Yearly more than a million succumb to ankle injuries in the United States alone, and it is not uncommon that individuals who have suffered such injuries lack the motivation necessary in order to successfully complete the rehabilitation process. In this article we describe three design and evaluation of three prototypes intended to provide individuals in need of ankle rehabilitation with the necessary motivation. The prototypes leverage video games potential as a source of intrinsic motivation by allowing individuals to control a game by means of a wobble board—an instrument used for ankle training–and thereby allow them to perform the needed exercises while playing. The design of the first prototype is informed by specific ankle exercises and theory pertaining to gameplay as emotional experience. An expert evaluation indicated that the prototype facilitated correct ankle training, and a user study suggested that participants generally found the act of playing intrinsically motivating. In a second study we compared a the wobble board interface with to commercially available input devices (the Wii balance board, and keyboard and mouse). The results provided insights about the relationship between perceived and actual performance, and suggests that novice players may underestimate their performance while balancing on the board—presumably because this task is difficult to novice users. Finally, we present a third prototype of the wobble board interface which is augmented with actuators and thereby provide vibrotactile feedback to the user while playing. The results of the performed evaluation indicated that the additional feedback need not improve performance in relation to the particular game being played. Moreover, the results suggest that one should be mindful when to add vibrotactile feedback during wobble board games since the feedback may distract the user, but also has the potential to make the experience more involving.
CCS Concepts: .Human-centered computing → Graphics input devices; Empirical studies in HCI;
Additional Key Words and Phrases: rehabilitation, exertainment, exergaming, intrinsic motivation
Ankle sprains are one of the most commonly occurring injuries amongst amateur and professional athletes alike [Ashton-Miller et al. 2001]. However, it is not just the physically active who succumb to this form of injury. Both the elderly and individuals who perform little or no regular physical activity are susceptible to ankle sprains [Asp et al. 2007]. Moreover, approximately 30% of the individuals who have sprained an ankle develop chronic ankle instability. The long term effects may also include a decrease in physical activity and early development of osteoarthritis [McKeon and Mattacola 2008]. The extent of the problem is reflected by the estimate that ankle sprains yearly amount to 1.6 million physician office visits and more than 8000 hospitalizations in the United States alone [McKeon and Mattacola 2008]. So, in addition to being a source of discomfort and an obstacle to the health of individuals, ankle injuries also constitute a considerable health care cost, estimated to amount to several millions of dollars of a year [McGuine and Keene 2006].
Fig. 1. Four common wobble board exercises [Asp et al. 2007]: (a) Balancing while keeping as steady as possible. (b) Moving the board back and forth. (c) Moving the board from side to side. (d) Clockwise and counterclockwise circular movement.
Luckily equipment does exist, that can be used to train ones ankles preemptively or as part of the rehabilitation process, once the damage has been done. The wobble board is one such piece of equipment. It consists of a circular disk, which is placed upon a half sphere. When training by means of the board the user stands on the disk while performing a series of specific movements (Figure 1). Even though this type of equipment has been proven to significantly reduce the risk of ankle sprains (see e.g. [McGuine and Keene 2006]), one problem seemingly remains in connection to some potential users. That is, individuals oftentimes stop performing the prescribed exercises after a relatively short period of time as they simply find the activity too boring or become discouraged by the absence of any palpable signs of improvement [Asp et al. 2007]. The problem is in other words the one of ensuring that individuals in need of ankle training remain motivated to perform the necessary exercises on a regular basis for the duration of the rehabilitation process.
Notably it is possible to distinguish between two distinct types of motivation, namely extrinsic and intrinsic motivation [Denis and Jouvelot 2005]. The former refers to the motivation that pushes us to perform a particular activity on account of factors that are external to the activity itself, such as punishment or reward. Intrinsic motivation does contrarily incite us perform the activity freely, for no other reason than our desire to do so. The fact that individuals in need of ankle training fail to successfully perform the necessary exercises can presumably be ascribed to insufficient or complete absence of extrinsic motivation. That is to say, the individual’s personal need to perform the exercises and the feeling of obligation towards the physiotherapist involved in the process do not necessarily suffice. Activities that by definition are intrinsically motivated, include the act of playing in general as well as instances of play revolving around a game [Nakamura and Csikszentmihalyi 2002]. This is a property which also extends to the act of playing digital games since people for the most part play such games for no reason other than the act of playing itself. Thus, it may be possible to incite otherwise unmotivated individuals to perform the necessary exercises by leveraging games’ seeming potential as a source of intrinsic motivation.
In this paper we describe three studies regarding the use of a traditional wobble board as a game controller. The paper is structured as follows: Section 2 outlines relevant related work. Section 3 describes our initial development and evaluation of a prototype enabling users to interact with a game by means of an augmented wobble board [Nilsson 2012a]. Section 4 details a study comparing three different ways of controlling a skiing game, namely, the augmented wobble board, the Wii Balance Board, and mouse and keyboard input. Section 5 presents a study exploring the role of audio-haptic feedback during use of the wobble board as a game controller [Nilsson 2012b]. Finally, Section 6 describes our conclusions on the presented work.
2. RELATED WORK
Systems combining physical movement with the act of playing games have been around for decades [Bogost 2005]. However, Nintendo’s Wii Fit is probably one of the games that have enjoyed the greatest commercial success, with millions of units sold worldwide [Nintendo 2008]. Interaction with the Wii Fit is first and foremost performed by means of the Wii balance board, which is a static board that detects the distribution of weight its surface, thus enabling the user to control virtual characters or objects by shifting the weight from one limb to another. The Wii Fit enables the user to train through two forms of activities, that is, explicit exercises and balance games. The explicit exercises include strength training, yoga and aerobics while the balance games comprise a selection of mini-games which encourage a variety of different forms of physical movement. Finally the Wii Fit enables users to keep track of their own progress and compare it with goals they have set for themselves.
Notably, it has been proposed that the Wii Balance Board may be used for training and assessing balance [Clark et al. 2010; Esculier et al. 2012; Gil-Go´ mez et al. 2011; Young et al. 2011], and the board has also been used as a virtual travel interface [Filho et al. 2012; Williams et al. 2011; Wang and Lindeman 2012].
A system, which is more explicitly aimed at ankle rehabilitation, is The “Rutgers Ankle” Rehabilitation interface [Girone et al. 1999]. The system does more specifically enable users to train in the comfort of their own homes while allowing a therapist to remotely monitor the exercise sessions. When interacting with the interface the user has one foot place on a movable circular platform that detects the pitch, roll and jaw of the foot as well as the forces applied to the platform. In addition to making it possible for users to interact with virtual worlds by means of his or her ankle movement the interface also provides resistive forces to the platform and thus the foot. One game-like application developed for The “Rutgers Ankle” Rehabilitation interface is a simple flight simulator where the user is tasked with piloting an airplane through a series of hoops. In this instance the therapist is able to alter the level of difficulty by changing the number and placement of the hoops, the speed of the airplane, the maximum force levels of the platform and the viewpoint of the simulation [Deutsch et al. 2001].
