Effect of Gameplay Uncertainty, Display Type, and Age on Virtual Reality Exergames
EEffect of Gameplay Uncertainty, Display Type, and Age onVirtual Reality Exergames
Wenge Xu
Xi’an Jiaotong-Liverpool UniversitySuzhou, Jiangsu, [email protected]
Hai-Ning Liang ∗ Xi’an Jiaotong-Liverpool UniversitySuzhou, Jiangsu, [email protected]
Kangyou Yu
Xi’an Jiaotong-Liverpool UniversitySuzhou, Jiangsu, [email protected]
Nilufar Baghaei
Massey UniversityAuckland, New [email protected]
ABSTRACT
Uncertainty is widely acknowledged as an engaging gameplay el-ement but rarely used in exergames. In this research, we explorethe role of uncertainty in exergames and introduce three uncertainelements (false-attacks, misses, and critical hits) to an exergame.We conducted a study under two conditions (uncertain and cer-tain), with two display types (virtual reality and large display) andacross young and middle-aged adults to measure their effect ongame performance, experience, and exertion. Results show that (1)our designed uncertain elements are instrumental in increasingexertion levels; (2) when playing a motion-based first-person per-spective exergame, virtual reality can improve performance, whilemaintaining the same motion sickness level as a large display; and(3) exergames for middle-aged adults should be designed with age-related declines in mind, similar to designing for elderly adults. Wealso framed two design guidelines for exergames that have similarfeatures to the game used in this research.
CCS CONCEPTS • Software and its engineering → Interactive games ; •
Human-centered computing → Virtual reality ; •
Applied computing → Computer games . KEYWORDS exergame, uncertainty, virtual reality, young adults, middle-agedadults
ACM Reference Format:
Wenge Xu, Hai-Ning Liang, Kangyou Yu, and Nilufar Baghaei. 2021. Effect ofGameplay Uncertainty, Display Type, and Age on Virtual Reality Exergames.In
CHI Conference on Human Factors in Computing Systems (CHI ’21), May ∗ Corresponding author.Permission to make digital or hard copies of all or part of this work for personal orclassroom use is granted without fee provided that copies are not made or distributedfor profit or commercial advantage and that copies bear this notice and the full citationon the first page. Copyrights for components of this work owned by others than ACMmust be honored. Abstracting with credit is permitted. To copy otherwise, or republish,to post on servers or to redistribute to lists, requires prior specific permission and/or afee. Request permissions from [email protected].
CHI ’21, May 8–13, 2021, Yokohama, Japan © 2021 Association for Computing Machinery.ACM ISBN 978-1-4503-8096-6/21/05...$15.00https://doi.org/10.1145/3411764.3445801
ACM, New York, NY, USA, 14 pages. https://doi.org/10.1145/3411764.3445801
Motion-based exergames, a combination of “motion-based exercise”and “gaming”, is a promising approach to encourage regular exer-cise, especially for unmotivated or inactive target groups [8, 70].Previous literature has shown the benefits of playing motion-basedexergame, which include but are not limited to enhanced postu-ral stability [68], muscle strength [72], and working memory [18].Because of the potential of these exergames in eliciting health ben-efits, much work has been conducted with different age groups(including children [30], young individuals [84], and older adults[25]).Age-related declines are common in older adults (i.e., aged 65and above) and middle-aged adults (i.e., aged 45 to 65) as previousstudies show that reductions (e.g., cognitive abilities) could starteven before the age of 50 [20, 77]. These age-related declines affectthe elderly’ game performance and experience and could also affectmiddle-aged adults in a similar way. Although there have been someattempts to understand whether middle-aged adults could obtainthe same health benefits from playing videogame as elderly adults[65, 80], there is very limited research on exploring the performanceand experience of middle-aged adults.Designing an enjoyable and effective exergame is challenging.Studies [3, 5, 33] have been conducted to improve the motivationand experience of these games. For instance, Ioannou et al. [33] pro-posed a virtual performance augmentation method for exergamesand found that it increased players’ immersion and motivation.Barathi et al. [3] implemented an interactive feedforward methodto an exergame and found that it improved players’ performance.One factor that has been widely applied in games is uncertainty,which has long been recognized as a key ingredient of engaginggameplay [11, 12, 35, 63]. Costikyan [12] argues that games requireuncertainty to hold players’ interest and that the struggle to mas-ter uncertainty is central to games’ appeal. Most importantly, hesuggested that game designers can harness uncertainty to framegameplay’s design. Several game designers and researchers havetried to identify uncertainty sources that can lead to a good game-play experience [15, 36, 43, 74]. Drawing on many of these sourcesand practical experience, Costikyan [12] listed an influential cate-gorization of eleven sources of uncertainty found or can be used a r X i v : . [ c s . H C ] J a n HI ’21, May 8–13, 2021, Yokohama, Japan Xu et al. in games. Recently, Kumari et al. [41] presented a grounded the-ory of uncertainty sources which can partially map onto existingtaxonomies, especially Costikyan’s [12], providing converging evi-dence of the validity of Costikyan’s categorization of uncertaintysources. Although uncertainty is recognized as a core componentof the gaming experience, there is relatively little research that haslooked specifically into the effect of uncertainty in games, espe-cially exergames. Based on uncertainty sources identified in [12],in this research, we propose the use of three uncertain elementsfor exergames, that cover four sources of uncertainty, and evaluatetheir effect in exergames on performance, game experience (andsickness when implemented in virtual reality), and exertion levels.Given the recent emergence of affordable virtual reality (VR)head-mounted displays (HMDs), VR exergames have been gainingrapid attention [3, 33, 82]. For instance, VR exergames are useful inpromoting physical activity in sedentary and obese children [64],especially to increase their motivation to exercise [50, 62]. Existingliterature has outlined that there are additional benefits of playingmotion-based exergames in VR than non-VR. In VR, players couldachieve a higher exertion and experience a game more positivelyin areas like the challenge, flow, immersion and a lower negativeaffect [84]. However, a major drawback is that VR might lead to ahigher level of simulator sickness, which must be taken into accountduring the design process to mitigate its effects.The aim of our research is to explore the effect of uncertain ver-sus certain elements and VR versus a typical TV large display (LD)on two main player groups of exergames regarding their game per-formance, experience, and exertion. In this paper, we first introduce
GestureFit , the game we developed for this research. We describethe rules and logic behind it, the game procedure, and risk controlfor middle-aged adults. We then present the study we conducted toinvestigate the effect of display type and game condition, focusingon differences between young adults and middle-aged adults. Wethen report the results and present a discussion of our findings thatare framed based on existing literature. Two main design guide-lines derived from the results are then proposed, followed by theconclusions.The contributions of the paper include: (1) an empirical evalua-tion of the effects of display type and game condition on exergameperformance, experience, and exertion between young and middle-aged adults; (2) a set of uncertain elements that can help increasethe exertion level for motion-based exergames; and (3) two rec-ommendations that can help frame the design of motion-basedexergames to contain uncertain gameplay elements and how tomotivate middle-age and older adults to engage with exergamesmore meaningfully.
Many motion-based exergames have been developed for non-VRdisplays since the introduction of Kinect. A typical motion-basedexergame requires players to move their body or perform certaingestures to interact with the game world. For instance, in
GrabApple [24], users need to jump or duck to pick up apples; they also need tomove around to locate them but also avoid touching other objects,like bombs. In a game reported in Gerling et al. [25], users need to perform static and dynamic gestures to grow plants and flowers andcatch birds. In
Sternataler [71], players use their hands to collectstars that appear sequentially in some predefined paths.Recent advances and the growing popularity of VR HMDs havecreated a substantial demand for motion-based exergames. Forinstance, games like
Virtual Sports for the HTC VIVE allow a userto play sports with his/her full body in fully immersive virtualenvironments. In another commercial game, FitXR , the users needto jab, weave, and do uppercuts following rhythmic music. In theresearch exergame KIMove [83], the players need to move theirhands to hit fruits floating in midair and use their feet to step oncubes moving towards them on the ground. In
GestureStar [84],users need to perform 6 different gestures to eliminate the objects,like cubes, flying towards them.Previous research has reported inconsistent findings when look-ing at the effect of display type on gameplay experience and per-formance. Xu et al. [84] suggested that players achieved a higherexertion and experienced a game more positively in VR than LD.However, they also found that VR could lead to a higher level of sim-ulator sickness. Results from [83] suggested that there was no effectof display type on gameplay performance and experience. There-fore, we have included this factor in our experiment to investigateit further and provide more insights.
Exergames integrate physical activity to engage players [55]. Be-cause findings from other types of games may not be applicableto exergames [51, 84], efforts have been focused on studying userexperience in exergames. For instance, it is reported in [83] that taskmode (single- and multi-tasking) could affect users’ exergame expe-rience; in particular, multi-tasking could not only make the gamemore challenging and cause a higher sickness, but also lead to worseperformance than single-tasking. Koulouris et al. [39] investigatedthe effect of customization and identification in a VR exergame,and found that customization significantly increased identification,intrinsic motivation, and performance in the exergame. Further,playing pose (i.e., standing and seated), performance augmentation(i.e., enabling players with superhuman capabilities in the virtualgame) could also affect the gameplay experience (e.g., sickness)[33, 81]. On the other hand, although uncertainty is a crucial el-ement in gameplay, it is underexplored in exergames. It is thisreason that we are interested in studying the effect of uncertaintyin exergames for both immersive VR and large displays.
Several design guidelines have been proposed by researchers in HCIand sport sciences for designing more attractive and effective full-body motion-based exergames [29, 44, 45]. According to these, todesign a playful exergame experience, designers should focus on (1)the player’s body (movement concept), (2) the mediating controllertechnology (transferring movement input into the virtual worldand providing feedback), and (3) the game scenario (audio-visualand narrative design and feedback) [48]. https://fitxr.com/ ffect of Gameplay Uncertainty, Display Type, and Age on Virtual Reality Exergames CHI ’21, May 8–13, 2021, Yokohama, Japan After criticizing existing exertion gamesand commercial exergames, Marshall et al. [45] proposed threedesign strategies based on the idea of movement, which are (1) thedesign of exertion trajectories (e.g., to create a trajectory acrossindividual play sessions for skill-learning that takes into accountplayers’ cognitive load and the exertion patterns), (2) design for,with, and around pain (e.g., celebrating positive pain), and (3) de-sign leveraging the social nature of exertion (e.g., players to besurrounded by other players like friends and family members orgame enthusiasts).
