Tei2014_brush And Learn

Transcript

Brush and learn: transforming tooth brushing behavior through interactive materiality, a design exploration Miguel Bruns Alonso Jelle Stienstra Rob Dijkstra Department of Industrial Design, Eindhoven University of Technology Den Dolech 2, 5612 AZ Eindhoven, The Netherlands [email protected] [email protected] [email protected] ABSTRACT To counteract the increased tendency in skill learning addressing our cognitive abilities we discuss an opportunity on how performance skills can be trained by means of inherent feed forward through interactive materiality. We address this approach in the context of designing an interactive toothbrush that supports users in learning a complex brushing technique by relying solely on their perceptual motor skills. We discuss how we designed a natural coupling according to the Frogger framework in the action-perception loops with the interactive toothbrush. We evaluated the toothbrush in context. The experimental results indicate that complex movements can be learned by providing inherent feed forward on the actions of users in skill training. This supports our argument and vision that the design-inspired approach or interactive materiality may offer new opportunities for behavioral transformation. Author Keywords Behavioral transformation; inherent feed forward; tangible interaction; perceptual motor skill; tooth brushing ACM Classification Keywords H.5.2 User Interfaces: Haptic I/O General Terms Design INTRODUCTION There is a growing movement in the field of human computer interaction that focuses on reducing the cognitive load in complex interactions to achieve a balanced way in addressing all human skills [6, 14, 19]. The main motivations of this movement are (1) reducing the cognitive load of complex computer applications or embedded systems by offering interactions that are closer to our natural way of interacting with the world and (2) to increase the pleasure and the inherent reward in interaction which we are loosing given the non-physical properties of today’s computerized systems. This movement departs from the traditionally physically oriented School of Industrial Design Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. Copyrights for components of this work owned by others than ACM must 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 a fee. Request permissions from [email protected]. TEI'14, February 16 - 19 2014, Munich, Germany Copyright 2014 ACM 978-1-4503-2635-3/14/02…$15.00. and finds inspiration in the theory of ecological psychology [8] and phenomenology of perception [11]. Gibson’s notion of affordance concerns a meaningful relationship between user’s bodily capabilities and the inherent properties of the object; and from a phenomenological perspective, meaning emerges in interaction. This is highly opposed to current developments in HCI, which often decouple action and reaction in the development of new interfaces. In this renewed line of thought, perception and action cannot be detached and the design of affordances requires a tight coupling between action and reaction. Frens et al. [6] argue for integration between form, function and interaction. The form informs the function and interaction upfront, to ensure the user’s understanding of the product behavior. Moreover, in products with embedded computing these relations are necessities for the user to understand the product behavior. According to Wensveen et al. [21] tangible interaction needs a link between our sensory richness, the possibility to perform an action with physical objects and our perceptual motor skills. With their alarm clock they demonstrate possible product interpretations of sensory richness in order to create new interaction styles. They propose the Frogger framework, and suggest feedback to be coupled naturally, i.e. inherently, when user action and product reaction coincide in time, location, direction, modality, dynamics and expression. As extension, when the action stops, feedback can become feed forward to guide a user towards the intended functionality. The concept of inherent feed forward thus is a natural reaction of a product that can guide the further actions of the user. Furthermore, closed loop interactions can be designed into intelligent systems to manage the relations between user’s action and product’s reaction. Elaborating on this concept Bruns Alonso et al. [1] describe a framework on how emotional aspects can be influenced by means of an affective feedback loop [18]. They observed that the stressful behaviors of a user could be affected through haptic feedback provided in a pen. By manipulating the stressful behaviors the pen could influence the visceral response of the user, resulting in a reduced heart rate. As a consequence they suggest that product feed forward can be experienced only viscerally, i.e. without the user being cognitively aware of it, when it is inherent. In human factors there is a large interest in facilitating the learning of complex movements by studying the human performance [10, 22]. In general, these complex movements are highly demanding on user’s cognitive skills [9]. We believe that inherent feed forward may offer an opportunity for learning complex movements through perceptual motor skills rather than cognitive skills. In the current paper we explore whether the inherent feed