The prototypes described in the current paper are not the first to use a wobble board as a game controller. The first iteration in the design described in section 3 was dubbed the WobbleActive. When using this interface the user’s movements on the wobble board are registered by four bending sensors, attached within hinges and distributed evenly underneath the disk which the user is standing on. The data gathered from these sensors are then used to control the user’s avatar a flying saucer during play and used to navigate a menu system [Asp et al. 2007]. The prototype allows users to play two different games at varying difficulty levels. In one game the users are faced with the challenge of keeping the flying saucer within an area at the center of the screen, thus forcing them to balance on the board. In the second game the challenge is to navigate through a maze, as this ensures movement from side to side and back and forth. The two game designs were primarily informed by the need to ensure that the user performed correct exercises while playing [Asp et al. 2007]. These requirements were necessarily essential, as a prototype that did not ensure appropriate ankle training would defeat its own purpose. However, it would seem that one major question was left unattended during the design of the prototype, namely, how to create gameplay that incites players to continue playing and exercising? The prototype described in section 3 was designed explicitly to meet this aim.
Fitzgeral et al.  describe a balance training system that, similarly, allows the user to control a game using a wobble board. When playing the game the user has to steer a ball through an obstacle course by tilting the virtual ground plane using the movement of the wobble board. The results of the performed evaluation indicate a high level of usability.
More recently, Fitzgeral et al.  performed a study investigating how their therapeutic exergaming system compare to traditional wobble board training in terms of the improvements to dynamic balance and intrinsic motivation. The results of the study suggested that the improvements to dynamic postural stability were similar across the two conditions, but the exergaming system were superior in terms of the perceived level of interest and enjoyment.
3. GAMEPLAY AS A SOURCE OF INTRINSIC MOTIVATION
The aim of initial prototype was as suggested created with the intention of facilitating intrinsic motivation on behalf of the player while ensuring correct ankle training. In order to meet this aim we created a game informed by Jävinen’s  conceptualization of gameplay as emotional experience.
3.1. The Feeling of Playing Games
Jävinen  describes that the enjoyment experienced when playing games to a large extent originates from the goals imposed on the player by the game. Particularly, it is the player’s aspiration to achieve these goals that makes it possible for the game to elicit the emotional responses which color the player’s experience, and in turn determine if said experience is pleasurable. Notably, it is not the goals themselves that elicit these responses but rather the events encountered while striving to achieve these goals [Jävinen 2009]. Jävinen  introduces a conceptualization of the pleasures brought about by games and other forms of entertainment that provides a potential answer to this question. The conceptualization has been adopted from experimental psychologist Kubovy , who has proposed that it is possible to identify at least five pleasures of the mind. A pleasure of the mind may in general terms be described as a temporally distributed sequence of emotions. A feature of these pleasures, which makes their connection to the experience of intrinsic motivation more readily apparent, is that they are sought out voluntarily. In other words, players strive to experience these pleasures of the mind and use games as vehicles to do so. While Jävinen  does not explicitly describe whether players are mindful of their pursuit for pleasure, it would seem that this pursuit, and the pleasures it leads to, can help explain why the act of playing games may be intrinsically motivated. That is to say, the conscious or unconscious prospect of experiencing one or more of these pleasures of the mind serves as an intrinsic goal, while the actual experience of these constitutes an intrinsic reward. The five types of pleasures of the mind outlined by Jävinen  are curiosity, virtuosity, nurture, sociality, and suffering:
Curiosity: This category of pleasures is derived from the process of satisfying one’s epistemic hunger; i.e., the process of acquiring knowledge pertaining to something previously unknown. This does in turn imply that the associated emotions by and large are leveled at the unknown and related to the inferences made about the outcome of current and future events [Jävinen 2009].
Virtuosity: This category of pleasures accompanies the experience of being proficient, implying that the underlying emotions are leveled at the individual’s own actions and level of proficiency [Jävinen 2009].
Nurture: Nurture relates to the pleasures experienced when taking care of “living” things. Thus, the emotions forming the basis for these pleasures may be leveled at the objects of the nurture or the act of nurturing itself [Jävinen 2009].
Sociality: As the name implies, this category relates to the pleasure of being a member of a social group. Thus, the emotions forming the basis for these pleasures are leveled at the remaining members of a social group and sociality may be closely tied to the joy of cooperating and receiving praise from one’s peers [Jävinen 2009].
Suffering: This fifth and final category of pleasures has its roots in the experience of negatively valenced emotions. Particularly, Jävinen  notes that pleasure may be derived from the experience of “mundane” psychological pains (e.g., guilt and shame), or from “existential” pains (e.g., fear of death). Even though the experience of negatively valenced emotions may be pleasurable, it seems doubtful that this type of emotions always will yield a pleasurable experience. Quite contrarily, it would appear that negatively valenced emotions only help make the experience intrinsically motivating if they fall within a tolerable range.
3.2. Prototype Design and Implementation
The prototype was developed based on principles of iterative player centered design (see e.g., [Gulliksen et al. 2003; Salen and Zimmerman 2004]). Thus, in addition to being informed by the theory outlined in subsection 3.1, the design was modified based on user feedback collected during four qualitative tests. Since small scale tests are sufficient when identifying usability issues [Nielsen 2000], no more than five individuals participated in each of the tests. The first two tests evaluated the usability of the prototype while the final two primarily were intended to fine-tune the gameplay (see subsections 3.2.3 and 3.2.4 for details). All participants were undergraduate or graduate students at Aalborg University.
Seeing as individuals of virtually all ages and backgrounds may succumb to ankle injuries, the potential user group is very diverse. Consequently, it was decided to focus explicitly on first-time users who had prior experience with computer games. This necessarily had implications on how the theory could be applied. Balancing on the wobble board already poses a challenge to novice users. Therefore it was deemed relevant to expand upon this existing element of challenge in order to facilitate the pleasure of virtuosity. However, the challenge of balancing might render novice users incapable of simultaneously engaging with challenges requiring high levels of cognitive capacity and mental acuity. Thus, when striving to elicit a sense of virtuosity it was decided to primarily focus on sensorimotor challenges. The need for allocating attentional resources to the act of balancing also led to the decision of omitting a complex narrative. However, the pleasure of curiosity was not disregarded all together. Particularly, this pleasure is largely the product of emotions leveled at the unknown (e.g., suspense), and it may therefore be facilitated by introducing uncertainty about the outcome of ongoing events. Finally, it was decided to focus on solitary ankle training, and the facilitation of sociality and nurture was therefore not regarded as a priority.
3.2.1. Gameplay At a Glance. The prototype does, as previously suggested, allow users to interact with a game by means of a wobble board. The game was implemented using Unity 3D. The player controls a archetypical flying saucer on a reconnaissance mission to Earth which involves three objectives, defining the goals of each of the game’s three levels: (1) Reaching Earth by maneuvering the saucer through an asteroid field. (2) Locating and abducting as many Earth specimens (cows) as possible within a limited period of time. (3) Returning as many of the specimens as possible to the mothership by shooting them through a gate on its side. Screenshots of the three levels are presented in Figure 2. Each goal is revealed to the player upon completion of the preceding level. This theme was chosen for three primary reasons: The visual resemblance between the wobble board and the flying saucer was believed to make the interaction more intuitive; the simple, humorous narrative should provide the user with a series of meaningful, yet bizarre, goals; and the theme appeared to have a relatively wide appeal when used for the original WobbleActive prototype [Asp et al. 2007].
