33 Participants

Fast (difficult) synchronous movements

n = 36 Participants

The experiment implemented the Virtual Reality game “Beat Saber” in all

experimental conditions. Not only are all experimental variables inherent in the game (music, movement, varying difficulties), but the game also offered feasible opportunities to

manipulate each of the variables. The experiment included six popular songs from various epochs and genres to (a) not only provide songs from one specific interpret/ genre and (b)

(14)

heighten the possibility that participants are familiar with the songs and rhythms. During the experimental procedure, all participants were exposed to the following six songs: Gwen Stefani (Hollaback Girl, 110 bpm), Doja Cat (Say So, 111 bpm), Ke$ha (Blow, 120 bpm), Rammstein (Du Hast, 125 bpm), Britney Spears (Toxic, 143 bpm), Wiz Khalifa (Black and Yellow, 164 bpm). All six songs included (a) a synchronous version, where movements and rhythm of the song were in line with each other, and (b) an asynchronous version, where movements and rhythm of the songs were noticeably misplaced. To manipulate synchrony the web–application Beatmapper was used, which allowed to misplace movement indications and detach them from the rhythm of the song. Indications were misplaced randomly either one, two, or three beats too early or too late in sets of 30, beats were on average 950 milliseconds detached from the rhythm. Additionally, all six songs were included in four difficulty stages (easy–normal–hard–expert).

A pilot test among ten participants was conducted to find thresholds for the

classification of players in the “matching” condition and to determine the difficulty of songs in each condition. Results showed that out of 300 movements to perform, struggling

participants usually ended up below 250 correct movements. Mediocre players scored between 251 and 290 correct movements, whereas highly skilled participants correctly performed between 291 and 300 movements. These values were used as thresholds for the assignment of players to starting points in the matching condition (low, mediocre, high). The pilot test also revealed that the too hard–condition needed to be divided into two stages of difficulty, to not demotivate players. Very lowly skilled players would be immediately overwhelmed and frustrated when in very hard difficulties. Hence it was decided to include two difficulty stages for the too hard condition (hard difficulty <280; expert difficulty >280).

All pilot test participants were asked to play all included songs and to rate them according to their personal experience of difficulty within each difficulty stage. Table 2 provides an

(15)

overview of the thresholds for all conditions. The order within the conditions is based on participant ratings, movements per second, and bpm of the songs:

Table 2

Overview of songs over the conditions

Song 1 Song 2 Song 3 Song 4 Song 5 Song 6

Too Easy Toxic (e) Blow (e) Say So (e)

Du Hast (e) Black and Yellow (e)

Hollaback Girl (e) Match/ Low

Skills (0–259)

Toxic (e) Blow (e) Say So (e)

Du Hast (e) Black and Yellow (e)

Hollaback Girl (e) Match/ Medioc.

Skills (260–290)

Du Hast (n) Black and Yellow (n)

Blow (n)

Say So (n) Toxic (n) Hollaback Girl (n) Match/ High

Skills(291–300)

Du Hast (h) Black and Yellow (h)

Say So (h)

Hollaback Girl (h)

Blow (h) Toxic (h)

Too Hard (Low Skills, < 280)

Du Hast (h) Black and Yellow (h)

Say So (h)

Hollaback Girl (h)

Blow (h) Toxic (h)

Too Hard (High

Skills, >280) Hollaback

Girl (ex) Du Hast

(ex) Toxic

(ex) Black and Yellow (ex)

Blow (ex) Say So (ex) (e) = easy, (n) = normal, (h) = hard, (ex) = expert

Measures

Game Flow. Measuring flow states a challenge in the sense that if flow took place, participants need to reflect on a period they did not consciously experience (Sites & Potter, 2018). Hence, the measurement of flow needs to cover every subpart to be accurate. An adapted extensive scale by Fang et al. (2013) to measure flow in video games was used, which is strongly attached to the described eight components of flow. 17 items on cover all eight contributors to game flow (e.g. Playing this game challenges me; My attention was focused entirely on the game that I was playing; I lost the consciousness of my identity and felt like “melted” into the game; Playing this game is rewarding in itself). Together 12 of these items form a reliable, unidimensional scale explaining 63.30 % of the variance (α = .81,

(16)

M = 5.56, SD = 0.74), which covers all components of game flow. Since concentration is described to be an integral part of game flow, isolating these two items also led to a reliable scale, α = .86, M = 5.83, SD = 1.18.