Studies have suggestedthat the participation of the body is a crucial variable not only in theefficacy of exergames in affecting users’ emotional experience [76],but also in improving user experience, energy expenditure, andintention to repeat the experience [38]. To achieve these positivegaming experiences, body-centered controllers should be designedto serve as an additional physical playground, so that they canbe easily integrated into players’ body scheme [60] and provide abalance of guided and free movements [48].
Exergame should involve specific prefer-ences for game mechanics, levels, visuals, audio, and narrative. Thisrequirement will unavoidably make it essential to involve the targetgroup in the design process from the start [46, 47]. The literatureoffers suggestions for key elements of game scenarios. For instance,games should include an immediate celebration of movement artic-ulation by providing direct and constrained amounts of feedback[52]. Also, games should involve achievable short-term challengesto foster long-term motivation and help players identify rhythmin their movements, for example, by setting movements that aremapped to specific sounds and visualizing previous and upcomingmovements [52, 53]. It is also important to provide a challenge thatmatches individual skill levels, for instance, balancing the challengelevel by monitoring the player’s heart rate [54].
Caillois [11] says that the outcome of a game should be uncertainfor it to be enjoyable. Similarly, Costikyan [12] argues about theimportance of uncertainty in the overall game experience and hasdeveloped an influential categorization of 11 sources of uncertaintywithin games. Typical uncertainty sources are (1) Performativeuncertainty: uncertainty of physical performance (e.g., hand-eyecoordination); (2) Solver’s uncertainty: weighting a group of op-tions against potential outcomes; (3) Player unpredictability: notknowing how the opponents/teammates will act; (4) Randomness:uncertainty emanating from random game elements. Recently, Ku-mari et al. [41] developed an empirically-based grounded taxonomyof seven sources of uncertainty across the input-output loop thatinvolves the game, the player, and their interaction in an outcome.This taxonomy partially maps onto existing taxonomies, especiallythe one proposed by Costikyan [12]. This, in turn, provides furtherevidence of its validity. Hence, in this research, we used Costikyan’ssources of uncertainty to guide the design of the uncertainty ele-ments in our exergame. To explore the effects of uncertainty in exergames, we appliedthree uncertain elements in an exergame we developed: (1)
False - Attacks : this concept is originally from sports (e.g., basketball) andhas been applied widely in sports videogames (e.g., NBA 2K series).(2)
Misses : this concept has been widely used in games (e.g., Dun-geon & Fighter) where an attack hits the opponent but is countedas a miss by the system. (3)
Critical Hits : this concept has also beenwidely used in games (e.g., Dungeon & Fighter). When a criticalhit happens, the player issuing the hit causes more damages to theopponent that a normal successful blow.
Users from different age groups often perceive gameplay elementsdifferently—for instance, what is motivating for one group maynot be so for another. Motivations can change with age: fantasy isa powerful motivational factor in younger children [27], whereascompetition and challenge-related motives are stronger in olderchildren and adolescents [69]. Young adults are more motivatedby rewarding experiences, while older adults are more inspired byperceived benefits to their health [73]. Young adults tend to prefervisually appealing graphics and music that fit the theme and natureof the game, but older adults pay more attention to the feedbackthat helps them complete a game [73]. Furthermore, there is an in-creased appreciation for the enjoyment that a game brings, greatersatisfaction for autonomy, and decreased competence as users age,especially after a certain threshold [6]. In other words, young adultsprefer exergames that allow them to challenge themselves physi-cally and cognitively, but older adults preferred exergames that arefun to play and are beneficial to their health [73].Gajadhar et al. [22] investigated the social elements of gameplayfor young adults. They found that gameplay is most enjoyablewhen gamers are co-located, less satisfying in mediated co-play,and the least enjoyable in virtual co-play. However, these threesocial contexts (virtual, mediated, and co-located co-play) do notpositively influence older users like younger adults [7, 23]. Gerlinget al. [26] explored the effect of sedentary and motion-based controltasks in games (such as pointing and tracking) for older adults andyounger adults, and found that older adults performed worse thanyoung adults.There is a large body of work on the experience of children[2, 17, 19] and young adults [81, 83, 84], and older adults [13, 14, 25]with videogames. However, there is only limited attention givento middle-aged players. Previous research suggested age-relateddeclines could start when people are in their mid-age; for instance,age-related memory impairment and executive dysfunction can befound in people before they reach 50 [20, 77]. Middle-aged adultssuffer from several age-related declines, including but not limited tolower working memory [49], grip strength [40], and muscle mass[10]. Given this above research, our work involves two groups,young adults (18-30) and middle-aged adults (45-65), to explore theeffect of age on exergames.
HI ’21, May 8–13, 2021, Yokohama, Japan Xu et al.
Table 1: Features and requirement for each move by the player a and the monster b . Name Description of the move
Kick a An attack move that inflicts 10 hp damage to the opponent in the kicking direction and requires a 3-second cooldown.
Punch a,b
An attack move that inflicts 10 hp damage to the opponent on the punching direction and requires a 3-second cooldown.
Zoom + Kick a A ranged attack move that inflicts 30 hp damage to the opponent in that attack range (1m) and requires a 5-secondcooldown.
Squat b A ranged attack move that deals 30 hp damage and requires a 5-seconds to cooldown.
Zoom + Squat a A defense move that releases a sphere to protect the user for 2 seconds and heals 20 hp if it could successfully defend theplayer from the monster’s attack. This move requires a 3-second cooldown.
The game was implemented in Unity3D with the Oculus Integrationplugin and the Kinect v2 Unity plugin . The design of our game was inspired by
Nintendo Ring Fit Adven-ture . The goal of the game is for the player to stay alive and defeata monster three times. To do this, the player needs to perform ges-tures to make attacks against the monster and defend themselvesfrom being attacked by it. The player begins with 100 health points(HP) while the monster has 500 HP. The monster or player dieswhen their HP reaches 0. Both the monster and the player have3 lives. The monster could move leftward or rightward within a2-meter range prior to its game starting position. Players’ lateralmovement is limited so that they are always within the operationaltracking range [33, 84]. The game is designed to take this into ac-count so that the gameplay experience is not affected. Both visualand audio feedback is provided to give a fuller range of sensoryexperience to players. There are three attack moves and one defense move. All moves canbe released by performing their corresponding gestures. These fourmoves are (i)
Kick : kicking using any leg, (ii)
Punch : single handpunching, (iii)
Zoom + Kick : kicking using any leg and leaning armsforward and stretching them out, and (iv)
Zoom + Squat : performinga squat and leaning arms forward and stretching them out. Theselected gestures were chosen based on design recommendationsfrom previous studies on young adults [84] and older adults [25].Table 1 lists pre-defined features and their requirements.
The uncertain condition includesthree uncertain elements, which covers four uncertainty sources[12]: • False-Attacks : There is a 20% chance that the monsterwould perform a false-attack (which lasts around 0.8 seconds)when the system triggers an attack-related animation to trickthe player into performing the defense move. False-attackscover the following uncertainty sources: a)
Performativeuncertainty : our game challenges eye-body coordination(i.e., would the players be able to cancel their defense move https://assetstore.unity.com/packages/tools/integration/oculus-integration-82022 https://assetstore.unity.com/packages/3d/characters/kinect-v2-examples-with-ms-sdk-and-nuitrack-sdk-18708 when they realize the monster is performing a false-attack?),b) Solver’s uncertainty : it is concerned with whether per-forming or not performing a defense move against potentialoutcomes (i.e., wasting a defense move to a false-attack orbeing successful in defending from an actual attack), and c)
Player unpredictability : this is about the uncertainty ofthe opponent’s movements (e.g., whether it is a false or realattack). • Misses : There is a 10% chance that the player’s or monster’sattack would be regarded as a miss even if it hits the oppo-nent.
Randomness : misses act as a random element in thegame. • Critical Hits : There is a 10% chance that the player’s ormonster’s attack could be a critical hit, which would deal50% more damage than a normal attack move.
Randomness :critical hits act as another random element in the game.The only difference between the certain and non-certain condi-tions is that the former does not include the above three uncertainfeatures.
In both conditions, the monster wouldperform an action every 2 sec. In the certain condition, if any at-tack skill is available, there is 80% chance that the action is anattack (either 100% for the only skill that is available or 50% foreach skill that is available); otherwise, it is a walk. The uncertaincondition also follows this attack mechanism; the only differenceis that if an attack skill is available, there is 80% chance the actionis attack-related (i.e., 8/10 = a real attack, 2/10 = a false attack).
The game starts with a training (warm-up) phase (see Figure 1a-b),where the player needs to use attack and defense moves. The orderof the moves required for the player to perform is
Kick , Punch , Zoom + Kick , Zoom + Squat . For attack moves, the player needs toperform the corresponding gesture, and its attack must damage themonster twice before proceeding to the next move. For the defensemoves, the player must successfully defend themselves from themonster’s attacks twice to finish the training. The player needs toperform a
Zoom gesture between each move training to switch tothe next move training. After the training phase, the player needsto perform another
Zoom gesture to start the gameplay phase.During the gameplay phase (see Figure 1c-d), players need toperform the gestures to attack and defend themselves. If the playershave no HPs, they need to perform
Zoom + Squat five times to regain ffect of Gameplay Uncertainty, Display Type, and Age on Virtual Reality Exergames CHI ’21, May 8–13, 2021, Yokohama, Japan
Figure 1: Screenshots of
GestureFit : (a) LD training phase, (b) VR training phase, (c) LD gameplay phase, and (d) VR gameplayphase. All variables are the same in all versions except in VR the player information is slightly tilted. life and perform
Zoom once to confirm they are ready to return. Ifthe monster has no HPs, the game will play an animation of themonster falling to the ground and is destroyed. After a 5-secondwait, the monster uses its second or third life and the game re-starts.The game ends when the monster or the player has no lives andHPs left.