forward can also be applied to transform behavior on a skill level. Therefore, we elaborate on the described framework and apply it to a motor action requiring a very specific and complex skill, tooth brushing. techniques measured from the way in which users employ the brush in their mouth) and perception (learning the right brushing techniques from a screen or projection). Thereby, the given examples focus strongly on the cognitive abilities of the user, as they need to translate what they have learned on-screen into a new way of tooth brushing. We believe that given the strong tactile properties of tooth brushing, a perceptual motor approach may be more effective to learn this complex motor skill. Therefore, in this paper, we present a method to change tooth-brushing behavior through inherent feed forward. INTERACTIVE MATERIALITY RELATED WORK Tooth brushing is an important element in personal hygiene. However, various researchers indicate a need for more dental education [7, 15, 16] as most adults still lack a good brushing technique [7]. Furthermore, although children are supported to learn the right tooth brushing techniques, after a while the low effort linear movements easily replace the complex, but proper, movements of tooth brushing [5]. As a consequence, persuasive games that support children in learning about oral hygiene and about proper tooth brushing behavior have been explored frequently in human-computer interaction. Molarcropolis for example is a mobile game that stimulates adolescents by means of information on oral illnesses and their causes, habits and activities to improve their oral health [17]. Although no longitudinal test was conducted, children appeared to learn correct brushing behaviors by means of the game. Furthermore, a study by Sylla et al [20] compared a tangible user interface to a graphical user interface. The tangible interface, in which children could physically brush an oversized physical tooth – with projected germs – until it was clean, was found to be more effective in learning children about oral hygiene than a graphical interface in which the graphical representation of the brush was controlled by a mouse. Two examples that address the brushing techniques are the Playful Toothbrush and the Virtual Aquarium. The Playful Toothbrush connects measured activity from a physical toothbrush to a computer game that guides children in learning proper brushing techniques and reinforces active participation of children in brushing activity. This example appears to improve the tooth brushing skills of kindergarten children within a relatively short training period [2]. Also the Virtual Aquarium aims at improving users’ dental hygiene by promoting correct tooth brushing practices. The system turns the bathroom mirror into a simulated aquarium in which the fish are affected by the users’ tooth brushing activity. When users brush their teeth properly, the fish prosper and procreate but weaken and may even perish when brushing is done improperly [13]. All the abovementioned examples take a classic HCI perspective and decouple action (current brushing To explore the possibilities for training tooth brushing skills by means of inherent feed forward, we have followed a research-through-design approach [23]. In this approach, we employed the three-step design process proposed by Stienstra, Bruns Alonso, Wensveen and Kuenen [19] for designing interactive systems that transform behavior through interactive materiality. Stienstra et al. [19] combine the theoretical concept of the Frogger Framework [21] with continuous action-perception loops and suggest to design according to the following the three steps: 1. Analyzing: affirming and appreciating the current behavior 2. Synthesizing: designing the transformation of behavior 3. Detailing: fine-tuning the sensitivities in the interactive materiality mapping for the Analyzing The desirable brushing technique According to Schlueter et al. [16] and Colgate [3] the Modified Bass technique (MBT) is a tooth brushing technique that proposes an effective way to clean teeth without destroying the mouths’ gums or enamel. The MBT (see Figure 1) requires complex hand-movements consisting of: • Making small strokes of a maximum of two teeth at the time • Holding the bristle at a 45 degree angle • Sweeping off the bristle after each tooth The studies by Schleuter [16] indicate that the most observed technique is the Horizontal Swipe technique (HST). This technique has several differences when compared to the MBT. When using the HST participants make a lateral movement over 5 or 6 teeth at the time. It can therefore cover a lateral movement as large as the maximum depth of the mouth. On the other hand the maximum lateral stroke length of the MBT should be of maximum 36 mm as this is the average arch length between the maxillary canine teeth [12] (two molars). A second difference between the two techniques is the orientation of the bristle. In the MBT the bristle should be at a 45-degree angle to clean between the teeth and gums whereas in the HST the toothbrush is held at a 90-degree angle. Finally, the off-sweeping movement of the MBT helps the bacteria to go off the teeth rather than spreading them. This rotational movement is not observed in the HST. Dynamics To transform the behavior, the dynamics of the action need to be de-coupled