Response States. A range of cognitive and emotional responses in line with Kendzierski & DeCarlo (1991) were measured (e.g. I enjoyed it, I disliked it, I found it

energizing, I found it frustrating). Ten items loaded on the underlying construct of enjoyment, forming a reliable scale (α = .91, M = 5.51, SD = 0.30). Four items formed a reliable scale for boredom, α = .81, M = 2.05, SD = 0.49. Three items formed a reliable scale for frustration, α

= .70; M = 1.85, SD = 0.92. To measure the physical component of enjoyment and boredom a Polar H10 heart rate monitor captured HR during the game, as previously used in studies regarding physiological effects of VR (Lemmens, Simon, & Sumter, 2022). An average value was recorded after each gaming interval (M = 113.95, SD = 19.63). Due to recognizable outliers, a z–standardization was performed. Two participants with z–scores below –2.00 were removed from further analyses.

Control Variables. As a control variable participants answered questions regarding their experience of the included music in the game. Participants indicated whether they are familiar with the songs they played and whether they enjoyed the songs to account for

possible intervening effects. On average, participants enjoyed the songs (M = 4.17, SD = 0.74) and were familiar with the songs (M = 3.98, SD = 1.41).

Results General Results

Table 3 provides an overview of the correlations between all relevant variables. Flow forms statistically significant, positive relationships with Song Enjoyment (r = .35, p < .001) and Enjoyment (r = .78, p < .001), as well as a negative relationship with boredom (r = –.67, p < .001). Hence, a higher sensation of flow is associated with high overall enjoyment and

(17)

lower levels of boredom. Players with higher game scores reported lower levels of boredom (r

= –.23, p = .001) and higher levels of enjoyment (r = .26, p < .001). Interestingly, song

enjoyment correlates with gender. Women (M = 4.07, SD = 0.63) enjoyed music slightly more than men (M = 3.76, SD = 0.76). The mean difference (of –.31) proved to be statistically significant, t (58.88) = –2.38, p = .021, 95% CI [–0.56, –0.05].

Table 3

Correlation Matrix

Variable n M SD 1. 2. 3. 4. 5. 6. 7.

1. HR 202 113.95 19.63

2. Flow 201 5.56 0.75 .13

3. % Score 199 80.92 17.56 –.13 .10

4. Song Enj. 183 4.01 0.67 –.02 .35** .04

5. Boredom 201 2.05 0.97 –.18* –.67** –.23** –.23**

6. Enjoyment 201 5.51 0.93 .11 .78** .26** .40** –.77**

7. Age 198 21.31 3.89 –.09 .12 –.04 .04 .01 .04

8. Gender 201 .01 –.05 –.07 .19** .11 –.06 –.11

*p < .05 **p < .01

Difficulty and Flow

To test whether matching game difficulty results in higher levels of flow compared to too high or too low difficulties (H1), a one–way analysis of variance was performed. All three groups were more or less of equal size and included more than 30 cases, the relevant

assumptions are satisfied. ANOVA with Bonferroni correction indicated that the effect of game difficulty on flow was statistically significant, F (2, 198) = 3.13, p = .046. However, the effect was of small size, explaining 3% of the variance, η2 = .03. Among those playing the matching condition, the average level of flow was notably higher (M = 5.74, SD = 0.67) than

(18)

among those individuals playing too easy levels in relation to their skill (M = 5.43, SD = 0.76). This mean difference (Mdif = 0.31) was statistically significant, p = .049, 95% CI [0.01, 0.62]. However, no significant differences could be detected between matching gameplay and those playing levels too hard for their skillset (M = 5.52, SD = 0.78, p = .243). Also, there were no significant differences in flow between those playing a condition too easy or too hard for their skill (Mdif = –0.09, p = 1.000, 95% CI [–0.22, 0.40]. Following this, an independent samples T–Test was conducted to inspect whether average levels of flow differed between matching and the combination of unmatching difficulties. In line with prior results, matching difficulty and skill (M = 5.74, SD = 0.67) evoked higher levels of game flow than unmatching conditions (M = 5.47, SD = 0.77). This mean difference was also statistically significant, t (199) = –2.40, p = .017, 95% CI [–0.48, –0.05] with a weak to moderate effect, d = 0.34.