Fig. 2. Screenshots of the game’s three levels. (a) Level 1: maneuvering the flying saucer through an asteroid field and reach Earth. (b) Level 2: locating and abducting cows (c) Level 3: return the cows to the mothership by shooting them through an opening in its side.
3.2.2. Physical Interface Design. In order to translate the movement of the wobble board into their virtual correlates, a Phidgets accelerometer was mounted inside the hollow spherical base of the board [Phidgets 2012]. The tilting angle of the board is derived from the acceleration based on a simple trigonometric calculation using the unit circle as a representation of the gravitational acceleration of 1G. The angular tilt (θ) on each of the two axis x and y can then be calculated from the acceleration (o) using the gravitational acceleration (h) by means of the formula θ = arcsin(o/h). As the fixed axes on the accelerometer are mapped to specific movements in the game it is important that the user is facing in the right direction when standing wobble board. A set of footprints were added to the surface of the board in order to indicate to the user where to place the feet.
3.2.3. Mapping and Control. The mapping between real and virtual movement was informed by the ongoing evaluation of the prototypes usability and the prescribed exercises (Figure 1). In order to reduce monotony and ensure that the game facilitates the appropriate ankle exercises a set of unique challenges were designed for each of the game’s three levels.
When facing the challenge of navigating though the asteroid field in the first level, the player can control the movement of the flying saucer in three dimensions. Particularly, the player only controls the pitch and roll of the saucer while the speed is kept constant; i.e., the mapping resembles the one employed when controlling a flight simulator using a joystick. The informal usability tests suggested that the participants generally found this control scheme intuitive. Since the player has to dodge the asteroids by flying over, under and around them, the player both has to tilt the board from side to side and back and forth. Moreover, semi-circular board movements are oftentimes necessary in order to perform evasive maneuvers.
In the second level, where the player has to locate and collect cows, movement of the saucer is restricted to two dimensions; i.e., lateral, and forward and backward motion along the ground plane. In addition to resolving various minor usability issues, the usability tests showed that different control schemes were suitable across the performed tasks.
One test compared two different ways of controlling the lateral movement of the saucer. When the user tilted board to the left or right, the saucer either moved to the left and right, or else it rotated around its vertical axis in either direction. The feedback provided by the participants suggested that the former was suitable when a higher level of precision was necessary while the latter was preferable when traveling over large distances.
The second usability test evaluated two methods for controlling forward and backward movement. In one case the forwards and backwards tilting angle of the board was directly mapped to the virtual speed, and in the second case forward and backward tilting of the board resulted in acceleration and deceleration, respectively. The participant’s feedback indicated that control of acceleration and deceleration was preferred when traveling through the environment while the direct mapping between tilt and speed was preferred when attempting to keep the saucer steadily hoover over a specific point.
Based on these evaluations it was decided to use an adaptive control scheme for the second level; i.e., a different control scheme was used depending on whether the user was navigating around the virtual environment or abducting cows. Particularly, when navigating the environment, forward and backward tilting of the board controls the acceleration and deceleration of the saucer, and left and rightward tilting rotate the saucer around its vertical axis. With respect to the exercises this tasks forces the user to both tilt the board from side to side and back and forth.
When abducting a cow the control scheme changes, and the tilting of the board is now mapped to the lateral, forward, and backward movement of the space ship. Particularly, the control scheme changes when the saucer enters a circular area with a cow at its center. The scheme changes back again when cow has been successfully abducted or if the saucer leaves the circular area. In order to indicate that the control scheme has changed a tractor beam is emitted from the saucer and the virtual point of view is changed so as to make it easier to see if the saucer is directly above the cow. This tasks forces the player to balance steadily on the board while the cow slowly is being lifted from the ground.
It is detrimental to the exercises if the edge of the board comes into prolonged contact with the ground. In order to discourage such actions a simple penalty mechanic was implemented. Specifically, prolonged contact with the ground causes the saucer to swirl around, leaving the player unable to control the saucer and loosing valuable time. Morever, if a cow has been abducted prolonged contact will result in the cow being dropped.
In the third level the player no longer has to navigate around the environment, but instead return the cows to a mothership by shooting them through a gate on the mothership’s side. The tilting angle of the board controls to movement of a crosshair; i.e., the more the board is tilting to one side the faster the crosshair will move in that direction. If the crosshair is in the vicinity of the mothership’s gate a launch sequence is initiated and after a 3 second countdown a cow is fired. Thus, the user has to balance steadily on the board in order to maintain the aim of the crosshair.
3.2.4. Level Design. Since the prototype was intended for novice users, the first level is meant to serve as an introduction to the challenge of balancing on the wobble board while controlling a game. Thus, to allow for experiences of virtuosity, the difficulty of the first level was adjusted to make it suitable for individuals with little or no experience with balancing on the board. Particularly, the size and density of the asteroid field was adjusted based on feedback from the two evaluations pertaining to playability. While the difficulty was kept low the challenge was designed to appear more difficult than it really is; i.e., from the starting position the asteroid field appears far denser than it really is. This was done in an attempt to elicit a feelings of suspense which may be related to the pleasure of curiosity, and to increase the pleasure of virtuosity experienced upon successful making it through the asteroid field.
In the second level the player is repeatedly faced with a sequence of three challenges; i.e., locating a cow, abducting it, and returning it to a container at the center of the environment. The player is simply tasked with collecting as many cows as possible in four minutes. The ongoing evaluation suggested that four minutes was a sufficient time for novice players to fully experience the level without becoming fatigued. No minimum number of cows was specified in order to reduce the risk that novice players would feel a sense of failure at this stage in the game. The level was designed to give the players an experience of gradual improvement when repeatedly facing the three challenges as this was believed to contribute to the pleasure of virtuosity. We generally refrained from including intellectually demanding challenges. However, one game mechanic did allow the players to employ some degree of strategic thinking as this might contribute to the sense of virtuosity. While navigating the environment players are able to pick up two types of power-ups that may be beneficial at certain points in time. One power-up reveals the shortest distance to the nearest cow by displaying an arrow in the corner of the screen, and the other increases the virtual speed for a limited period of time. The size of the environment, the distribution of power-ups and cows, and the visual appearance of the icons representing these power-ups were based on the ongoing evaluation.
In the third level the player is challenged with returning the collected cows to the mothership by shooting them through a gate on its side. This challenge requires the user to adjust the aim of a crosshair while taking into account the current wind conditions. The aim is maintained by balancing steadily on the board until the cow is launched. Since the user has a limited amount of ammunition (the number of collected cows) the degree of success or failure is more readily apparent. This knowledge was believed to give rise to feelings of anticipation or suspense associated with the pleasure curiosity. As in the previous levels it was the hope that players might experience virtuosity upon successful completion of the level.
The objective of the performed evaluations was twofold: to determine if the game facilitated correct ankle training; and to assess if the participants experienced some of the pleasures of the mind outlined in section 3.1 as this might suggest that the participants found the act of playing intrinsically motivating. Consequently, an expert evaluation and a user study were performed.