We controlled the risk, if any, to a minimal level. As pointed in[46, 47], having users involved in the development process is useful.As such, for our game prototype, we had two middle-aged adultsfrequently involved during the development process to test thegestures’ suitability, tune parameters (e.g., cooldown time, shieldprotection’s duration) and ensure accurate and meaningful execu-tion of movements. The selected gesture worked quite well sinceall middle-aged participants had no issues performing them duringthe experimental gaming sessions (as our results would show; moreon this later).Besides, we minimized any risks by (1) making a first-personviewing perspective game so that players can see their motions,(2) limiting the number of monster’s attack skills and having gapsin its attacks, (3) restricting players’ position, (4) allowing them 5sec rests after they took a monster’s life, (5) allowing them to restas much as they want after they lost one life, and (6) displayinginformation (user’s skills, player’s HP, and monster’s HP) in frontof the users without the need for additional head movement.
The experiment followed a 2 × 2 within-subjects design with twowithin-subjects factors: Display Type (DT: VR and LD) and (2) GameCondition (GC: certain and uncertain). The order of DT × GC wascounterbalanced in the experiment.To determine participants’ task performance, we collected thefollowing (1) completion time on each of the three lives of themonster; (2) success rate of each move; and (3) the total number ofeach type of gestures performed.Participants’ experience was measured with Game ExperienceQuestionnaire (GEQ) [32] and Simulator Sickness Questionnaire(SSQ) [37]. We used the 33-item core module of the GEQ to measuregame experience, which consists of seven components: competence,immersion, flow, tension, challenge, negative affect, and positive af-fect. Simulator sickness was assessed using the 16-item SSQ, which produces 3 measures of cybersickness (nausea, oculomotor, anddisorientation).Exertion was evaluated by (1) the average heart rate (avgHR%)expressed as a percentage of a participant’s estimated maximumheart rate (211-0.64 × age) [58], (2) calories burned, and (3) Borg RPE6-20 scale [9].We measured the acceptability of the uncertain elements used inour games with three questions: “ I like the design of the false-attacks ”,“
I like the design of attacks that could be missed by chance ”, and “
Ilike the design of attacks that could be a critical hit by chance ”. Thequestions followed a 1-7 Likert scale, with 1 indicating “extremelydisagree” and 7 indicating “extremely agree”.After completing the above questionnaires, we conducted a semi-structured interview for participants with the following open-endedquestions: “
Overall, what did you think about the game ?”, “
Whatdid you like about the game ?”, “
What did you not like about thegame ?”, “
Was there anything more difficult than you expected in thegame ?”, and “
Was there anything more confusing than you expectedin the game ?” [16]. Answers were recorded and transcribed in textand later analyzed by two of the researchers following an infor-mal, simplified inductive open coding approach [66]. Themes wereconcluded by the two researchers independently and agreed in apost-coding meeting with a third researcher. Details of the themescan be found in the feedback section (Section 4.5.5). There was nolimit for the length of participants’ responses.
We used an Oculus Rift CV1 as our VR HMD and a 50-inch 4K TVas our LD. Both devices were connected to an HP Z workstationwith an i7 CPU, 16GB RAM, and a Nvidia Quadro P5200 GPU.Players’ gestures were detected via a Microsoft Kinect 2, which wasalso connected to the HP Z workstation. The heart rate (HR) wasmonitored by a Polar OH1 optical HR sensor, which has been provento be reliable compared to the gold standard of HR measurementwith an electrocardiography device [31, 67]. Figure 2 shows theexperiment setup and devices used in the experiment.The experiment was conducted in an indoor laboratory roomthat could not be seen from the outside. The laboratory room waswell illuminated, and its temperature was controlled by an air con-ditioner that regulated the room temperature to 24℃ during theexperiment.
HI ’21, May 8–13, 2021, Yokohama, Japan Xu et al.
Figure 2: Experiment setup and the devices used in the ex-periment: (1) the Oculus Rift CV1; (2) a 50-inch 4K TV; (3)the HP Z backpack; (4) the Microsoft Kinect 2; and (5) PolarOH1.
Participants were recruitedfrom a local university campus and a local community centerthrough posters, social media platforms, and a mailing list for youngadults between 18 and 30 years old and middle-aged adults between45 to 65 years old. The study included participants who were notdisabled, were not pregnant (because of the physical exertion re-quired to play the game), and had not consumed any alcohol duringthe day (because blood alcohol level of approximately 0.07% couldreduce symptoms of cybersickness [34], which might affect theresults of our study).Participants were excluded from the experiment if they (1) an-swered “yes” to any of the Physical Activity Readiness Question-naire [75] questions, (2) had resting blood pressure higher than140/90 mmHg, and (3) had an extremely good or poor resting heartrate (RestHR) level (i.e., heart rate range were the top 10% or thelast 10% of the population) depending on their age and gender [59].
Thirty-two (32) participants partic-ipated in our study—16 young adults (6 females; mean age = 20.6,SD = 1.31, range 18 to 23; BMI = 20.3, SD = 2.62), and 16 middle-agedadults (5 females; mean age = 47.7, SD = 2.68, range 45 to 54; BMI= 23.8, SD = 2.04). Among young adults, 7 of them had experiencewith VR HMDs, but none were regular users. Fourteen of themplayed videogames before; 6 of them played regularly. For middle-aged adults, none had experience with VR HMDs and videogames.There were no dropouts in this experiment.
The duration of each session was about one hour. Before the ex-periment began, participants needed to fill out a pre-experimentquestionnaire that gathered demographic information (e.g., age,gender, and experience with the VR device) and Physical Activity Readiness Questionnaire [75]. After a brief description of the exper-imental procedure, participants signed the consent to participatein the experiment and collected their RestHR and resting bloodpressure level. They were also asked to enter their age, gender,height, and weight into the Polar Beat app.Before each condition started, a researcher would help eachparticipant to wear the required devices (e.g., Polar OH1). Oncetheir HR reached the equivalent RestHR level, they were led to theexperiment stage, beginning with a training (warm-up) phase andthen the gameplay phase (see Figure 1 and Section 3.2). After eachcondition, they were asked to fill in post-condition questionnaires(GEQ [32], SSQ [37], Borg RPE 6-20 scale [9]). They proceeded tothe next condition when they felt rested and their HR was at theresting level. Once they completed all conditions, they needed tocomplete a post-experiment questionnaire and a semi-structuredinterview.
We used SPSS version 24 for windowsfor data analysis. We employed a three-way mixed ANOVA withGC (uncertain and certain) and DT (VR and LD) as within-subjectsvariables and Age (young adults—YA and middle-aged adults—MA)as the between-subjects variable. We applied Age as the between-subjects variable because we want to follow existing approaches inthe literature [26, 57, 78]. Bonferroni correction was used for pair-wise comparisons. Effect sizes ( 𝜂 𝑝 ) were added whenever feasible.To minimize any impact on the readability of the paper, we haveplaced all the data results in the tables of an appendix located afterthe references. Completion Time on Each Life . Figure 3a presentsthe mean completion time of each life (i.e., monster’s life1, life2,life3). ANOVA tests yielded a significant effect of Age on life2( 𝐹 , = . , 𝑝 < . , 𝜂 𝑝 = . 𝐹 , = . , 𝑝 < . , 𝜂 𝑝 = . Success Rate . Table 2 shows the ANOVA tests of the successrate for
Zoom + Squat , Kick , Zoom + Kick . Corresponding success ratedata can be found in Figure 3b,c and Figure 4a. In summary, (1)participants have a higher defense (i.e.,
Zoom + Squat ) success rate incertain GC than uncertain GC, (2) YA have a higher defense successrate in VR than LD, (3) participants have a higher
Kick success ratein VR than LD, (4) YA had a higher
Zoom + Kick success rate thanMA in VR, (5) YA had a higher
Zoom + Kick success rate in VR thanLD, and (6) YA had a higher
Zoom + Kick success rate than MA inuncertain GC.
Total Number of Gestures Performed . Table 3 shows the ANOVAtests of the total number of gestures performed for
Zoom + Squat , Punch , Zoom + Kick . Corresponding success rate data can be foundin Figure 4b,c. In summary, (1) YA and MA both performed moredefense moves (i.e.,
Zoom + Squat ) in uncertain GC than certain GC,(2) MA performed more defense moves than YA in both certain anduncertain GC, (3) YA performed more
Punch than MA in LD, (4) MAperformed more
Punch in VR than LD, (5) participants performedmore
Zoom + Kick in uncertain GC than in certain GC. ffect of Gameplay Uncertainty, Display Type, and Age on Virtual Reality Exergames CHI ’21, May 8–13, 2021, Yokohama, Japan
Figure 3: (a) Mean completion time on each monster’s life according to age group, (b) mean success rate of
Kick and
Zoom + Squat according to DT, and (c) mean success rate of
Zoom + Squat and
Zoom + Kick according to GC and Age. Error bars indicate ±2standard errors.Table 2: Three-way mixed ANOVA test results for success rate. Significant results where 𝑝 < . are shown in light green, 𝑝 < . in green, and 𝑝 < . in dark green. Punch , Age, DT × GC, DT × Age × GC have no significant results and therefore not shownfor better clarity. No sig indicates no significant results.
Kick Zoom + Squat Zoom + Kick DT 𝐹 , = . , 𝑝 < . , 𝜂 𝑝 = . 𝐹 , = . , 𝑝 < . , 𝜂 𝑝 = .
324 No sigGC No sig 𝐹 , = . , 𝑝 < . , 𝜂 𝑝 = .
421 No sigDT × Age No sig 𝐹 , = . , 𝑝 < . , 𝜂 𝑝 = . 𝐹 , = . , 𝑝 < . , 𝜂 𝑝 = . 𝐹 , = . , 𝑝 < . , 𝜂 𝑝 = . DT : VR > LD ( 𝑝 < .