Thus, H1 can be positively answered. There was a significant main effect of matching game difficulty to the players’ skillset on a higher sensation of flow, especially in comparison to unchallenging gameplay. A matching difficulty level elicited a higher average sensation of flow than a difficulty not matching the players’ skillset.

Synchrony and Flow

An independent samples T–Test was conducted to test whether synchrony between music and movements produced higher levels of game flow compared to rhythmically

asynchronous gameplay (H2). However, music–movement synchrony elicited lower levels of perceived game flow (M = 5.41, SD = 0.82) than asynchronous movements (M = 5.72, SD = 0.63). This mean difference (of 0.31) was statistically significant, t (190.86) = –2.98, p = .003, 95% CI [–0.51, –0.10], and represented a moderate effect, d = 0.43. Hence, H2 cannot be answered positively. Contrary to the expectations, asynchrony between music and movements elicited higher levels of perceived game flow than music–movement synchrony within the gameplay. A possible explanation is the higher amount of concentration needed when playing

(19)

asynchronously, which is considered a vital component of flow (Sweetser & Wyeth, 2005). A T–Test confirmed this assumption. The average levels of concentration were significantly different across levels of synchrony. Concentration was substantially higher in asynchronous conditions (M = 6.07, SD = 1.22) compared to synchronous conditions (M = 5.60, SD = 1.10), t (199) = –2.92, p = .004, 95% CI [–0.80, –0.16].

Looking at the combination between matching difficulty and synchrony (H1 and H2), H3 addresses whether matching difficulty results in higher game flow when music and

movement are synchronous compared to when asynchronous. To test this, a two–way analysis of variance was conducted. There was a statistically significant interaction effect of matching rhythmic difficulty and music–movement synchrony, F (5, 195) = 4.85, p < .001, η2 = .04.

Thus, 4% of the variance in game flow could be explained by the interaction between rhythmic difficulty and synchrony. The size of this effect was small. Indeed, the effect of music movement synchrony on game flow was different across conditions of difficulty. The means plot revealed that while within too easy and too hard conditions the sensation of flow was lower for players in synchronous conditions, this changed when difficulty matched the players’ skillset (Figure 1). When the challenge was too easy, the sensation of flow was lower when playing synchronously (M = 5.09, SD = 0.72) than among those playing asynchronously (M = 5.78, SD = 0.63). It was in this condition that the most pronounced difference in flow between synchronous and asynchronous could be observed. Similar, but smaller effects were observable when the challenge was too high for the players’ skillset. When playing

synchronously, average levels of flow were notably lower (M = 5.40, SD = 0.87) compared to when playing asynchronously (M = 5.65, SD = 0.66). This effect however changed, when difficulty matched the players’ skillset. Those playing in a synchronous condition reported slightly higher levels of flow (M = 5.75, SD = 0.74) than those playing asynchronously (M = 5.73, SD = 0.61). This effect is even more pronounced when looking at the interaction effect of difficulty as a dichotomy (unmatching vs. matching) and synchrony. A Two–Way ANOVA

(20)

showed a significant interaction effect between difficulty and synchrony, F (1, 201) = 6.97, p

= .009, η2 = .56. Again, there are significant differences between synchrony and asynchrony:

Too easy and too hard conditions combined show significantly lower levels of flow (M = 5.09, SD = 0.72) than matching difficulty (M = 5.57, SD = 0.82) in synchronous gameplay, whereas this changes in asynchronous gameplay. Switching to asynchronous conditions, there is a pronounced increase in flow for unmatching difficulties (M = 5.78, SD = 0.63) compared to only a slight increase for matching gameplay (M = 5.69, SD = 0.63). Hence, H3 can only partly be positively answered. The effect of matching rhythmic difficulty on game flow was indeed different between players that played in a synchronous condition compared to those that played in an asynchronous condition, albeit surprisingly not as pronounced as expected.