3.3.1. Expert Evaluation of Training Efficacy. In order to determine whether the prototype did facilitate correct ankle training an expert on the topic was consulted, namely Anders Laun who at the time was finishing his degree as a physiotherapist at the University College Metropol in Copenhagen. Even though Laun had yet to become a fully certified physiotherapist, he was considered sufficiently knowledgeable as he was specialized within the field of proprioceptive ankle training. The expert consultation was divided into two sessions, an expert test and expert observation. The expert test involved Laun’s subjective and professional evaluation of the training, after and while he was playing the game. During the expert evaluation Laun was asked to observe the ankle movement of an individual playing the game, and subsequently evaluate the efficacy of the performed training. The data obtained during the expert test was gathered using a think-aloud approach and the data obtained from the expert observation was the result of an informal open-ended interview.
3.3.2. Results of the Expert Evaluation. The results obtained from the expert consultation with Anders Laun are summarized below. The summary of the interview data has been read and approved by Laun himself.
(1) The gameplay complies with all the necessary demands by affording controlled, proactive and reactive movements (Proactive balance is the ability employ sensory and motor skills related to expected postural demands, while reactive balance relates to the ability to regain balance after an unexpected disruptive action).
(2) Compared to traditional wobble board exercises the gameplay adds an extra reactive dimension, since the player has to compensate for the onscreen disruptive
actions such as avoiding suddenly appearing asteroids or controlling the movement of the flying saucer when accidently hitting a prop or a speed power-up.
(3) In comparison with ordinary wobble board training, the gameplay prompts a lot of static tension where the ankles are constantly working. This tension, resulting in physical fatigue, is more intense than the one caused by a regular wobble board where the user can rest a split second every time the board is tilted from one side to another.
(4) Since the gameplay prompts reactive balancing, it would be irresponsible to recommend the prototype to people who had just recently suffered from an ankle sprain. It would, however, be useful during the later stages of the rehabilitation process.
3.3.3. User Study. In order to assert whether the act of playing the game facilitated intrinsic motivation and lead to the experience of some of the pleasures of the mind outlined in section 3.1 a user study was performed. A total of 40 adult volunteers (mean age 28 years, 30 males and 10 females) participated in the evaluation. All participants were students at Aalborg University.
The participants played the game on an identical setup and were placed approximately two meters from the 5” plasma monitor (Samsung PPM50H3Q). Auditory feedback was displayed using a set of stereo speakers (Creative SBS 250). Figure 3 illustrates a scematic drawing of the setup used for the study. To ensure the safety and comfort of the participants a chair was placed in front of the board during play. Once done playing, the participants answered a questionnaire and were offered a beverage as compensation for participating. To reduce the risk of acquiescence bias the participants were lead to believe that they were evaluating one out of a series of possible prototype designs.
The presented questionnaire was designed to assess three dimensions of the participants’ experience of playing the game: (1) If they experienced a willing a continued commitment to the act of playing. (2) If they experienced one or more of the described pleasures of the mind (3) The participants perception of the prototype’s usability. The ad-hoc questionnaire was used instead of an existing motivation questionnaire (e.g., the Intrinsic Motivation Inventory [Ryan 1982]) since we wanted the questionnaire to be specific to the type of experience the game was designed to elicit. The final questionnaire comprised a total of 19 items, which were intended to answer the following questions:
Fig. 3. Schematic drawing of the setup used for the user study.
(1) How was the usability of the prototype? Items related to usability asked about the participants’ experience of the controls, the information presented on the heads up display, and how clear the goals of the game were.
(2) Did the participants experience the pleasure virtuosity? Items related to virtuosity primarily asked about how satisfied the participants were with their own performance and how difficult they found the three levels. Moreover, one item asked whether the participants had employed strategic thinking while playing the second level which allowed for some degree of strategic decision-making.
(3) Did the participants experience the pleasure curiosity? Since is closely tied to uncertainty about what is to come, the related items pertained to the sense of worry experienced during the games first and third levels which were designed to elicit such emotions. Moreover, an item pertaining to the experience of the audiovisual presentation was included as these stimuli might positively influence whether the participants wished to continue playing or not.
(4) Did the participants experience a willing and continued commitment to the act of playing? Finally, three items related to whether the participants had adopted the goals of the game, whether they wanted to continue playing once the game was over, and whether they experienced fatigue as a result of playing the game.
All of the questionnaire items were answered by means of Likert-type scale items. The items related to the perceived level of difficulty asked the participants to rate whether said level was ‘too low’, ‘low’, ‘moderate’, ‘high’, or ‘too high’. The remaining items asked the participants to rate their level of agreement with different statements on a six point scale (‘1’ signified strong disagreement and ‘6’ indicated strong agreement). Notes about the participants’ utterances and behaviors were made during and after exposure to the game.
3.3.4. Results of The User Study. The results of the user study are summarized in Tables I and II. The data obtained from the six point Likert-scales (Table I) was treated as interval data, and central tendencies are be presented as the mean rating of each item and variability is presented as standard deviations. The data related to the perceived difficulty of the three levels was treated as ordinal data and central tendency is summarized by the mode associated with each item (Table 2).
Based on the results of the expert evaluation it would seem that the prototype does afford both controlled proactive and reactive training. However, it is necessary to introduce a few caveats. Since the training facilitated by the prototype seemingly is more physically straining than traditional wobble board training, it seems likely that it, in its current form, will be less suitable during the earliest stages of a rehabilitation process. Thus, the game would need to be modified in order to reduce physical strain. Alternatively, the gameplay could also be modified to accommodate the abilities of more proficient users who are able to assign greater attention to challenges and events occurring in the game.
The results of the questionnaire items related to usability indicated that the participants generally found the prototype easy to use; i.e., the participants generally thought the movement of the board and spaceship corresponded well; the text and symbols were easy to understand and not regarded as an intrusion; and the participants were generally not in doubt about what to do next.
The means and standard deviations pertaining to the sense of achievement during the first level suggests that the participants may have experienced varying degrees of virtuosity upon completion of this level. The mean related to participants’ satisfaction with the number of collected cows was relatively low in comparison, and it is therefore unlikely that the experience of virtuosity was the norm during the second level. A possible explanation for the difference is that the only criterion for success in the first level was to make it to Earth while the two subsequent levels provided the participants with more explicit information about their performance, i.e. number of cows collected and delivered to mothership.
Table I. Results pertaining to the six-point Likert-type scale items.
Based on the item related to the experience of gradual improvement, it would appear that a number of the participants had a sense of gradually becoming better as the game progressed. Notably, a comparison of the data pertaining to the experience of improvement and the general willingness to continue playing indicated that the participants, who had experienced a sense of progressive improvement, in average also provided higher ratings when asked if they would have liked to continue playing (Pearson r =
0.47). Moreover, it is interesting to note that each level only received one rating reflecting a negative experience of the difficulty level (‘too low’ or ‘too high’). Assuming that the participants refrained from making these ratings because the challenges neither were too trivial nor too difficult, this may be viewed as a positive indication.
The results of the two items related to the experience of worry during the first and third level are too scattered to warrant any conclusions. That is, the means and standard deviations indicated little or no agreement amongst the participants. Thus, some participants may have experienced some level of suspense or worry, but this was by no means the norm. Therefore, it is unlikely that such negatively valenced, prospect-based emotions led to the pleasure of curiosity.