5; see Figure 3b) GC : uncertain < certain ( 𝑝 < . YA : VR > LD ( 𝑝 < . VR : YA > MA ( 𝑝 < .
05; see Figure 4a); YA :VR > LD ( 𝑝 < .
05; see Figure 4a);
Uncer-tain : YA > MA ( 𝑝 < .
05; see Figure 3c)
Table 3: Three-way mixed ANOVA test results for the total number of gestures performed. Significant results where 𝑝 < . are shown in light green, 𝑝 < . in green, and 𝑝 < . in dark green. Kick , DT, GC × DT, Age × GC × DT have no significantresults and therefore not shown for better clarity. No sig indicates no significant results.
Punch Zoom + Squat Zoom + Kick
GC No sig 𝐹 , = . , 𝑝 < . , 𝜂 𝑝 = . 𝐹 , = . , 𝑝 < . , 𝜂 𝑝 = . 𝐹 , = . , 𝑝 < . , 𝜂 𝑝 = . 𝐹 , = . , 𝑝 < . , 𝜂 𝑝 = .
383 No sigGC × Age No sig 𝐹 , = . , 𝑝 < . , 𝜂 𝑝 = .
235 No sigDT × Age 𝐹 , = . , 𝑝 < . , 𝜂 𝑝 = .
142 No sig No sigPost-hoc LD : YA > MA ( 𝑝 < .
01; see Figure 4b); MA : VR > LD ( 𝑝 < .
01; see Figure 4b)
YA and MA : uncertain > certain (both 𝑝 < . Uncertain andcertain : MA > YA (both 𝑝 < . GC : uncertain > certain ( 𝑝 < .
05; see Fig-ure 4c)
Game Experience . ANOVA tests yielded a signif-icant effect of Age on competence ( 𝐹 , = . , 𝑝 < . , 𝜂 𝑝 = . 𝐹 , = . , 𝑝 < . , 𝜂 𝑝 = . 𝐹 , = . , 𝑝 < . , 𝜂 𝑝 = . 𝐹 , = . , 𝑝 < . , 𝜂 𝑝 = . 𝐹 , = . , 𝑝 < . , 𝜂 𝑝 = . 𝐹 , = . , 𝑝 < . , 𝜂 𝑝 = . 𝐹 , = . , 𝑝 < . , 𝜂 𝑝 = . 𝑝 < . 𝑝 < .
05) and MA ( 𝑝 < . HI ’21, May 8–13, 2021, Yokohama, Japan Xu et al.
Figure 4: (a) Mean success rate of
Zoom + Kick and
Zoom + Squat according to DT and Age, (b) mean total number of
Punch performed according to DT and Age, and (c) mean total number of
Zoom + Kick and
Zoom + Squat performed according to GCand Age. Error bars indicate ±2 standard errors.Figure 5: (a) Game experience questionnaire rating of subscales according to Age, (b) mean flow rating according to DT andAge, and (c) mean nausea and oculomotor rating according to Age. Error bars indicate ±2 standard errors.
Simulator Sickness . ANOVA tests yielded a significant effect ofAge on nausea ( 𝐹 , = . , 𝑝 < . , 𝜂 𝑝 = . 𝐹 , = . , 𝑝 < . , 𝜂 𝑝 = . 𝐹 , = . , 𝑝 = . , 𝜂 𝑝 = . Uncertain Elements’ Ratings . We employed a two-way mixedANOVA with Elements (false-attack, hit, miss) as the within-subjectsvariable and Age as the between-subjects variable. The ANOVAtests yielded a significant effect of Elements ( 𝐹 . , . = . , 𝑝 < . , 𝜂 𝑝 = . 𝐹 . , . = . , 𝑝 = . 𝐹 , = . , 𝑝 < . , 𝜂 𝑝 = . Table 4 shows the ANOVA tests of all exertionmeasures. In summary, (1) YA had lower avgHR% than MA in uncertain GC, (2) MA had a higher avgHR% in uncertain GC thancertain GC, (3) participants burned more calories in uncertain GCthan certain GC, (4) MA participants burned more calories thanYA participants (see Figure 6b), (5) Borg RPE for uncertain GC washigher than certain GC among YA and MA, (6) the Borg RPE forYA was higher than MA in certain GC and uncertain GC.
The VR uncertain version wasrated the best version among the four versions by 23 participants(12 YA). Only 5 participants (4 YA) selected VR certain as their topoption and 4 MA chose LD uncertain version as their top selection.
Feedback . From the coded transcripts, three main themes emerged(element of the games, general gaming experience, and exercisingfor health) from the two researchers, who first reviewed the tran-scripts independently. They were agreed by a third researcher aftera second discussion. Thirty-two participants were labeled P1-P16(YA group) and P17-P32 (MA group).Overall, both user groups perceived the game as “ enjoyable ” (10YA, 9 MA), “ novel ” (9 YA, 8 MA), and “ good for their health ” (9 YA,14 MA) and none of them perceive anything that was confusing inthe game. Both groups perceived the false-attacks more difficultthan expected (P3, P13, P20, P22, P24-27), but only MA participants ffect of Gameplay Uncertainty, Display Type, and Age on Virtual Reality Exergames CHI ’21, May 8–13, 2021, Yokohama, Japan
Table 4: Three-way mixed ANOVA test results for exertion measurements. Significant results where 𝑝 < . are shown in lightgreen, 𝑝 < . in green, and 𝑝 < . in dark green. DT, GC × DT, Age × DT, Age × GC × DT have no significant results andtherefore not shown for better clarity. No sig indicates no significant results. avgHR% Calories Burned Borg RPEGC 𝐹 , = . , 𝑝 < . , 𝜂 𝑝 = . 𝐹 , = . , 𝑝 < . , 𝜂 𝑝 = . 𝐹 , = . , 𝑝 < . , 𝜂 𝑝 = . 𝐹 , = . , 𝑝 < . , 𝜂 𝑝 = . 𝐹 , = . , 𝑝 < . , 𝜂 𝑝 = . 𝐹 , = . , 𝑝 < . , 𝜂 𝑝 = . 𝐹 , = . , 𝑝 < . , 𝜂 𝑝 = .
248 No sig 𝐹 , = . , 𝑝 < . , 𝜂 𝑝 = . Uncertain : YA < MA (both 𝑝 < . MA : uncertain > cer-tain ( 𝑝 < . GC : uncertain > certain ( 𝑝 < . Age : MA > YA ( 𝑝 < .
01; seeFigure 6b) YA : uncertain > certain ( 𝑝 < .
01; see Figure6c); MA : uncertain > certain ( 𝑝 < . Certain : YA > MA ( 𝑝 < . Uncertain : YA > MA ( 𝑝 < .
01; see Figure 6c)
Figure 6: (a) Mean avgHR% according to GC and Age, (b) mean calories burned, and (c) mean Borg RPE rating according to GCand Age. Error bars indicate ±2 standard errors. mentioned that sometimes they could not perform the defense movein time.Regarding the elements that they liked about the game, the com-ments from the two groups came from two different perspectives.Most YA focused on the game elements (e.g., “ the false-attack by theopponents ” [P3, P14, P16], “ critical hits ” [P5], “ misses ” [P11], “ usinggestures to trigger attacks are fun and easy to understand ” [P6, P9,P13]) while only a few mentioned about the health benefits as theirpreferred elements (P8, P10, P15). This is a completely different forthe MA, where 13 MA mentioned they liked the game because itcould be a good exercise activity while only 6 comments focused ondesign elements (e.g., “ false-attacks by the monster is a good design ”[P23, P27, P30], “ it tricks me into performing defense moves, which isgood for my health ” [P20, P24, P25]).The two generations focused on the different perspectives againregarding the elements that they did not like. Most comments fromYA were about the graphics and models used in the game, that theyshould be improved and more moves could be added. On the otherhand, most MA believed that the uncertain elements are sometimesoverused, which caused them to perform too many defense movesand made them feeling exhausted during the game.
In general, the performance (i.e., completion times, success ratesfor both attack and defense moves) of middle-aged adults wereworse than young adults in our motion-based first-person exergame,which is in line with previous studies of similar games [26]. Onepossible reason could be age-related declines in mobility; for in-stance, middle-aged adults typically require more time to performgestures [21]. They also were not able to react to the monster’sattack sometimes or cancel their defense moves when realizing thatthe monster was performing false-attacks; for example, P20, P22,P24-25, P27-28: “
I could not react in time .” Hence, it is necessaryto take into account age-related declines (e.g., working memory[49], grip strength [40], and muscle mass [10]) when designingexergames for middle-aged adults.In addition, the two age groups perceived the game experiencedifferently. We found that young adults were more immersive (im-mersion, flow) in the game than middle-aged adults and had a higherpositive emotion, efficacy, competence. However, young adults stillfelt more annoyed and experienced more negative emotions thanmiddle-aged adults even though they had a better performance (e.g.,the successful attack rate is much higher). One possible reason isthat young adults might have expected that they should performmuch better due to their competitive expectations of themselves and
HI ’21, May 8–13, 2021, Yokohama, Japan Xu et al. the game, while the competition was downplayed in middle-agedadults [73].Previous research has suggested that there may be a decline insusceptibility to VR sickness as people age [4]. Our results also sup-port this, as we found that young adults felt sicker during gameplaythan middle-aged adults. Overall, sickness level for all participantswere either negligible or very low, with no participants experienc-ing severe simulator sickness. That is, all participants had no issuesin playing the game.Existing literature in the exercise domain (e.g., tai chi [42], armstraining [28], arm abduction [61]) have suggested that age doesnot affect the exertion level of the exercise. However, this is notsupported by our results because we found that our two groupsof participants produced different levels of exertion (middle-agedadults had a higher avgHR% in the uncertain condition and burnedmore calories than young adults but gave a lower Borg RPE ratings).Further study is required to explain this.