Figure 1

Estimated Marginal Means of Game Flow

Enjoyment, Boredom and Physiological Arousal

Following former argumentations, enhanced flow states are considered to be prerequisites to enjoyment, respectively low flow states might lead to boredom. This assumption can also be confirmed in this context: A regression analysis showed that higher levels of flow positively predicted enjoyment, F (1, 200) = 304.79, p < .001, b = .98 and negatively predicted boredom F (1, 200) = 157.76, p < .001, b = –.86. Hence, a one–unit

5 5,2 5,4 5,6 5,8

Easy Match Hard

Estimated Marginal Means

Condition: Easy, Match, Hard

Synchronous Not Synchronous

(21)

increase in flow led to no less than a 0.98 unit increase in enjoyment, whereas the same led to a 0.86 unit decrease in boredom. These effects are both strong, b*enjoyment = 0.78 and b*boredom

= –0.67. Whereas H1 confirmed the positive influence of matching gaming difficulty on flow, too high difficulties should lead to frustration and too easy difficulties should lead to

boredom. To test this, two T–Tests were conducted. The expected effect can be confirmed for frustration: too high difficulties (M = 2.10, SD = 1.16) led to significantly higher levels of frustration than too easy or matching difficulties (M = 1.72, SD = 0.74), t (199) = –2.77, p = .006, 95% CI [–0.64, –0.11]. However, boredom did not fulfill the expectations. Playing a level too easy for the skillset (M = 1.97, SD = 0.80) did not lead to significantly more boredom than matching or too hard conditions (M = 2.08, SD = 1.04), t (159.53) = 0.83, p = .406, 95% CI [–0.15, 0.37].

In the current study, players’ physiological arousal evolving from enjoyment or boredom was measured in HR. H4 addresses this physical influence of boredom, suspecting that higher levels of boredom subsequently lead to lower levels of HR. The regression model predicting HR from boredom while controlling for % of game score (as indicated by existing correlations) was statistically significant, F (2, 195) = 6.58, p = .002. The model was likely to appropriately predict HR in the population. However, the model only explained 6% of the variance, R2 = .06. As suspected, boredom negatively predicted HR. All else equal, an

additional unit of boredom averagely accounted for a decrease in HR of four beats, b = –4.27.

This effect was also statistically significant, t = –3.11, p = .002, 95% CI [–6.98, –1.56] and of moderate size, b* = –.22. Additionally the covariate had a negative, significant but rather small effect on heart rate, t = –2.53, p = .012, 95% CI [–0.34, –0.04], b* = –.18. The higher the percentage of correctly performed movements, the lower HR tended to be, b = .19. Hence, H4 can be supported. Boredom does indeed negatively predict physiological arousal.

Lastly, a mediation analysis was conducted to determine whether the effect of too high difficulty on enjoyment is mediated by physiological arousal (H5). The regression model was

(22)

statistically significant, F (2, 196) = 6.90, p = .001 and explained 7% of the variance, R2 = 0.07. In Step 1 of the mediation model, the regression of HR on too difficult gameplay was significant, b = 7.77, t = 2.81, p =.005, 95% CI [2.31, 13.22]. Everything else equal, too hard gameplay leads to a an increase in average heart rate of 7.77 beats compared to too easy or matching gameplay. Step 2 showed that the effect of HR as the mediator on enjoyment was also significant, albeit small (b = 0.01, t = 2.26, p =.025, 95% CI [0.01, 0.02]). A one-unit increase in heart rate predicts a 0.01 unit increase in enjoyment. Step 3 of the mediation process showed a significant direct effect of too hard difficulty on enjoyment, b = –0.46, t = – 3.33, p =.001, 95% CI [–0.74, –0.19]. On average, too hard gameplay predicts a decrease in enjoyment compared to too easy and matching gameplay. Step 4 of the analyses revealed that there was a significant indirect effect of too hard difficulty on enjoyment through an increase of HR (see Figure 2), meaning that H5 can be supported. Parts of the effect of too hard gameplay are mediated by increasing HR. This represents a relatively small effect, b* = 0.06.