Judging from the mean pertaining to the question of whether the participants devised strategies while playing it would appear that they did so to some extent. However, this results does not reveal what the participants consider to be a strategy, if they managed to execute it, or if the successful execution led to an experience of virtuosity. Moreover, the results suggest that sound and visuals contributed positively to the participants experience.
Table II. Frequency of responses to the items pertaining to the perceived difficulty the levels.
Fig. 4. A screenshot of the skiing game used for the user study.
The results related to the question of whether it mattered to the participants how many cows they managed to collect and return to the mothership indicates that it in some capacity mattered to most of the participants how well they performed.
Notably, a number of participants were curious to know how many cows they had successfully collected, and several participants asked to try the game again so that they could beat their previous score. This may be viewed as an indication that at least some participants adopted the goals imposed by the game. On a similar note, the results pertaining to the question of whether the participants wanted to continue playing suggested that the participants did so to some extent. Finally, the results suggested that the degree of perceived fatigue varied greatly from participant to participant.
4. VIRTUAL SKIING USING THREE DIFFERENT INTERFACES
This section details a study comparing the augmented wobble board with two commercially available input devices: the Wii balance board, and keyboard and mouse input. The latter was included in order to provide information about how the alternative input devices would measure up against conventional devices. A game was created as a test vehicle for the comparison of the three interfaces.
4.1. Gameplay at a Glance
The designed game was in inspired by the classic skiing game Mogul Maniac—a two-dimensional slalom course skiing game—and the associated input device, the Joy board, which generally can be viewed as a technologically inferior predecessor to the Wii Balance Board [Bogost 2007]. This game was chosen because of the simplicity of the challenge the player is faced with; i.e., the player simply has to navigate down the slope. This navigational challenge may be very easy to perform but it is difficult to master. Particularly, the game required the player to ski down a slope along a fixed path as fast as possible while attempting to pass through a series of 31 gates along the way. While skiing the player is visually presented with information about the elapsed time and the current score. One point is awarded for every gate the player passes through. The game was implemented using Unity 3D. A screenshot of the game is presented in Figure 4.
Fig. 5. The three mappings used for the three input devices. (a) Wobble board: Tilting the board to the left or the right results in a turn in either of the two directions and tilting the board forward entails forwards movement. (b) Wii balance board: Leaning forwards or backwards results in acceleration and deceleration, respectively, and leaning to the left or right result in movement in the corresponding directions. (c) Keyboard and mouse: The keys w, a, s, and d controls the acceleration and deceleration and right and leftwards movement. The mouse is use to rotate to either the left or right.
4.2. Mapping and Control
Each of the three input devices involved very different movements on behalf of the player, and three different mappings between real and virtual movements were therefore create. The mappings used for each of the three input devices are illustrated in Figure 5. Despite the different mappings, the highest possible speed and acceleration were identical for all mappings.
4.3. User Study
The aim of the study was, as suggested, to explore how the wobble board compared to two commercially available input devices, namely, the Wii balance board, and mouse and keyboard.
4.3.1. Method and Materials. A total of 10 volunteers partook in the evaluation (8 males and 2 females) with a mean age of 23.9 ± 2.5). In order to prevent any distractions and to ensure unrestrained movement on behalf of the participants, the evaluation was performed in a closed and spacious room. An identical setup was used for all interfaces, namely, a 27 inch iMac with a 3.1 GHz Intel core i5 processor and 4GB of memory running Mac OS X version 10.6.8 and Unity 3D, displayed full screen at the resolution of 2560 x 1440. An iMac wireless keyboard and Mac USB mouse were used for the keyboard trails. Phidget21 plugin version 2.18 was used for the wobble board data acquisition and OSCulator version 220.127.116.11 for Wii Balance Board data.
The participants tried the skiing game with the three interfaces in randomized order, and the slope was identical across the three trials. Before each trial the participants were asked to familiarize themselves with each interface by going down the slope once.
Once the participants had completed the three trials, they were required to fill out a questionnaire asking them to compare the three interfaces. The questionnaire consisted of a series of items asking the participants to rate the degree to which they agreed with particular statements on a scale from ‘1’ to ‘7’ (‘1’ signified strong disagreement and ‘7’ signified strong agreement). Particularly, they were asked to rate each input device in terms of ease of use and the level of enjoyment. Moreover, the participants were asked to rate their overall experience of the game from poor and excellent, and rate their desire to play the exergame again given the chance. Finally the participants were asked to comment on their experience of the three interfaces. Two performance measures were employed, namely, completion time and the number of cleared gates.
4.3.2. Results. The results pertaining to the self-reports and performance measures are summarized in Table III. Repeated-measures analyses of variance (ANOVAs) were used for statistical comparison of the data obtained from all measures (α = .05). Significant measures were subjected to post-hoc analyses using paired-sample, one tailed t-tests with Bonferroni corrected alpha values (α = 0.017). The ANOVAs revealed significant differences in relation to completion time (F (2, 29) = 24.6, p = .000007), game score (F (2, 29) = 19.6, p = .00003), and perceived ease of use (F (2, 29) = 4.4, p = .03). No significant difference was found between the means obtained from the self-reported measure of enjoyment.
Table III. Means ± one standard deviation pertaining to the self-reports and performance measures used in the study.
In relation to completion time, the post-hoc analysis revealed signifcant differences between all conditions: wobble board and Wii balance board (p = .0031), wobblar board and keyboard and mouse (p = .0013), and Wii balance board and keyboard and mouse (p = .0001).
In relation to game score, significant differences were found between the wobble board and keyboard and mouse (p = .0002), and between the Wii balance board and keyboard and mouse (p = .0003).
In relation to ease of use, a significant difference was found between the wobble board and mouse and keyboard (p = .00112).
Moreover, the results were tested for correlations using Pearson’s correlation. These tests revealed a significant correlation between completion time and game score (r(30) = −.63, p = .0001) indicating a strong negative relationship between the two variables. Positive relationships were found between perceived ease of use and game score (r(30) = .55, p = .0008) and between perceived ease of use and enjoyment (r(30) = .51, p = .002).
Finally, the questionnaire item asking the participants to rate their overall experience on a scale from poor to excellent, yielded a mean of 5.3 ± 0.7, and the question related to whether they would like to play the game again produced a mean score of 4.8 ± 1.5.
The user study suggested that the participants generally performed better when using the keyboard and mouse to control the game; i.e.,they completed the game significantly faster and managed to pass through significantly more gates. Interestingly, the participants were significantly faster when controlling the game by means of the wobble board compared to the Wii Balance Board. Moreover, the increased speed did not appear to negatively influence precision because no significant difference was found between the mean game score pertaining to the two interfaces.
The superiority of the mouse and keyboard controls was also apparent from the means pertaining to the perceived ease of use of the three interfaces. However, a significant difference was only found between the means related to the mouse and keyboard, and the wobble board, which in average scored the highest and lowest respectively. However, it hardly comes as a surprise that the wobble board was perceived as the most difficult interface since the act of balancing on the board may pose a challenge to novices even when they are not using the board as a controller.