Our results suggest that participants had a better performance inVR (i.e., higher success rates in attack and defense moves in VR thanLD). This is understandable because the greater flow experiencebrought by VR to the players had a positive effect on performancein the game [1]. A previous study [84] that also focused on theeffect of DT versus VR showed that VR could provide a greaterpositive game experience (e.g., challenge, flow, immersion) to theplayers than LD, which was also found in our results (i.e., VR led toa higher flow rating than LD). Existing literature also indicated thatgame experience (from GEQ) could be perceived the same in bothVR and LD [83]. One reason could be that in [83], participants onlyexperienced 4 minutes of gameplay, which is relatively short fordeveloping a fuller picture of the technologies. Hence, we suggestthat future studies consider a longer game duration, like 7- 8 min-utes used in our research and in [84], to let the players experiencea game in each technology more fully.In addition, our findings indicate there was no significant differ-ence regarding the level of sickness that participants experiencedbetween VR and LD when playing the motion-based exergame,which is in line with [83] but not [84] where researchers reportedthat playing a motion-based exergame in VR could lead to a highersickness than LD. One possible explanation could be that the typeof game used in the experiment was different. Our game and thegame used in [83] involved more interaction with the virtual worldthan the game in [84]. For instance, players had direct contactswith the virtual objects (either through attacking and defendingagainst the monster in our game or directly using the hands or feetto hit the objects in the game from [83]), which is not the case for[84] where the gestures performed by the users did not have directcontact with the virtual objects in the form of cubes.
The purpose of the design of false-attacks, one uncertain elementin our exergame, was to trick the players into using the defensemoves. Our results show that this element achieved its intended goalbecause participants performed more defense moves (
Zoom + Squat )in the uncertain condition than the counterpart condition. We also observed during the experiment that this design tricked all playersacross both groups.In addition, the design of misses had also forced them to performmore attack moves in their attempts to kill the monster. Hence,participants had a higher exertion level (i.e., avgHR%—MA, caloriesburned, Borg RPE) in the uncertain condition. Furthermore, whatis interesting to note is that participants did not feel a worse ex-perience by these design features since (1) they did not complainabout the features, and (2) the gameplay experience and sicknessin both game conditions were not significantly different. Therefore,we believe that involving uncertain elements (i.e., false-attacks,misses, and critical hits) in the type of exergame similar to ourscould increase players’ energy costs without incurring negativegameplay experiences in both VR and LD.
As our results show, theproposed uncertain elements in our exergame could be useful inenhancing exertion levels during game sessions. We list with ex-amples of how these uncertain elements can be applied to otherexergames.For sports exergames, false-attack can be used in several ways.For example, in the boxing game
Creed: Rise to Glory , a false-attackcan be directly applied to Creed’s attack strategy to trick playersinto making defense moves. False-attacks can be enhanced furtherby following a real attack after the animation of a false-attack. For Eleven Table Tennis VR , this can be added as a way for NPC topretend they want to move into one direction but not moving intothat direction. This type of false moves can be used in designingbasketball and football exergames where trickery is a key to make adefending player go into one direction so that the player can moveinto the opposite way (e.g., Kinect Sports: Soccer ).For exergames that involve one-way interaction with the enemy(i.e., player to NPC), critical hits and misses can be used. For instance,in the tower defense game Longbowman [79], critical hits and missescan be designed with additional features. A critical hit can dealadditional damage and also slow down the movement of the enemy.In contrast, a miss does not damage the enemy and would makethe enemy become angry and move faster.For exergames that involve two-way interaction with the enemy(player to NPC and NPC to player), all three elements can be used.For instance, in
Ring Fit Adventure , a motion-based active game forthe Nintendo Switch, all these elements can be added in a similarway that we did in our exergame since it is designed based on thiscommercial game.
Like older adults [73], middle-aged adults believe that exergamesare helpful to their health. We suggest making the potential healthbenefits to middle-aged adults explicit and clear inside the gameand as part of the gameplay experience. For instance, designerscould (1) introduce the benefits of each gesture before the game,(2) present the energy cost like calories burned during the game as https://marketplace.xbox.com/en-US/Product/Kinect-Sports/66acd000-77fe-1000-9115-d8024d5308c9 ffect of Gameplay Uncertainty, Display Type, and Age on Virtual Reality Exergames CHI ’21, May 8–13, 2021, Yokohama, Japan part of any dynamic visual and audio feedback, (3) give a summaryreport of the overall performance (e.g., for each type of gestures,providing the total number the player performed) after the game. There are some limitations in this research, which can also serveas directions for the future. One limitation is that we tested threeelements of uncertainty (false-attacks, misses, and critical hits) thatcovers four uncertainty sources. Future work could explore moreuncertainty sources [12] in motion-based exergames. For example,we can use (1) analytical complexity , by allowing more skills forthe player but require the player to kill the monster in a limitedtime so that the player needs to analyze the best strategy to fightagainst their opponent carefully. It is possible to integrate (2) hiddeninformation , by not showing information of the opponent’s attackmoves. Addition, (3) narrative anticipation can be used by adding astoryline to a game and fighting an opponent would reward themwith the corresponding piece of the storyline. By doing this, theplayer has the desire to know the next piece of the storyline [56].In addition, there are some limitations related to the choice ofVR HMD and exergames in current commercial VR HMDs. We usedthe Oculus Rift CV1. Newer VR HMDs (i.e., VIVE Pro Eye) thatcome with a higher resolution could impact simulator sickness andgame experience. We used the Oculus Rift CV1 because we wantedto have consistency with prior studies [83, 84]. The Rift CV1, asa tethered helmet, has a limited range of motion because of theattached cables. While standalone devices like Oculus Quest donot have this limitation, they suffer from latency issues when usedwith external motion sensors (i.e., Kinect) to capture motion data.In addition, long gameplay sessions wearing any current HMDscould result in sweats in the glasses; thus, the length of gameplayshould be carefully designed to prevent this issue. Also, to makeMA-friendly exergames, future games should involve more simplegestures (like zoom—hands stretching out, hands-up) to eliminateany risks when wearing a VR HMD.Our study only involved a single session. Longer-term studieswill be useful to determine if the same results hold and to determineadditional effects that may come with long-term exposures. Inaddition, due to the COVID-19, we cannot to include the elderlyadults (i.e., those 65 years old and above) in the experiment. Futurework could have all these three groups of adults (i.e., young, middle-aged, elderly) to assess their relative performance and experiencewith exergames.
In this research, we have investigated the effect of display type(virtual reality and large display) with or without elements of un-certainty in motion-based first-person perspective exergames. Wealso have explored the impact of age by comparing game perfor-mance, gameplay experience, and level of energy exertion betweenyoung adults and middle-aged adults. Our results suggest the fol-lowing three conclusions: (1) For the type of exergame like ours,virtual reality could improve game performance while maintainingthe same level of sickness as large displays. (2) Uncertain elementslike those used in this research’s motion-based exergame might nothelp enhance the overall game experience, but are instrumental in increasing exertion levels, which is one of the essential features ofexergames. (3) Exergames for middle-aged adults should be care-fully designed with consideration to age-related declines, similar toolder adults. We also proposed two main design guidelines whichcan pave the way for improving the acceptability of VR exergamesamong young and middle-aged adults.
ACKNOWLEDGMENTS
The authors would like to thank the anonymous reviewers for theirvaluable comments and helpful suggestions and the CommitteeMember who guided the revision of our paper. The work is sup-ported in part by Xi’an Jiaotong-Liverpool University (XJTLU) KeySpecial Fund (KSF-A-03) and XJTLU Research Development Fund.
REFERENCES [1] Wilfried Admiraal, Jantina Huizenga, Sanne Akkerman, and Geert ten Dam. 2011.The concept of flow in collaborative game-based learning.
Group Awareness inCSCL Environments
27, 3 (May 2011), 1185–1194. https://doi.org/10.1016/j.chb.2010.12.013[2] Valentina Andries and Judy Robertson. 2019. Designing Social Play to SupportYoung Hospitalised Children. In
Proceedings of the 18th ACM International Confer-ence on Interaction Design and Children (IDC ’19) . Association for Computing Ma-chinery, New York, NY, USA, 550–555. https://doi.org/10.1145/3311927.3325317event-place: Boise, ID, USA.[3] Soumya C. Barathi, Eamonn O’Neill, Christof Lutteroth, Daniel J. Finnegan,Matthew Farrow, Alexander Whaley, Pippa Heath, Jude Buckley, Peter W.Dowrick, Burkhard C. Wuensche, and James L. J. Bilzon. 2018. InteractiveFeedforward for Improving Performance and Maintaining Intrinsic Motivationin VR Exergaming. In
Proceedings of the 2018 CHI Conference on Human Fac-tors in Computing Systems - CHI ’18 . ACM Press, Montreal QC, Canada, 1–14.https://doi.org/10.1145/3173574.3173982[4] Aurore Bardey, Gaëlle Quarck, Paolino F., Pierre Denise, Michel Paolino, JohnGolding, and Venera GHULYAN-BEDIKIAN. 2013. Motion sickness susceptibilityin healthy subjects and vestibular patients: effects of gender, age and trait-anxiety.
Journal of Vestibular Research
23, 4-5 (Nov. 2013), 203–209. https://doi.org/10.3233/VES-130501[5] Shlomo Berkovsky, Mac Coombe, Jill Freyne, Dipak Bhandari, and Nilufar Baghaei.2010. Physical Activity Motivating Games: Virtual Rewards for Real Activity. In
Proceedings of the SIGCHI Conference on Human Factors in Computing Systems (Atlanta, Georgia, USA) (CHI ’10) . Association for Computing Machinery, NewYork, NY, USA, 243–252. https://doi.org/10.1145/1753326.1753362[6] Max V. Birk, Maximilian A. Friehs, and Regan L. Mandryk. 2017. Age-BasedPreferences and Player Experience: A Crowdsourced Cross-Sectional Study. In
Proceedings of the Annual Symposium on Computer-Human Interaction in Play(CHI PLAY ’17) . Association for Computing Machinery, New York, NY, USA,157–170. https://doi.org/10.1145/3116595.3116608 event-place: Amsterdam, TheNetherlands.[7] Kenneth Blocker, Timothy Wright, and Walter Boot. 2014. Gaming Preferencesof Aging Generations.