Figure 2

Mediation Model for H5

Discussion

The main aim of this study was to establish how rhythmic gaming difficulty and the synchrony between music and movements influence the players’ experience of game flow in VR games. Game flow is described to be highest when tasks within a game are of adequate

(23)

challenge in relation to a player's skillset. Subsequently, too low challenges should lead to boredom, whereas high challenges lead to frustration due to continuously too demanding tasks. In line with expectations, results showed that the player's sensation of flow is highest when difficulty and skill match. This is not only supported by theoretical considerations regarding game flow (Sweetser & Wyeth, 2005), but reinforces prior findings (Rutrecht et al., 2021; Schmierbach et al., 2014). Comparing too hard vs. too easy gameplay conditions provided further insights into the basic assumptions of flow. Too hard gameplay led to frustration in line with all expectations, however too easy gameplay did not lead to significantly more boredom compared to matching and too easy gameplay. One possible explanation might be the overall high level of enjoyment (M = 5.51) and flow (M = 5.56) paired with the novelty effect of VR (Elor et al., 2022). It is well possible to be frustrated by unachievable demands of the game, however, there is an observable ceiling effect. It is less likely for players to be bored by a highly enjoyable experience, even if tasks are too easy.

Respectively, too low difficulty leads to a decrease in flow, but not necessarily to boredom.

Rhythmic synchrony however did not have the expected effect on game flow.

Although based on prior findings rhythmic synchrony should lead to higher levels of flow, this could not be confirmed within this experiment. Overall, asynchronous gameplay led to higher average levels of flow than synchrony. Contradicting Costello (2020), rhythmic synchrony between music and movements was not able to counterbalance intervening, flow–

enhancing effects of asynchronous conditions. To clarify this effect, the interaction between synchrony and difficulty was observed. Although participants reported the overall highest level of flow in a synchronous condition, this only held when difficulty matched the skills (Mueller & Isbister, 2014). However, in all other conditions synchrony did not elicit higher states of flow. In both too easy and too hard conditions participants reported higher levels of game flow in asynchronous conditions compared to synchronous gameplay. There are two possible explanations for this phenomenon.

(24)

On the one hand, the rhythmic displacement of movement and music in combination with the overall exciting and novel character of VR might have led to a cognitive overload and hence exceed processing capacities, following the idea of the limited capacity model of attention (Norman & Bobrow, 1975). When tasks do not interfere with each other, they seem to be controlled automatically. However, when this is not the case this incongruency elicits higher levels of attention, cognitive elaboration, and hence concentration (Zhang & Gao, 2014). It is this increase in concentration, that according to Sweetser & Wyeth (2005) constitutes one of the main components of game flow. Hence, rhythmic incongruency might theoretically also lead to higher levels of game flow due to the increased concentration it requires. Within that, another possible explanation might be the switch of focus following incongruency. As a consequence of unmatching music and increased attention, there is increased activity in the auditory cortex. To counterbalance this distraction there might be a switch of focus and processing to the visual cortex, which in turn attributes less attention to the actual musical stimulus (Zimmer et al., 2010). This is also supported by the McGurk effect, describing how visual information overrides auditory processing (McGurk &

Macdonald, 1976). When auditory information is distracting or of poor quality paired with extensive, high–quality visual information, processing capacities tend to override auditory senses. Thus, rhythmic asynchrony might in this case be initially detected, but evolve to be less disturbing due to more spent resources to decrease overload and visual fatigue (Souchet et al., 2022). The data also shows indications in line with this argumentation: not only is the sensation of flow notably higher in asynchronous conditions, but this is also confirmed when looking specifically at concentration. Reported levels of concentration were substantially higher for asynchronous conditions compared to synchronous gameplay due to the higher amount of necessary attention and visual focus, which might lead to higher levels of flow.

Furthermore, states of enjoyment and boredom were described to evolve from

different types of difficulty and are hence connected to flow. In line with flow theory, higher

(25)

flow states led to enjoyment, whereas lower flow states are connected to boredom (Bodzin et al., 2021). Expanding this by a physiological component of excitement and enjoyment, lower levels of excitement were expected to predict a lower average HR. This could also be

confirmed, following Elor et al. (2022). Additionally, a mediation model confirmed that the effect of higher difficulty on enjoyment is indeed mediated by physiological arousal.

Although too difficult gameplay alone led to a decrease in enjoyment, the higher

physiological arousal determined by higher intensity in turn positively mediated the effect on enjoyment in line with (Bevilacqua et al., 2018). An increase in flow and physical enjoyment was also suspected to influence time perception as another constituting variable of flow.

However, neither flow, enjoyment nor boredom did have the expected effects on time perception.