While no significant difference was found between the three interfaces in relation to the participants experience of enjoyment, it is worth noting that the perceived difficulty may have negatively influenced the experience. Indeed a positive correlation was found between the perceived ease of use and enjoyment across the three interface types.
Moreover, it is interesting to note that there is a discrepancy between the objective measures of the participants’ performance and their subjective experience of their own proficiency. That is, the participants found the wobble board harder than the Wii Balance Board even though they in average achieved a better completion time with the wobble board and got a similar number of points with the two interfaces. Notably, no correlation between perceived ease of use and completion time was found, but a positive relationship was found between perceived ease of use and game score. A plausible explanation is that the participants did not have any success criteria to compare their own performance against in regards to the completion time. Conversely, the participants did know that there were a total of 31 gates along the ski slope, and they therefore had an ideal results with which they could compare their game score.
Finally it is worth noting that, while the analysis of the data revealed a positive correlation between perceived ease of use and enjoyment, no significant correlations were found between the two performance measures and the experience of enjoyment.
5. CONTINUOUS AUDIO-HAPTIC FEEDBACK FOR BALANCE CONTROL
The third study was performed with the intention of exploring the use of audio-haptic feedback for balance control.
5.1. The audio-haptic wobble board
A modified version of the traditional wobble board was augmented with sensors and actuators. Particularly, two vibrotactile actuators [TactileLabs 2052] were mounted between the two circular, wooden plates of the board that the user stands on (Figure 6). The user’s movement on the board was registered using the same type of accelerometer used in the previous studies. The accelerometer was placed between the two wooden plates.
5.2. Gameplay and Mapping
In order to test the role of auditory and haptic feedback in facilitating balance control, we designed a simple computer game. The goal of the game is to control the movement of an ice floe by means of the wobble board. The tilting angle of the board is directly mapped to the tilt of the ice floe. A penguin is located on the surface of the ice floe and the objective of the game is simply to prevent the penguin from falling off the ice floe and into the water for as long as possible. This challenge amounts to more than balancing, since the penguin slowly walks in circles on icy surface of ice flow which entails that the player at times will have to tilt the board in order to avoid that the penguin walks of the edge. Moreover, the difficulty of the game increases over time since a piece of the ice floe falls off every 15th second. This game is first and foremost designed with the intention of eliciting a pleasure of virtuosity, and since the player does not have to respond to unexpected obstacles it is the belief that there is little need for reactive balancing on behalf of the player.
The movement of the penguin in response to the tilt of the ice floe is governed by the NVIDIA PhysX physics engine built in to Unity 3D which was used to create the game. A screenshot of the game and a picture of a user balancing on the board while playing the game can be seen in Figure 6.
5.3. Audio-Haptic Feedback
When sliding on the ice floe, the penguin creates friction which can be both heard and felt at the feet. Particularly, the velocity of penguin’s movement is sent from Unity 3D to the real-time synthesis platform Max/MSP, that is used to simulate the auditory and haptic feedback. The audio-haptic synthesis engine is based on a physically-based simulation of sliding on a solid surface, initially developed as part of the Sounding Objects EU project [Avanzini et al. 2005]. In our simulations, designers have access to a sonic palette making it possible to manipulate different parameters, including material properties. However, for this particular prototype several parameters related to the material properties of the objects in contact were fixed, which allowed the simulation to be controlled only by the velocity of the penguin sliding on the ice. In order to increase the realism of the simulation, the remaining parameters were defined based informal evaluations prior to the study current study. The same simulation was used to drive both the the auditory and haptic synthesis engine. Thus, in the case of the auditory feedback, subjects hear the sound of the penguin sliding on ice, while on the case of haptic feedback they feel the vibrations representing the penguin sliding. Two laptops were used for the prototype: one was used to display the visual feedback and another generated the auditory and haptic feedback. A sound card (Fireface 800) was used to provide feedback both to the auditory channel and to the haptic channel. The auditory feedback was delivered through a pair of Beyerdynamic DT 770 closed headphones.
5.4. User Study
The purpose of the performed user study was threefold: (1) to determine whether the added feedback would positively influence the participants proficiency at playing the game; (2) to investigate whether it would affect the perceived difficulty of the game; and (3) to explore whether the feedback positively contributed to the participants experience of the game. Finally it was also regarded relevant to determine whether the gameplay afforded by the game could serve as a source of intrinsic motivation on behalf of the participants.
5.4.1. Methods and Materials. A total of 20 volunteers participated in the study (15 males and 5 females), with a mean age of 23.3 ± 2.5. The study was based on a within-subjects design, and the participants were exposed to the following types of feedback: visual feedback only, visual and auditory feedback, and visual and audio-haptic feedback. Henceforth, the three conditions will be referred to as the visual, auditory, and audio-haptic condition. This design was chosen for two reasons. First, it allowed the participants to compare the three combinations of feedback, and secondly, it should minimize the influence of individual differences in terms of how proficient the participants were at using the board as an input device. No condition involving only visual and haptic feedback was included because the actuators also emit sounds. This undesired auditory feedback could have been masked with another auditory stimulus (e.g., pink noise) delivered through the headphones. However, prior experiences with the use of such stimuli during balance tasks have led to the belief that it causes great discomfort on behalf of the participants. The participants were exposed to each of the three conditions three times. Initially they tried each condition once as part of a training session and subsequently they tried each condition twice in randomized order.
Fig. 6. The augmented wobble board (left), screenshot of the penguin game (middle), and a user balancing on the wobble board while playing the game (right).
Table IV. Means ± one standard deviation pertaining to the time spent playing.
The influence of the added feedback was assessed by means of two metrics: a performance metric and self-reports. Since the objective of the game is to keep the penguin on the ice floe for as long as possible, the participants proficiency at playing the game was determined based on how long time they were able to do so. These times were recorded for all nine trials, but the data obtained from the training sessions was not analyzed.
After the final trail a questionnaire was administered. This questionnaire included items pertaining to the perceived difficulty of the game, their preference in regards to the provided feedback and the gameplay’s ability to serve as a source of intrinsic motivation. The items pertaining to perceived difficulty and preference required the participants to rank the three conditions based on which one they found the hardest and which one they preferred. The participants were encouraged to elaborate on their ranking. Finally the gameplays general potential as a source of intrinsic motivation was assessed by means of a subset of the Intrinsic Motivation Inventory (IMI) [Ryan 1982]. The inventory assesses the individuals experience of a particular activity and considers six dimensions of this experience: interest/enjoyment, perceived competence, effort, value/usefulness, felt pressure and tension, and perceived choice. Since the subscale pertaining to interest/enjoyment is considered to be the measure of intrinsic motivation, the questionnaire used for the current study only comprised items belonging to this subscale. The questionnaire included five items which were adapted from a version of the IMI used by McAuley, Duncan, and Tammen . Three of items required the participants to rate the degree to which they had found the act of playing the game enjoyable, fun and interesting. The final two items related to whether the activity had been able to hold their attention and whether they had been conscious of the enjoyable nature of the activity while performing it. All five items required the participants to rate their level of agreement on a scale from ‘1’ to ‘7’ where ‘1’ signified strong disagreement and ‘7’ signified strong agreement.