Gerontechnology
12, 3 (June 2014), 174–184. https://doi.org/10.4017/gt.2014.12.3.008.00[8] Ian Bogost. 2005. The Rhetoric of Exergaming. In
Proceedings of the Digital Artsand Cultures (DAC) . Copenhagen, Denmark, 10. https://doi.org/10.1.1.500.1480[9] Gunnar Borg. 1982. Psychophysical bases of perceived exertion.
Medicine &Science in Sports & Exercise
14, 5 (1982), 377–381. insights.ovid.com[10] Marybeth Brown and Eileen M Hasser. 1996. Complexity of age-related changein skeletal muscle.
The Journals of Gerontology Series A: Biological Sciences andMedical Sciences
51, 2 (1996), B117–B123. Publisher: The Gerontological Societyof America.[11] Roger Caillois. 2001.
Man, Play and Games (reprint edition ed.). University ofIllinois Press, Urbana.[12] Greg Costikyan. 2013.
Uncertainty in games . Mit Press.[13] Bob De Schutter. 2011. Never Too Old to Play: The Appeal of Digital Gamesto an Older Audience.
Games and Culture
6, 2 (May 2011), 155–170. https://doi.org/10.1177/1555412010364978 Publisher: SAGE Publications.[14] Bob De Schutter and Vero Vanden Abeele. 2010. Designing Meaningful Playwithin the Psycho-Social Context of Older Adults. In
Proceedings of the 3rdInternational Conference on Fun and Games (Fun and Games ’10) . Associationfor Computing Machinery, New York, NY, USA, 84–93. https://doi.org/10.1145/1823818.1823827 event-place: Leuven, Belgium.[15] Bernie DeKoven. 2002.
The Well-Played Game: A Playful Path to Wholeness .iUniverse, San Jose.
HI ’21, May 8–13, 2021, Yokohama, Japan Xu et al. [16] Anders Drachen, Pejman Mirza-Babaei, and Lennart Nacke (Eds.). 2018.
GamesUser Research . Oxford University Press, Oxford, UK ; New York.[17] Henry Been-Lirn Duh, Sharon Lynn Chu Yew Yee, Yuan Xun Gu, and VivianHsueh-Hua Chen. 2010. A Narrative-Driven Design Approach for Casual Gameswith Children. In
Proceedings of the 5th ACM SIGGRAPH Symposium on VideoGames (Sandbox ’10) . Association for Computing Machinery, New York, NY,USA, 19–24. https://doi.org/10.1145/1836135.1836138 event-place: Los Angeles,California.[18] Patrick Eggenberger, Vera Schumacher, Marius Angst, Nathan Theill, and Eling Dde Bruin. 2015. Does multicomponent physical exercise with simultaneous cogni-tive training boost cognitive performance in older adults? A 6-month randomizedcontrolled trial with a 1-year follow-up.
Clinical interventions in aging
10 (Aug.2015), 1335–1349. https://doi.org/10.2147/CIA.S87732[19] Eva Eriksson, Gökçe Elif Baykal, Staffan Björk, and Olof Torgersson. 2019. UsingGameplay Design Patterns with Children in the Redesign of a CollaborativeCo-Located Game. In
Proceedings of the 18th ACM International Conference onInteraction Design and Children (IDC ’19) . Association for Computing Machinery,New York, NY, USA, 15–25. https://doi.org/10.1145/3311927.3323155 event-place:Boise, ID, USA.[20] Daniel Ferreira, Rut Correia, Antonieta Nieto, Alejandra Machado, YaizaMolina Rodríguez, and José Barroso. 2015. Cognitive decline before the ageof 50 can be detected with sensitive cognitive measures.
Psicothema
27, 3 (May2015), 216–222. https://doi.org/10.7334/psicothema2014.192[21] Luigi Ferrucci, Rachel Cooper, Michelle Shardell, Eleanor Simonsick, JenniferSchrack, and Diana Kuh. 2016. Age-Related Change in Mobility: PerspectivesFrom Life Course Epidemiology and Geroscience.
The Journals of GerontologySeries A: Biological Sciences and Medical Sciences
71, 9 (March 2016), 1184–1194.https://doi.org/10.1093/gerona/glw043[22] Brian J. Gajadhar, Yvonne A. W. de Kort, and Wijnand A. IJsselsteijn. 2008. SharedFun Is Doubled Fun: Player Enjoyment as a Function of Social Setting. In
Fun andGames , Panos Markopoulos, Boris de Ruyter, Wijnand IJsselsteijn, and DuncanRowland (Eds.). Springer Berlin Heidelberg, Berlin, Heidelberg, 106–117. https://doi.org/10.1007/978-3-540-88322-7_11[23] Brian J. Gajadhar, Henk Herman Nap, Yvonne A. W. de Kort, and Wijnand A.IJsselsteijn. 2010. Out of Sight, out of Mind: Co-Player Effects on Seniors’ PlayerExperience. In
Proceedings of the 3rd International Conference on Fun and Games(Fun and Games ’10) . Association for Computing Machinery, New York, NY, USA,74–83. https://doi.org/10.1145/1823818.1823826 event-place: Leuven, Belgium.[24] Yue Gao and Regan Mandryk. 2012. The acute cognitive benefits of casualexergame play. In
Proceedings of the 2012 ACM annual conference on HumanFactors in Computing Systems - CHI ’12 . ACM Press, Austin, Texas, USA, 1863.https://doi.org/10.1145/2207676.2208323[25] Kathrin Gerling, Ian Livingston, Lennart Nacke, and Regan Mandryk. 2012. Full-body motion-based game interaction for older adults. In
Proceedings of the 2012ACM annual conference on Human Factors in Computing Systems - CHI ’12 . ACMPress, Austin, Texas, USA, 1873. https://doi.org/10.1145/2207676.2208324[26] Kathrin M. Gerling, Kristen K. Dergousoff, and Regan L. Mandryk. 2013. Is Move-ment Better? Comparing Sedentary and Motion-Based Game Controls for OlderAdults. In
Proceedings of Graphics Interface 2013 (GI ’13) . Canadian InformationProcessing Society, CAN, 133–140. event-place: Regina, Sascatchewan, Canada.[27] Paul E. Greenberg, Tamar Sisitsky, Ronald C. Kessler, Stan N. Finkelstein, Ernst R.Berndt, Jonathan R. T. Davidson, James C. Ballenger, and Abby J. Fyer. 1999.The economic burden of anxiety disorders in the 1990s.
The Journal of ClinicalPsychiatry
60, 7 (1999), 427–435. https://doi.org/10.4088/JCP.v60n0702 Place:US Publisher: Physicians Postgraduate Press.[28] Alain Groslambert, Céline C Grange, Stéphane Perrey, Jérôme Maire, NicolasTordi, and Jean Denis Rouillon. 2006. Effects of Aging on Perceived Exertion andPain During Arm Cranking in Women 70 to 80 YEARS OLD.
Journal of sportsscience & medicine
5, 2 (June 2006), 208–214. https://pubmed.ncbi.nlm.nih.gov/24259993 Publisher: Asist Group.[29] Sandro Hardy, Tim Dutz, Josef Wiemeyer, Stefan Göbel, and Ralf Steinmetz.2015. Framework for personalized and adaptive game-based training programsin health sport.
Multimedia Tools and Applications
74, 14 (July 2015), 5289–5311.https://doi.org/10.1007/s11042-014-2009-z[30] Hamilton A. Hernandez, Zi Ye, T.C. Nicholas Graham, Darcy Fehlings, and LaurenSwitzer. 2013. Designing action-based exergames for children with cerebral palsy.In
SIGCHI Conference on Human Factors in Computing Systems . ACM Press, Paris,France, 1261–1270. https://doi.org/10.1145/2470654.2466164[31] Imali T. Hettiarachchi, Samer Hanoun, Darius Nahavandi, and Saeid Nahavandi.2019. Validation of Polar OH1 optical heart rate sensor for moderate and highintensity physical activities.
PLOS ONE
14, 5 (May 2019), e0217288. https://doi.org/10.1371/journal.pone.0217288[32] W. A. IJsselsteijn, Y. A. W. De Kort, and Karolien Poels. 2008. The game experiencequestionnaire.
Eindhoven: Technische Universiteit Eindhoven (2008).[33] Christos Ioannou, Patrick Archard, Eamonn O’Neill, and Christof Lutteroth. 2019.Virtual Performance Augmentation in an Immersive Jump & Run Exergame. In
Proceedings of the 2019 CHI Conference on Human Factors in Computing Systems- CHI ’19 . ACM Press, Glasgow, Scotland Uk, 1–15. https://doi.org/10.1145/ 3290605.3300388[34] Aliya Iskenderova, Florian Weidner, and Wolfgang Broll. 2017. Drunk VirtualReality Gaming: Exploring the Influence of Alcohol on Cybersickness. In
Pro-ceedings of the Annual Symposium on Computer-Human Interaction in Play (CHIPLAY ’17) . Association for Computing Machinery, New York, NY, USA, 561–572. https://doi.org/10.1145/3116595.3116618 event-place: Amsterdam, TheNetherlands.[35] Mark R. Johnson. 2018.
The Unpredictability of Gameplay . Bloomsbury Academic,New York.[36] Jesper Juul. 2011.
Half-Real: Video Games between Real Rules and Fictional Worlds .The MIT Press, Cambridge, Mass.[37] Robert S. Kennedy, Norman E. Lane, Kevin S. Berbaum, and Michael G. Lilienthal.1993. Simulator Sickness Questionnaire: An Enhanced Method for QuantifyingSimulator Sickness.
The International Journal of Aviation Psychology
3, 3 (July1993), 203–220. https://doi.org/10.1207/s15327108ijap0303_3[38] Sung Yeun (Su) Kim, Nathan Prestopnik, and Frank A. Biocca. 2014. Body inthe interactive game: How interface embodiment affects physical activity andhealth behavior change.