from the reaction. The toothbrush should counteract the horizontal swipe movement of the HST, which covers a larger distance than the MBT in the same time. The faster acceleration of the HST should be diminished to support smaller strokes. Expression Wensveen et al. [21] indicate that the expression has to do with the emotional effects when controlling an interactive product. Since the current toothbrush focuses on the skill, rather than the emotional experience, we will not take the expressive coupling into account for the current design. Detailing Design explorations We explored three different approaches to reduce the acceleration of the movement in the HST. Figure 1: Schematic interaction overview of the modified bass technique Synthesizing When designing the brush we focus on how users can learn to make small strokes of a maximum of two teeth at the time. We will consequently discuss how we can design for a tight coupling between the six aspects of interactive materiality and the toothbrush. A short analysis provides insight into the coupling. Location The brushing technique has an effect on the teeth. The relation between the teeth and the hand movement is localized in the brush. The feed forward should therefore be provided inside the brush. Direction When comparing the axial movement of both techniques there is an observable difference. When using the MBT, users need to swipe two teeth at the time, while the HST uses five or six teeth at the time. This direction is the lateral direction of the teeth and the feed forward should therefore be in the axial direction of the toothbrush. Modality When a user brushes her teeth, she can feel the teeth through the toothbrush in her hands. Therefore the feed forward should address the tactile senses. Through haptic feed forward users can perceive what happens in their mouth through their hands. Time In contrast to periodical indications of appropriate brushing behavior by the dentist, the toothbrush should act immediately upon the brushing patterns. Firstly, we investigated how the inertia effect could increase the difficulty of switching the direction of the brushing by adding a mass to the brush. When a user makes a fast movement in a specific direction the mass continues in that same direction after the user changes her direction. Unfortunately this method requires a very heavy toothbrush to achieve the desired effect. Secondly, we explored an elastic delay in which we assumed that an elastic band could reduce the handle’s action radius. When a movement is made, the delay between the handle and bristle is great enough to overcome the resistance. The effect on the interaction with the brush is an increase in stroke length before the bristle starts to move. When the startup-resistance is overcome, the elastic band makes the bristle snap back to the handle. Unfortunately, it is too difficult to make small movements with this brush. Thirdly, we explored whether we could change the resistance between the handle and the toothbrush head, by fixing a ball bearing between the two. This allowed us to design an interaction where the resistance determines the way in which one can interact with the toothbrush. It allowed us to set the toothbrush from a normal brush to one where the handle and toothbrush head operate separately. The interaction remains the same for the user as she can still move the handle. However, the hands can anticipate upon the resistance of the handle without moving the bristle. Resistance brush The resistance brush was used for further elaboration of the concept. The toothbrush should fully control the way in which the resistive difference is mapped and presented to the user. Therefore, we developed a prototype that could actively extend between the handle and the toothbrush bristle. The delay provided by the system caused the Figure 2: The linear ball bearing prototype participants to feel an offset during the interaction and response of the system. An active DC motor rotated a bolt to extend the toothbrush. However, this response was simply too slow to overcome the action speed of the user. Consequently, the linear ball-bearing system (see Figure 2) provided an opportunity to improve the coupling between the feed forward and the action. Thus the system uses the forces of the user to decouple the bristle from the handle. Thereby providing a tighter fit in time. The mapping was designed such that it feels as if the participant has a regular toothbrush when the right behavior is performed and handle loosens when the undesired behavior is performed. our toothbrush. Therefore, our quantifiable research question is: APPARATUS DESIGN Pre-experimental interview The base component of the toothbrush is a metal u-bar with a hole to which the head of a Lactona Travel Toothbrush is attached by means of a bolt. For hygiene purposes, this part was changed for each participant. Removing the bottom of the Lactona Travel Toothbrush Handle and attaching bolts altered the handle part of the travel toothbrush. The bolts in the handle provided a tight and low resistance fit to slide over the metal u-bar. A DC motor was integrated in the handle and had an extension to push itself onto the U of the metal bar creating a fixed toothbrush. This automated lock and release mechanism was designed to decouple the handle from the rest of the toothbrush. When