These results speak for a successful manipulation of the state of game flow. This has several implications. Firstly, findings from traditional gaming settings were partly confirmed and secondly, more profound insights about the effects of difficulty and synchrony in a VR gaming setting could be gained. Especially differences in flow between synchronous and asynchronous gameplay were surprising and seem to be unique to the VR gaming setting. The described cognitive overload resulting from the novelty and excitement of VR as a medium, paired with the incongruency of music and movement indicate that on average processing capacities of participants were exceeded in these conditions. A possible conclusion might be that rhythmic synchrony is simply not as important for producing a state of flow as initially suspected. These findings generate further insights into what constitutes pleasurable, flow–

enhancing experiences (Faric et al., 2019b) and open up new research areas in the field of music and rhythmic synchrony in VR gaming settings.

From a theoretical perspective, the general assumptions of game flow could be

confirmed. Participants felt more immersed and present, they enjoyed the experience and had a higher sensation of flow when difficulty matched a player's skillset. However, especially the

(26)

results regarding rhythmic synchrony may also provide further insight into the weighting of game flow components. Whereas music–movement synchrony seems not to be an integral part of flow, it does indicate how the related response states (concentration, attention) relate to flow. If attention and concentration rise, other components appear to be less influential on the overall sensation of flow. Nevertheless, the results allow no conclusion about a hierarchy between these components. Future research could examine possible moderating or mediating effects among flow components.

Overall, the methodological considerations were able to appropriately reflect the research aim. However, this experiment faced two methodological limitations: First, as reported in former studies the novelty effect of VR might have led for many participants to higher degrees of arousal and enjoyment (Elor et al., 2022), as indicated by the overall high levels of enjoyment and flow. This could be undermined with more experienced participants.

Second, the manipulation of synchrony did not have the intended effect. Most likely this can be attributed to the manipulation itself, which might have been too subtle. The misplacement of movement indication by on average 950 ms was noticed by participants, but could have been more prominent. Especially in more complicated gameplay phases with high

concentration required, the implemented asynchrony might just have been not prominent enough to influence the experience of flow.

Conclusion

In conclusion, this study generates new insights into the research area of flow and music in rhythmic VR games. Gaming in VR has implications that are beneficial for the optimal experience and enjoyment, described as game flow due to higher immersion, presence, and prominent physical movement. Results showed that matching skill and difficulty influences flow as expected. Matching difficulty and skills led to higher levels of flow, whereas unmatched conditions led to a decrease in flow. Higher states of flow were in

(27)

turn cause for higher enjoyment and lower levels of boredom. Synchrony however did not prove to be an integral part of the manipulation of flow, asynchrony between music and movement proved to be more beneficial to flow. Combined effects of rhythmic synchrony and matching difficulty led to highest levels of flow. Therefore, the combination of music and VR is potentially overwhelming when incongruent, but if applied with caution, might be able to amplify the sensation of flow. This state of flow in turn can be helpful for a variety of application areas in VR, even beyond gaming contexts. Rising fields such as exergaming, work, or leisure time activities in the metaverse might profit from users that from time to time simply Flow to the Beat.

(28)

References

Bacon, C. J. (2012). Effect of music-movement synchrony on exercise oxygen consumption.

The Journal of Sports Medicine and Physical Fitness, 52(4), 359–365.

Bevilacqua, F., Engström, H., & Backlund, P. (2018). Changes in heart rate and facial actions during a gaming session with provoked boredom and stress. Entertainment

Computing, 24, 10–20. https://doi.org/10.1016/j.entcom.2017.10.004

Bodzin, A., Junior, R. A., Hammond, T., & Anastasio, D. (2021). Investigating engagement and flow with a placed-based immersive Virtual Reality game. Journal of Science Education and Technology, 30(3), 347–360. https://doi.org/10.1007/s10956-020- 09870-4

Bronner, S., Pinsker, R., & Noah, J. A. (2013). Energy cost and game flow of 5 exer-games in trained players. American Journal of Health Behavior, 37(3), 369–380.

https://doi.org/10.5993/AJHB.37.3.10

Caronongan, A. P., & Marcos, N. A. (2021). Predicting chart difficulty in rhythm games through classification using chart pattern derived attributes. In R. Alfred, H. Iida, H.