5.4.2. Results. The results pertaining to the participant’s performance are summarized in Table IV, and the results related to the participants’ rankings in terms of preference and perceived difficult are summarized in Table V and VI, respectively. One participant failed to perform the ranking of preference correctly and this data is therefore omitted. Finally, Table VII details the means and standard deviations corresponding to each of the five questions adopted from the IMI as well as the aggregate subscale score (the grand mean).
Table V. Frequency of rankings pertaining to preference.
Table VI. Frequency of rankings pertaining to difficulty.
The data obtained from the performance measure suggested that the there was a difference in how well the participants performed across the three condition, albeit a statistically insignificant one. On average the performance was best when the visuals were accompanied by auditory feedback, while it was worst during exposure to the audio-haptic feedback. Four participants remarked that they found the added feedback to be an unnecessary distraction from the task at hand. Considering that the task of controlling the game by means of the board has yet to become automated in case of novice users this seems like a plausible explanation. Here, automatization refers to how the repeated execution of a task may result in increased speed and efficiency. Once a task is fully automated its demands on cognition diminishes and performance becomes effortless and possibly even unconscious, thus enabling it to be performed alongside a more controlled main task [Saariluoma 2005]. Notably, three participants made opposing comments. They described that they believed the audio-haptic feedback to provide additional information which made the task easier. An additional two made similar remarks in regard to the auditory feedback. It does, however, seem probable that the added feedback did not positively influence the participants’ performance. That does not necessarily mean that the addition of auditory or audio-haptic feedback cannot influence performance during balance tasks. Instead it seems plausible that the feedback was of little or no use in relation to the particular game used for the study. To be more exact, the game frequently requires the player to make rapid adjustments to the tilt of the board as the penguin is sliding across the surface of the ice floe. While such actions may save the penguin from falling off the edge, they oftentimes result in the penguin moving at high velocity towards the opposing edge, which in turn forces the player to move rapidly. During such situations the player seemingly relies on the visual feedback since this stimulus unlike to the auditory and audio-haptic feedback provides information about the penguin’s position and direction of movement. Since the auditory and audio-haptic feedback is produced from the velocity of the penguin as it slides across the ice floe, this feedback is inconsequential when the board is level and the penguin is not sliding.
The differences in opinion about how helpful the added feedback had been also seems to be reflected by the measure perceived difficulty. While the visual condition was ranked as the easiest by nine participants an equal number found this condition to be the hardest. Five participants rated the auditory and audio-haptic conditions as the easiest. The highest level of agreement was reached in relation to the auditory feedback, which was rated as the second easiest condition by 11 participants. The rankings pertaining to the audio-haptic condition are almost equally distributed amongst the three levels of difficulty. The opinions of the participants also seemed to diverge somewhat in regards to what feedback combinations they preferred. The audio-haptic condition was preferred by a total of nine participants, while eight regarded it as the least preferable. n equal number of participants regarded the visual condition as the least preferable. The highest level of agreement was again reached in relation to the auditory feedback, which was rated as the second most preferable by 12 participants. The participants who expressed a preference towards the audio-haptic condition qualified their answers by stating that it made the experience more immersive, involving, and enjoyable or believed it to make the task easier. The participants who preferred the visual feedback generally elaborate on their rating by stating that it was the least distracting condition, or described that they had found the other feedback combinations ambiguous. Finally the data pertaining to the IMI items yielded relatively positive results across the board and the subscale suggests that the participants generally had experienced the activity as intrinsically motivated, albeit to varying extents.
Table VII. Means ± one standard deviation pertaining to the IMI items and the subscale score.
In this paper we have described the design, implementation and evaluation of three prototypes allowing users to control a video game by means of a wobble board.
The first prototype was designed to meet two aims, namely, to facilitate correct ankle training by means of a wobble board while leveraging games’ potential as a source of intrinsic motivation. The design of the gameplay and mapping was informed by the prescribed ankle exercises and theory related to the emotions elicited by games. In order to assess the prototype an expert evaluation and a user study were performed. The expert evaluation confirmed that the prototype did ensure correct ankle training even though it was not ideal during the early stages of a rehabilitation process. Al though the user study did not unequivocally prove that all of the participants found the act of playing the game intrinsically motivating, it did indicate that this was the case for a number of said participants. However, further assessment by means of well-established measures of intrinsic motivation is necessary in order to further substantiate this conclusion. Moreover, additional evaluations of the efficacy of the training facilitated by the prototype are needed in order to determine how the training compares to the one achieved with conventional wobble boards.
The second prototype was based on the same wobble board as the first one, but involved a different game. Particularly, the prototype was used as a test vehicle for comparing the performance and experience of playing a game using a wobble board, the Wii balance board, and keyboard and mouse. The results of the study suggested the interface that was most enjoyed, the Wii Balance Board, also was the slowest, and the number of cleared gates was largely identical to the one related to the wobble board (both performed worse than mouse and keyboard). However reported ease of use pertaining to the Wii balance board was relatively high which may seem contradictory. Csikszentmihalyi’s  notion of flow may provide one possible explanation. An individual may enter a state of flow when performing an activity of interest where the equilibrium between perceived skills and perceived difficulty is maintained. So, while the participants may actually have performed worse when using the Wii balance board compared to the wobble board, it is entirely possible that they found the activity more enjoyable, since they perceived the match between their skills and the difficulty level to be suitable. In regards to the wobble board the opposite may be true. The participants may have performed better, but since they did not feel proficient they erroneously found the interface to be harder to use than it really was. With that being said, it should be stressed that novice users indeed may find balancing on the board difficult. While flow theory may help explain the inconsistent results, it is remains uncertain whether the participants indeed did experience flow while playing. In continuation hereof, it is worth mentioning that, the qualitative feedback obtained during the evaluation showed that participants were vary of correcting their foot placement when using the wobble board as they would lose control.
The final study was performed in order to explore the use of audio-haptic feedback for balance control while playing a simple game. The study involved a third prototype, augmented with sensors and actuators, and compared three conditions (visual feedback only, visual and auditory feedback, and visual and audio-haptic feedback). No significant difference was found in terms of performance, and different participants expressed a preference for and against the added feedback. Common reasons for these answers were that the additional feedback made the experience more involving and enjoyable, or that the the participants found the feedback disturbing. A possible explanation for the mixed results is that, in relation the specific game, the feedback might not have provided any information that could not already be gleaned from the visuals. The results pertaining to whether the participants found the game intrinsically motivating, suggested that this generally was the case, albeit to varying extents. Based on the three studies, we can with cautious optimism conclude that the combination of wobble boards and digital games may may serve as a source of intrinsic motivation while enabling users to perform correct ankle training. Moreover, we found that novice generally find it difficult to balance on the board, which suggests that different game-mechanics, and possibly entirely different games, needs to used at the different stages of the training process. Finally, we found no evidence suggesting that audio-haptic feedback improves performance while balancing on the board. However, the particular game used for the final study may not have been an ideal candidate for additional feedback, and it therefore seems possible that the feedback could have a positive influence in relation to other types of games.
James A Ashton-Miller, Edward M Wojtys, Laura J Huston, and Donna Fry-Welch. 2001. Can proprioception really be improved by exercises? Knee surgery, sports traumatology, arthroscopy 9, 3 (2001), 128–136.