the DC motor provides a high force it locks the slider and the handle and the toothbrush can be used as a regular toothbrush. When the toothbrush is used improperly, the DC motor releases the lock from the slider and the handle becomes loose. In the design, the threshold to release the handle was set when the maximum stroke of 36mm range of the canine arch, was exceeded for two consecutive strokes. The system could control the resistance between the handle and slider within approximately 50Hz. Also an accelerometer was integrated in the handle to measure the stroke length. Both DC motor and accelerometer were connected to a computer by means of an Arduino. The software program that determines the reaction of the DC motor based upon the data retrieved from the accelerometer and logs the sensory output was developed in Processing. See Figure 3 for a schematic overview of the system. EVALUATION An evaluation study was set up to analyze the differences between regular brushing and brushing with the designed toothbrush inherently responding to long brush strokes. Through this evaluation we intend to analyze the difference between the participants’ regular brushing techniques and the techniques stimulated by the inherent feed forward in • Does inherent feed forward support learning to brushing with shorter strokes? METHOD Participants Twenty-one students, all Bachelor students from the department of Industrial Design of TU Eindhoven, participated voluntarily in this study (8 female, evenly distributed, mean age: 19.55, STD 2.16). A semi-structured interview was conducted before the start of the experiment to gain insight into the brushing habits and techniques of participants. Questions on habits were employed to gain insight in the consistency of the participants (occurrence and duration), and their motivation to change behavior. Questions on technique were employed to verify if the users did not already use the MBT and to gain insight in the way they perceive their own ability to brush teeth. Experimental procedure After the interview, participants signed an informed consent form. To prevent side effects, participants did not get any instructions on how to brush their teeth upfront. Participants were distributed evenly over two groups, one group used the active brush the other used the passive brush (control group) during the tooth brushing session. The active condition provided insight into the effect of an inherent feedback system on the participants’ tooth brushing technique and the passive condition controlled for fatigue effects. A tooth brushing session consisted of two minutes (recommended brushing time) and was divided in three parts. The first 30 seconds served as baseline to determine the participants’ normal brushing behavior. The second part consisted of one minute brushing with the toothbrush in either active or passive condition. During the active condition the toothbrush responded to the participants’ brushing behavior. During the final 30 seconds the toothbrush returned to a passive mode. Experimental setup The participants were invited to brush their teeth at a sink with a water tap. They were provided with a mirror, a cup, a napkin, toothpaste and a toothbrush to conduct a regular tooth brushing session. DATA ANALYSIS RESULTS Stroke length analysis Qualitative Results To analyze the movement of the participants, the Arduino attached to a computer collected the movement data from the accelerometer (see Figure 3). From a pilot study, we obtained a normal frequency of tooth brushing of 5 Hz. To be able to integrate the acceleration data we sampled at 10 data points per stroke, consequently a sample rate of 50 Hz was taken. The length of the stroke was calculated by the following formula for distance: d = v*t + (1/2)*a*t^2. Before calculating the distance, the raw acceleration data was corrected for gravitational forces to obtain the actual acceleration. Therefore, the preceding 50 measurements of the accelerometer were averaged (averageOutput) and subtracted from the current output (currentOutput) value of X providing the acceleration value: a = currentOutputX – averageOutputX. The raw data was reviewed in MS Excel, where the length of the strokes was obtained by determining the local maximums of the stroke length. When the brushing stopped, data was excluded, as these breaks did not contain data on the movement. The average strokes were processed in SPSS and a paired-samples T-test was conducted to determine the differences between baseline measurement and brushing behavior in the final 30 seconds. The effect of the control group was analyzed to explore for a possible fatigue effect. Finally, the effect of the active condition was compared to the passive condition to determine the effect of our inherent feed forward intervention in which a longer stroke length indicates a poor brushing technique as it damages both the gums and enamel [15]. To control for equality between the two groups, Levene’s test on homogeneity of variances was conducted between the baseline measurements of both groups. Evaluation Feedback After the experiment an exit interview was conducted to understand how participants perceived the feed forward, how they would like to receive feed forward and how they deal with regular and proposed feedback mechanisms. Habits Most participants indicated that they brush their teeth twice a day: in the morning, before going to the university and at night before going to bed. The main reasons to not brush for two minutes are a constructive lack of feedback, as most of the participants do not look at a clock while they brush their teeth. Half of the participants indicate that being in a hurry affects their tooth brushing technique. Technique One participant was excluded from the current study as he already used the Modified Bass Technique and this behavior is envisioned as good brushing. The other participants all used the horizontal swipe technique or the horizontal swipe technique combined with a vertical swipe movement. Most of the participants indicated that they know what the right technique is because their dentist told them to use a specific method. 