Computers in Human Behavior
36 (2014), 376 – 384.https://doi.org/10.1016/j.chb.2014.03.067[39] Jordan Koulouris, Zoe Jeffery, James Best, Eamonn O’Neill, and Christof Lutteroth.2020. Me vs. Super(Wo)Man: Effects of Customization and Identification in aVR Exergame. In
Proceedings of the 2020 CHI Conference on Human Factors inComputing Systems (CHI ’20) . Association for Computing Machinery, New York,NY, USA, 1–17. https://doi.org/10.1145/3313831.3376661 event-place: Honolulu,HI, USA.[40] Rumi Kozakai, Fujiko Ando, Heung Kim, Atsumu Yuki, Rei Otsuka, and HiroshiShimokata. 2016. Sex-differences in age-related grip strength decline: A 10-year longitudinal study of community-living middle-aged and older Japanese.
The Journal of Physical Fitness and Sports Medicine
5, 1 (March 2016), 87–94.https://doi.org/10.7600/jpfsm.5.87[41] Shringi Kumari, Sebastian Deterding, and Jonathan Freeman. 2019. The Role ofUncertainty in Moment-to-Moment Player Motivation: A Grounded Theory. In
Proceedings of the Annual Symposium on Computer-Human Interaction in Play (Barcelona, Spain) (CHI PLAY ’19) . Association for Computing Machinery, NewYork, NY, USA, 351–363. https://doi.org/10.1145/3311350.3347148[42] Ching Lan, Ssu-Yuan Chen, and Jin-Shin Lai. 2004. Relative Exercise Inten-sity of Tai Chi Chuan is Similar in Different Ages and Gender.
The AmericanJournal of Chinese Medicine
32, 1 (Jan. 2004), 151–160. https://doi.org/10.1142/S0192415X04001746 Publisher: World Scientific Publishing Co.[43] Thomas W. Malone. 1982. Heuristics for Designing Enjoyable User Interfaces:Lessons from Computer Games. In
Proceedings of the 1982 Conference on Hu-man Factors in Computing Systems (Gaithersburg, Maryland, USA) (CHI ’82) .Association for Computing Machinery, New York, NY, USA, 63–68. https://doi.org/10.1145/800049.801756[44] Elena Márquez Segura, Annika Waern, Jin Moen, and Carolina Johansson. 2013.The Design Space of Body Games: Technological, Physical, and Social Design. In
Proceedings of the SIGCHI Conference on Human Factors in Computing Systems (Paris, France) (CHI ’13) . Association for Computing Machinery, New York, NY,USA, 3365–3374. https://doi.org/10.1145/2470654.2466461[45] Joe Marshall, Florian ‘Floyd’ Mueller, Steve Benford, and Sebastiaan Pijnappel.2016. Expanding exertion gaming.
International Journal of Human-ComputerStudies
90 (2016), 1 – 13. https://doi.org/10.1016/j.ijhcs.2016.02.003[46] Anna Lisa Martin-Niedecken. 2018. Designing for Bodily Interplay: Engagingwith the Adaptive Social Exertion Game "Plunder Planet". In
Proceedings of the17th ACM Conference on Interaction Design and Children (Trondheim, Norway) (IDC ’18) . Association for Computing Machinery, New York, NY, USA, 19–30.https://doi.org/10.1145/3202185.3202740[47] Anna Lisa Martin-Niedecken and Elisa D. Mekler. 2018. The ExerCube: Participa-tory Design of an Immersive Fitness Game Environment. In
Serious Games , StefanGöbel, Augusto Garcia-Agundez, Thomas Tregel, Minhua Ma, Jannicke Baal-srud Hauge, Manuel Oliveira, Tim Marsh, and Polona Caserman (Eds.). SpringerInternational Publishing, Cham, 263–275.[48] Anna Lisa Martin-Niedecken, Katja Rogers, Laia Turmo Vidal, Elisa D. Mekler,and Elena Márquez Segura. 2019. ExerCube vs. Personal Trainer: Evaluatinga Holistic, Immersive, and Adaptive Fitness Game Setup. In
Proceedings of the2019 CHI Conference on Human Factors in Computing Systems (Glasgow, ScotlandUk) (CHI ’19) . Association for Computing Machinery, New York, NY, USA, 1–15.https://doi.org/10.1145/3290605.3300318[49] Yuko Meguro, Toshikatsu Fujii, Atsushi Yamadori, Takashi Tsukiura, KyokoSuzuki, Jiro Okuda, and Mariko Osaka. 2000. The Nature of Age-Related Declineon the Reading Span Task.
Journal of Clinical and Experimental Neuropsychology
22, 3 (June 2000), 391–398. https://doi.org/10.1076/1380-3395(200006)22:3;1-V;FT391 Publisher: Routledge.[50] Daniel R. Mestre, Virginie Dagonneau, and Charles-Symphorien Mercier. 2011.Does Virtual Reality Enhance Exercise Performance, Enjoyment, and Dissocia-tion? An Exploratory Study on a Stationary Bike Apparatus.
Presence: Teleopera-tors and Virtual Environments
20, 1 (Feb. 2011), 14. https://doi.org/10.1162/pres_a_00031 ffect of Gameplay Uncertainty, Display Type, and Age on Virtual Reality Exergames CHI ’21, May 8–13, 2021, Yokohama, Japan [51] Diego Monteiro, Hai-Ning Liang, Wenge Xu, Marvin Brucker, Vijayakumar Nan-jappan, and Yong Yue. 2018. Evaluating enjoyment, presence, and emulator sick-ness in VR games based on first- and third- person viewing perspectives.
ComputerAnimation and Virtual Worlds
29, 3 (2018), 12. https://doi.org/10.1002/cav.1830[52] Florian Mueller and Katherine Isbister. 2014. Movement-Based Game Guidelines.In
Proceedings of the SIGCHI Conference on Human Factors in Computing Systems (Toronto, Ontario, Canada) (CHI ’14) . Association for Computing Machinery,New York, NY, USA, 2191–2200. https://doi.org/10.1145/2556288.2557163[53] Florian Mueller, Rohit Ashok Khot, Kathrin Gerling, and Regan Mandryk. 2016.Exertion Games.
Found. Trends Hum.-Comput. Interact.
10, 1 (Dec. 2016), 1–86.https://doi.org/10.1561/1100000041[54] Florian Mueller, Frank Vetere, Martin Gibbs, Darren Edge, Stefan Agamanolis,Jennifer Sheridan, and Jeffrey Heer. 2012. Balancing Exertion Experiences. In
Proceedings of the SIGCHI Conference on Human Factors in Computing Systems (Austin, Texas, USA) (CHI ’12) . Association for Computing Machinery, New York,NY, USA, 1853–1862. https://doi.org/10.1145/2207676.2208322[55] Florian “Floyd” Mueller, Darren Edge, Frank Vetere, Martin R. Gibbs, StefanAgamanolis, Bert Bongers, and Jennifer G. Sheridan. 2011. Designing Sports:A Framework for Exertion Games. In
Proceedings of the SIGCHI Conference onHuman Factors in Computing Systems (CHI ’11) . Association for ComputingMachinery, New York, NY, USA, 2651–2660. https://doi.org/10.1145/1978942.1979330 event-place: Vancouver, BC, Canada.[56] Elizabeth L. Murnane, Xin Jiang, Anna Kong, Michelle Park, Weili Shi, ConnorSoohoo, Luke Vink, Iris Xia, Xin Yu, John Yang-Sammataro, Grace Young, JennyZhi, Paula Moya, and James A. Landay. 2020. Designing Ambient Narrative-BasedInterfaces to Reflect and Motivate Physical Activity. In
Proceedings of the 2020CHI Conference on Human Factors in Computing Systems (CHI ’20) . Associationfor Computing Machinery, New York, NY, USA, 1–14. https://doi.org/10.1145/3313831.3376478 event-place: Honolulu, HI, USA.[57] Lennart E. Nacke, Anne Nacke, and Craig A. Lindley. 2009. Brain Training forSilver Gamers: Effects of Age and Game Form on Effectiveness, Efficiency, Self-Assessment, and Gameplay Experience.
CyberPsychology & Behavior
12, 5 (2009),493–499. https://doi.org/10.1089/cpb.2009.0013 PMID: 19772440.[58] Bjarne. M Nes, Imre Janszky, Ulrik Wisløff, Asbjørn Støylen, and Trine Karlsen.2013. Age-predicted maximal heart rate in healthy subjects: The HUNT FitnessStudy.
Scandinavian Journal of Medicine & Science in Sports
23, 6 (2013), 697–704.https://doi.org/10.1111/j.1600-0838.2012.01445.x[59] Yechiam Ostchega, Kathryn Porter, Jeffery Hughes, Charles F Dillon, and TatianaNwankwo. 2011. Resting pulse rate reference data for children, adolescents, andadults: United States, 1999-2008.
National health statistics reports
41 (Aug. 2011),1–16.[60] Marco Pasch, Nadia Bianchi-Berthouze, Betsy van Dijk, and Anton Nijholt.2009. Movement-based sports video games: Investigating motivation andgaming experience.
Entertainment Computing
1, 2 (2009), 49 – 61. https://doi.org/10.1016/j.entcom.2009.09.004 Intelligent Technologies for InteractiveEntertainment.[61] Danny Pincivero, Mark Timmons, and D Elsing. 2010. RPE Angle Effects inYoung and Middle-Aged Adults.
International journal of sports medicine
31, 4 (Feb.2010), 257–260. https://doi.org/10.1055/s-0030-1247551[62] Thomas G. Plante, Arianna Aldridge, Denise Su, Ryan Bogdan, Martha Belo, andKamran Kahn. 2003. Does Virtual Reality Enhance the Management of StressWhen Paired With Exercise? An Exploratory Study.