Haviluddin, & P. Anthony (Eds.), Computational Science and Technology (Vol. 724, pp. 193–205). Springer Singapore. https://doi.org/10.1007/978-981-33-4069-5_17 Costello, B. M. (2020). Moving in rhythms: what dancers and drummers can teach us about

designing digital interactions. Digital Creativity, 31(3), 181–191.

https://doi.org/10.1080/14626268.2020.1782944

Csikszentmihalyi, M. (2009). Flow: The psychology of optimal experience. Harper & Row.

Csikszentmihalyi, M., & Nakamura, J. (2010). Effortless attention in everyday life: A

systematic phenomenology. In B. Bruya (Ed.), Effortless Attention (pp. 179–190). The MIT Press. https://doi.org/10.7551/mitpress/9780262013840.003.0009

Dillman Carpentier, F. R., & Potter, R. F. (2007). Effects of music on physiological arousal:

explorations into tempo and genre. Media Psychology, 10(3), 339–363.

(29)

https://doi.org/10.1080/15213260701533045

Elor, A., Powell, M., Mahmoodi, E., Teodorescu, M., & Kurniawan, S. (2022). Gaming beyond the novelty effect of immersive Virtual Reality for physical rehabilitation.

IEEE Transactions on Games, 14(1), 107–115.

https://doi.org/10.1109/TG.2021.3069445

Fang, X., Zhang, J., & Chan, S. S. (2013). Development of an instrument for studying flow in computer game play. International Journal of Human-Computer Interaction, 29(7), 456–470. https://doi.org/10.1080/10447318.2012.715991

Faric, N., Potts, H. W. W., Hon, A., Smith, L., Newby, K., Steptoe, A., & Fisher, A. (2019a).

What players of Virtual Reality exercise games want: Thematic analysis of web-based reviews. Journal of Medical Internet Research, 21(9), e13833.

https://doi.org/10.2196/13833

Faric, N., Yorke, E., Varnes, L., Newby, K., Potts, H. W., Smith, L., Hon, A., Steptoe, A., &

Fisher, A. (2019b). Younger adolescents’ perceptions of physical activity,

exergaming, and Virtual Reality: Qualitative intervention development study. JMIR Serious Games, 7(2), e11960. https://doi.org/10.2196/11960

Feniger-Schaal, R., Schönherr, D., Altmann, U., & Strauss, B. (2021). Movement synchrony in the mirror game. Journal of Nonverbal Behavior, 45(1), 107–126.

https://doi.org/10.1007/s10919-020-00341-3

Georgiou, O., Jeffrey, C., Chen, Z., Xiao Tong, B., Hei Chan, S., Yang, B., Harwood, A., &

Carter, T. (2018). Touchless haptic feedback for VR rhythm games. 2018 IEEE Conference on Virtual Reality and 3D User Interfaces (VR), 553–554.

https://doi.org/10.1109/VR.2018.8446619

Harmat, L., de Manzano, Ö., Theorell, T., Högman, L., Fischer, H., & Ullén, F. (2015).

Physiological correlates of the flow experience during computer game playing.

International Journal of Psychophysiology, 97(1), 1–7.

(30)

https://doi.org/10.1016/j.ijpsycho.2015.05.001

Jin, S.-A. A. (2012). “Toward integrative models of flow”: Effects of performance, skill, challenge, playfulness, and presence on flow in video games. Journal of Broadcasting

& Electronic Media, 56(2), 169–186. https://doi.org/10.1080/08838151.2012.678516 Kendzierski, D., & DeCarlo, K. J. (1991). Physical activity enjoyment scale: Two validation

studies. Journal of Sport and Exercise Psychology, 13(1), 50–64.

https://doi.org/10.1123/jsep.13.1.50

Lemmens, J. S., Simon, M., & Sumter, S. R. (2022). Fear and loathing in VR: The emotional and physiological effects of immersive games. Virtual Reality, 26(1), 223–234.

https://doi.org/10.1007/s10055-021-00555-w

Lim, H. B. T., Karageorghis, C. I., Romer, L. M., & Bishop, D. T. (2014).