Stefan E. Asp, Kolbrún Ó. Halldórsdóttir, Camilla Hägg, Micael L. Møller, Bodil Mikkelson, Lars B. Pedersen, and David Skaarup. 2007. Wobbleactive. In In Proceedings of the 1st international symposium on Ludic Engagement Design for All (LEDA) 2007. Aalborg University, Esbjerg.
Federico Avanzini, Stefania Serafin, and Davide Rocchesso. 2005. Interactive simulation of rigid body interaction with friction-induced sound generation. Speech and Audio Processing, IEEE Transactions on 13, 5 (2005), 1073–1081.
Ian Bogost. 2005. The rhetoric of exergaming. Proceedings of the Digital Arts and Cultures (DAC) (2005). Ian Bogost. 2007. Persuasive games: The expressive power of videogames. Mit Press.
Ross A Clark, Adam L Bryant, Yonghao Pua, Paul McCrory, Kim Bennell, and Michael Hunt. 2010. Validity and reliability of the Nintendo Wii Balance Board for assessment of standing balance. Gait & posture 31, 3 (2010), 307–310.
Mihaly Csikszentmihalyi and Mihaly Csikzentmihaly. 1991. Flow: The psychology of optimal experience. Vol. 41. HarperPerennial New York.
Guillaume Denis and Pierre Jouvelot. 2005. Motivation-driven educational game design: applying best practices to music education. In Proceedings of the 2005 ACM SIGCHI International Conference on Advances in computer entertainment technology. ACM, 462–465.
Judith E Deutsch, Jason Latonio, Grigore C Burdea, and Rares Boian. 2001. Post-stroke rehabilitation with the Rutgers Ankle System: a case study. Presence: Teleoperators and Virtual Environments 10, 4 (2001), 416–430.
Jean-Francois Esculier, Joanie Vaudrin, Patrick Beriault, Karine Gagnon, and Louis E Tremblay. 2012. Home-based balance training programme using Wii Fit with balance board for Parkinson’s disease: a pilot study. Journal of rehabilitation medicine 44, 2 (2012), 144–150.
H. K. Filho, W. Sarmiento, V. Jorge, C. Collazos, and N Luciana. 2012. Walk in Place Using a Balance Board Matrix. In Proceedings of Workshop on Works In Progress, SIBGRAPI 2012. 1–2.
Diarmaid Fitzgerald, Nanthana Trakarnratanakul, Lucy Dunne, Barry Smyth, and Brian Caulfield. 2008. Development and user evaluation of a virtual rehabilitation system for wobble board balance training. In Engineering in Medicine and Biology Society, 2008. EMBS 2008. 30th Annual International Conference of the IEEE. IEEE, 4194–4198.
Diarmaid Fitzgerald, Nanthana Trakarnratanakul, Barry Smyth, and Brian Caulfield. 2010. Effects of a wobble board-based therapeutic exergaming system for balance training on dynamic postural stability and intrinsic motivation levels. J. orthopaedic & sports physical therapy 40, 1 (2010), 11–19.
José-Antonio Gil-Gómez, Roberto Lloréns, Mariano Alcañiz, and Carolina Colomer. 2011. Effectiveness of a Wii balance board-based system (eBaViR) for balance rehabilitation: a pilot randomized clinical trial in patients with acquired brain injury. J. neuroengineering and rehabilitation 8, 1 (2011), 30.
M Girone, G Burdea, and Mourad Bouzit. 1999. The Rutgers Ankle orthopedic rehabilitation interface. Proc. ASME Dyn. Syst. Control Div 67 (1999), 305–312.
Jan Gulliksen, Bengt Göransson, Inger Boivie, Stefan Blomkvist, Jenny Persson, and Åsa Cajander. 2003. Key principles for user-centred systems design. Behaviour and Information Technology 22, 6 (2003), 397–409.
Aki Järvinen. 2009. Understanding video games as emotional experiences. The video game theory reader 2 (2009), 85–108.
Michael Kubovy. 1999. On the pleasures of the mind. Well-being: The foundations of hedonic psychology (1999), 134–154.
Edward McAuley, Terry Duncan, and Vance V Tammen. 1989. Psychometric properties of the Intrinsic Motivation Inventory in a competitive sport setting: A confirmatory factor analysis. Research quarterly for exercise and sport 60, 1 (1989), 48–58.
Timothy A McGuine and James S Keene. 2006. The effect of a balance training program on the risk of ankle sprains in high school athletes. The American journal of sports medicine 34, 7 (2006), 1103–1111.
Patrick O McKeon and Carl G Mattacola. (2008) Interventions for the prevention of first time and recurrent ankle sprains. Clinics in sports medicine 27, 3 (2008), 371–382.
Jeanne Nakamura and Mihaly Csikszentmihalyi. 2002 The concept of flow. Handbook of positive psychology (2002), 89–105.
Jakob Nielsen. 2000. Why you only need to test with 5 users. (2000). http://www.nngroup.com/articles/why-you-only-need-to-test-with-5-users/
Nilsson, N.C., Serafin, S. and Nordahl, R., 2012a. Gameplay as a Source of Intrinsic Motivation for Individuals in Need of Ankle Training or Rehabilitation. Presence: Teleoperators and Virtual Environments, 21(1), pp.69-84; http://www.mitpressjournals.org/doi/abs/10.1162/PRES_a_00090#.WOKqu1WGOM8/
Nilsson, N.C., Serafin, S. and Nordahl, R., 2012b. The Fwobble: Continuous audio-haptic feedback for balance control. In 3D User Interfaces (3DUI), 2012 IEEE Symposium, pp. 153-154; http://ieeexplore.ieee.org/document/6184207/
Nintendo. 2008. Annual Report 2008. http://www.nintendo.co.jp/ir/pdf/2008/annual0803e.pdf
Phidgets. 2012. PhidgetAccelerometer 3-Axis. (2012). http://www.phidgets.com/products.php?product id=1059
Richard M Ryan. 1982. Control and information in the intrapersonal sphere: An extension of cognitive evaluation theory. J. personality and social psychology 43, 3 (1982), 450.
Pertti Saariluoma. 2005. Explanatory frameworks for interaction design. In Future interaction design. Springer, 67–83.
Katie Salen and Eric Zimmerman. 2004. Rules of play: Game design fundamentals. MIT press. TactileLabs. 2052. Tactile Labs Products. (2052). http://tactilelabs.com/
Jia Wang and Robert W Lindeman. 2012. Comparing isometric and elastic surfboard interfaces for leaning-based travel in 3D virtual environments. In Proceedings of the 2012 IEEE Symposium on 3D User Interfaces. IEEE, 31–38.
B. Williams, S. Bailey, G. Narasimham, M. Li, and B. Bodenheimer. 2011. Evaluation of walking in place on a wii balance board to explore a virtual environment. Proceedings of the ACM Transactions on Applied Perception 8, 3 (2011), 19.
William Young, Stuart Ferguson, Sebastien Brault, and Cathy Craig. 2011. Assessing and training standing balance in older adults: a novel approach using the Nintendo WiiBalance Board. Gait & posture 33, 2 (2011), 303–305.