International Journal of StressManagement
10, 3 (2003), 203–216. https://doi.org/10.1037/1072-5245.10.3.203[63] Christopher Power, Paul Cairns, Alena Denisova, Themis Papaioannou, and RuthGultom. 2019. Lost at the edge of uncertainty: Measuring player uncertainty indigital games.
International Journal of Human–Computer Interaction
35, 12 (2019),1033–1045. https://doi.org/10.1080/10447318.2018.1507161 Publisher: Taylor &Francis.[64] Albert Skip Rizzo, Belinda Lange, Evan A. Suma, and Mark Bolas. 2011. VirtualReality and Interactive Digital Game Technology: New Tools to Address Obesityand Diabetes.
Journal of Diabetes Science and Technology
5, 2 (March 2011),256–264. https://doi.org/10.1177/193229681100500209[65] Daniel M Rosney and Peter J Horvath. 2018. Exergaming Intervention in Seden-tary Middle-Aged Adults Improves Cardiovascular Endurance, Balance andLower Extremity Functional Fitness.
Health Science Journal
12, 6 (2018), 1–10.https://doi.org/10.4172/2165-7904-C1-55 Publisher: Technological EducationalInstitute of Athens.[66] Liz Sanders and Pieter Jan Stappers. 2013.
Convivial Toolbox: Generative Researchfor the Front End of Design (illustrated edition ed.). Laurence King Publishing,Amsterdam.[67] Matt Schubert, Amy Clark, and Annie Rosa. 2018. The Polar OH1 Optical HeartRate Sensor is Valid during Moderate-Vigorous Exercise.
Sports Medicine Inter-national Open
2, 3 (March 2018), 67–70. https://doi.org/10.1055/a-0631-0920[68] Dwayne P. Sheehan and Larry Katz. 2013. The effects of a daily, 6-week exergam-ing curriculum on balance in fourth grade children.
Journal of Sport and HealthScience
2, 3 (Sept. 2013), 131–137. https://doi.org/10.1016/j.jshs.2013.02.002 [69] John L. Sherry, Kristen Lucas, Bradley S. Greenberg, and Ken Lachlan. 2006. VideoGame Uses and Gratifications as Predicators of Use and Game Preference. In
Playing video games: Motives, responses, and consequences.
Lawrence ErlbaumAssociates, New York, US, 213–224. Place: Mahwah, NJ, US Publisher: LawrenceErlbaum Associates Publishers.[70] Jeff Sinclair, Philip Hingston, and Martin Masek. 2007. Considerations for thedesign of exergames. In
Proceedings of the 5th international conference on Computergraphics and interactive techniques in Australia and Southeast Asia - GRAPHITE’07 . ACM Press, Perth, Australia, 289. https://doi.org/10.1145/1321261.1321313[71] Jan Smeddinck, Sandra Siegel, and Marc Herrlich. 2013. Adaptive Difficulty inExergames for Parkinson’s Disease Patients. In
Proceedings of Graphics Interface2013 (GI ’13) . Canadian Information Processing Society, CAN, 141–148. event-place: Regina, Sascatchewan, Canada.[72] Antônio Soares, Noé Borges Júnior, Marcelo Hounsell, Elessandra Marcelino,Gabriel Rossito, and Yoshimasa Sagawa. 2016. A serious game developed forphysical rehabilitation of frail elderly.
European Research in Telemedicine / LaRecherche Européenne en Télémédecine
5, 2 (June 2016), 45–53. https://doi.org/10.1016/j.eurtel.2016.05.003[73] Sruti Subramanian, Yngve Dahl, Nina Skjæret Maroni, Beatrix Vereijken, andDag Svanæs. 2019. Assessing Motivational Differences Between Young and OlderAdults When Playing an Exergame.
Games for Health Journal
9, 1 (Sept. 2019),24–30. https://doi.org/10.1089/g4h.2019.0082 Publisher: Mary Ann Liebert, Inc.,publishers.[74] Katie Salen Tekinbas and Eric Zimmerman. 2003.
Rules of Play: Game DesignFundamentals . The MIT Press, Cambridge, Mass.[75] Scott Thomas, Jeff Reading, and Roy Shephard. 1993. Revision of the PhysicalActivity Readiness Questionnaire (PAR-Q).
Canadian journal of sport sciences
Computers in Human Behavior
57 (2016), 267 – 273. https://doi.org/10.1016/j.chb.2015.12.033[77] Paul Verhaeghen and Timothy Salthouse. 1997. Meta-analyses of age-cognitionrelations in adulthood: Estimates of linear and nonlinear age effects and structuralmodels.
Psychological bulletin
Games for Health Journal
6, 4 (June 2017), 237–248. https://doi.org/10.1089/g4h.2017.0005 Publisher: Mary Ann Liebert, Inc.,publishers.[79] Biao. Xie, Yongqi. Zhang, Haikun. Huang, Elisa. Ogawa, Tongjian. You, andLap. Fai. Yu. 2018. Exercise Intensity-Driven Level Design.
IEEE Transactions onVisualization and Computer Graphics
24, 4 (2018), 1661–1670. https://doi.org/10.1109/TVCG.2018.2793618[80] Wenge Xu, Hai-Ning Liang, Nilufar Baghaei, Bing Wu Berberich, and YongYue. 2020. Health Benefits of Digital Videogames for the Aging Population:A Systematic Review.
Games for Health Journal
9, 6 (2020), 389–404. https://doi.org/10.1089/g4h.2019.0130 PMID: 32589482.[81] Wenge Xu, Hai-Ning Liang, Qiuyu He, Xiang Li, Kangyou Yu, and Yuzheng Chen.2020. Results and Guidelines From a Repeated-Measures Design ExperimentComparing Standing and Seated Full-Body Gesture-Based Immersive VirtualReality Exergames: Within-Subjects Evaluation.
JMIR Serious Games
8, 3 (2020),15. https://doi.org/10.2196/17972[82] Wenge Xu, Hai-Ning Liang, Xiaoyue Ma, and Xiang Li. 2020. VirusBoxing: AHIIT-Based VR Boxing Game. In
Extended Abstracts of the 2020 Annual Symposiumon Computer-Human Interaction in Play (Virtual Event, Canada) (CHI PLAY ’20) .Association for Computing Machinery, New York, NY, USA, 98–102. https://doi.org/10.1145/3383668.3419958[83] Wenge Xu, Hai-Ning Liang, Yifan Yu, Diego Monteiro, Khalad Hasan, and CharlesFleming. 2019. Assessing the Effects of a Full-body Motion-based Exergame inVirtual Reality. In
Proceedings of the Seventh International Symposium of ChineseCHI on - Chinese CHI 2019 . ACM Press, Xiamen, China. https://doi.org/10.1145/3332169.3333574[84] Wenge Xu, Hai-Ning Liang, Zeying Zhang, and Nilufar Baghaei. 2020. Studyingthe Effect of Display Type and Viewing Perspective on User Experience in VirtualReality Exergames.
Games for Health Journal
9, 4 (Feb. 2020), 10. https://doi.org/10.1089/g4h.2019.0102 Publisher: Mary Ann Liebert, Inc., publishers.
A APPENDIXA.1 Data Results
We list all data in Table 5 and 6. VR_Cer represents VR certaingame condition (GC), VR_Unc represents VR uncertain GC, TV_Cerrepresents TV certain GC, TV_Unc represents TV uncertain GC.
HI ’21, May 8–13, 2021, Yokohama, Japan Xu et al.
Table 5: Means (SDs) of participants’ performance data regarding the completion time on each of the three lives of the monster,total number of gestures performed, and success rate of each move.
Young Adults Middle-aged AdultsType VR_Cer VR_Unc TV_Cer TV_Unc VR_Cer VR_Unc TV_Cer TV_UncCompletion Time on Each of The Three Lives of The MonsterLife1 126.83(34.74) 126.14(36.79) 133.69(54.42) 140.05(51.73) 149.38(41.71) 152.82(48.91) 148.16(60.72) 144.64(44.30)Life2 113.80(22.07) 114.60(33.05) 119.71(33.22) 121.71(37.82) 133.96(25.06) 138.09(39.38) 131.36(29.37) 142.75(47.57)Life3 105.45(19.23) 109.30(35.73) 112.21(29.10) 115.39(23.88) 134.27(36.88) 126.20(28.26) 121.08(19.63) 139.17(46.41)Total Number of Gestures PerformedKick 33.19 (7.88) 35.13 (7.44) 38.00 (8.33) 34.75 (10.08) 36.50 (10.41) 36.56 (8.60) 35.19 (9.09) 35.00 (9.35)Push 45.56 (13.77) 44.25 (8.31) 47.50 (9.64) 44.13 (14.06) 40.94 (10.97) 41.31 (8.54) 34.31 (13.80) 35.88 (11.91)Zoom+Kick 35.25 (3.62) 35.75 (5.01) 32.81 (3.29) 35.50 (3.12) 34.06 (4.55) 35.75 (5.42) 35.06 (5.20) 35.56 (5.67)Zoom+Squat 29.19 (7.88) 33.88 (13.50) 24.56 (9.24) 36.63 (13.44) 33.75 (4.93) 46.69 (7.43) 34.44 (5.68) 50.44 (12.93)Success Rate of Each MoveKick 82.19%(12.31%) 83.72%(10.28%) 77.28%(13.50%) 75.87%(21.27%) 80.03%(12.01%) 80.44%(13.10%) 75.96%(11.01%) 76.04%(7.98%)Push 52.04%(25.06%) 55.34%(24.32%) 61.05%(18.84%) 60.43%(20.57%) 54.76%(13.12%) 49.18%(18.52%) 56.79%(15.61%) 55.31%(15.83%)Zoom+Kick 98.32%(2.14%) 99.50%(1.08%) 96.16%(3.46%) 98.13%(2.17%) 97.61%(4.08%) 97.31%(4.03%) 99.14%(2.20%) 96.19%(3.81%)Zoom+Squat 74.07%(13.17%) 73.03%(13.36%) 61.88%(17.56%) 55.46%(18.25%) 73.51%(10.20%) 64.12%(9.21%) 70.13%(11.33%) 63.10%(6.78%)