Psychophysiological effects of synchronous versus asynchronous music during cycling. Medicine & Science in Sports & Exercise, 46(2), 407–413.

https://doi.org/10.1249/MSS.0b013e3182a6378c

Llobera, J., Charbonnier, C., Chagué, S., Preissmann, D., Antonietti, J.-P., Ansermet, F., &

Magistretti, P. J. (2016). The subjective sensation of synchrony: An experimental study. PLOS ONE, 11(2), e0147008. https://doi.org/10.1371/journal.pone.0147008 McGurk, H., & Macdonald, J. (1976). Hearing lips and seeing voices. Nature, 264(5588),

746–748. https://doi.org/10.1038/264746a0

Mueller, F., & Isbister, K. (2014). Movement-based game guidelines. Proceedings of the SIGCHI Conference on Human Factors in Computing Systems, 2191–2200.

https://doi.org/10.1145/2556288.2557163

Norman, D. A., & Bobrow, D. G. (1975). On data-limited and resource-limited processes.

Cognitive Psychology, 7(1), 44–64. https://doi.org/10.1016/0010-0285(75)90004-3 Pallavicini, F., & Pepe, A. (2019). Comparing player experience in video games played in

Virtual Reality or on desktop displays: Immersion, flow, and positive emotions.

(31)

Extended Abstracts of the Annual Symposium on Computer-Human Interaction in Play Companion Extended Abstracts, 195–210. https://doi.org/10.1145/3341215.3355736 Rutrecht, H., Wittmann, M., Khoshnoud, S., & Igarzábal, F. A. (2021). Time speeds up

during flow states: A study in Virtual Reality with the video game thumper. Timing &

Time Perception, 9(4), 353–376. https://doi.org/10.1163/22134468-bja10033 Sanders, T., & Cairns, P. (2010, September 1). Time perception, immersion and music in

videogames. BCS ’10: Proceedings of the 24th BCS Interaction Specialist Group Conference. 24th BCS Interaction Specialist Group Conference, Dundee, United Kingdom.

Schmierbach, M., Chung, M.-Y., Wu, M., & Kim, K. (2014). No one likes to lose: The effect of game difficulty on competency, flow, and enjoyment. Journal of Media

Psychology, 26(3), 105–110. https://doi.org/10.1027/1864-1105/a000120

Sites, R. D., & Potter, R. F. (2018). Everything merges with the game: A generative music system embedded in a videogame increases flow. Game Studies, 18(2).

Souchet, A. D., Philippe, S., Lourdeaux, D., & Leroy, L. (2022). Measuring visual fatigue and cognitive load via eye tracking while learning with Virtual Reality head-mounted displays: A Review. International Journal of Human–Computer Interaction, 38(9), 801–824. https://doi.org/10.1080/10447318.2021.1976509

Sweetser, P., & Wyeth, P. (2005). GameFlow: A model for evaluating player enjoyment in games. Computers in Entertainment, 3(3), 3–3.

https://doi.org/10.1145/1077246.1077253

Walther, B. K., & Larsen, L. J. (2020). Bicycle kicks and camp sites: Towards a phenomenological theory of game feel with special attention towards ‘rhythm.’

Convergence: The International Journal of Research into New Media Technologies, 26(5–6), 1248–1268. https://doi.org/10.1177/1354856519885033

Wechselberger, U. (2016). Music game enjoyment and natural mapping beyond intuitiveness.

(32)

Simulation & Gaming, 47(3), 304–323. https://doi.org/10.1177/1046878116651024 Yoo, S., Carter, M., & Kay, J. (2018). VRmove: Design framework for balancing enjoyment,

movement and exertion in VR Games. Proceedings of the 2018 Annual Symposium on Computer-Human Interaction in Play Companion Extended Abstracts, 295–307.

https://doi.org/10.1145/3270316.3272054

Zhang, J., & Gao, X. (2014). Background music matters: Why video games lead to increased aggressive behavior? Entertainment Computing, 5(2), 91–100.

https://doi.org/10.1016/j.entcom.2014.01.002

Zimmer, U., Roberts, K. C., Harshbarger, T. B., & Woldorff, M. G. (2010). Multisensory conflict modulates the spread of visual attention across a multisensory object.

NeuroImage, 52(2), 606–616. https://doi.org/10.1016/j.neuroimage.2010.04.245

Figure

Updating...

References

Related subjects :