11 participants indicated that the dentist wants them to make smaller movements possibly with rotations, although the researcher did not observe these movements. Only three participants indicated that they follow up on the advice of the dentist (because they had dental braces). The most common behavior is to follow up on the advice of the dentist for two weeks after having the bi-annual check-up. Quantitative Results Levene’s test on homogeneity of variances indicates that there is no significant difference between the active and passive group of student volunteers during the baseline measurement F(1,18)=1.275, ns. There is no significant effect of fatigue in the passive control group between the zero-measurement (m=3.25±1.02) and the after-experimentmeasurement (m=2.86±0.93) t(9) = 1.82, p = ns. There is a significant difference between the active condition zero measurement (m=3.07±0.582) and the after-experimentmeasurement (m=2.44±0.602) t(9) = 13.61, p < .00 (see Figure 5). When the mean brush strokes and its standard deviations of all 3-second periods are mapped over time, a decreasing brush stroke trend can be observed when the active brush is set on passive (see Figure 4). Figure 3: A schematic overview of the used apparatus. Toothbrush head, metal u-bar, handle, DC motor, Accelerometer, Arduino and Mac Figure 4: Mean (M) brush strokes (in cm) and Standard deviations (SD) per three-second periods mapped over time for passive toothbrush: black (M) and blue (SD); active toothbrush: white (M) and red (SD) Evaluation Feedback Several participants were familiar with the feedback clock systems that indicate how long you are brushing. The students indicated that they would like to see visualizations of where they have brushed, or vibrations to gain insight when to switch sides. Participants who had the passive condition all indicated that they did not get any feedback of the system. Seven out of ten participants that had used the active toothbrush indicated doubtfully that they felt that the handle loosened at certain moments. They indicated that they did not understand what to do with the handle. Three participants did not feel any feed forward at all (while they did receive the feed forward). Figure 5: Mean brush strokes (in cm) and confidence intervals 95% with the passive toothbrush (blue) and the active toothbrush (green) DISCUSSION In the current experiment we have demonstrated that it is possible to transform a behavior by creating a tight coupling between the action and perception of the user. The results of our experiment indicate a significant difference between the brush strokes of participants using the active toothbrush and participants using the passive toothbrush. The brush strokes seemed to continue decreasing even after the feed forward had ceased. Consequently, this could indicate that giving guidance to users with a toothbrush providing inherent feed forward can support users in learning a more appropriate tooth brushing skill. In its current setup, we could only evaluate a behavioral change over one session of brushing. Given the context of tooth brushing we also considered expanding the experiment in a longitudinal study. However, such attempt would have required a much larger scale for statistical validity [13]. As we first wanted to evaluate whether the research question could be answered positively with the current prototype, we opted for the described experimental setup. As a follow up we should consider whether the change in behavior lasts after using the toothbrush for a prolonged period, and consequently if the behavior persists when the active toothbrush is replaced by a regular one. Our approach of transforming behavior by means of interactive materiality [19] may have considerable added value over the traditional approaches to learn new tooth brushing techniques. In a typical study such as [15] the main research question is whether one technique is superior to another. In these situations it is just assumed (or hoped) that participants do as they are told: “Instruction on MBT was performed using a model and seeing the video `Periodoncia No 1'”. There is no first person perspective in the training and the dependent variable is a Plaque Index Score. We believe that it makes more sense to improve technique by motivating the users to execute the right movement and sustain it. The cognition-based [17] and second- or third-person perspective-based instruction methods [2, 20] can be enriched, extended or even replaced by new tools, which give more accurate and more direct feedback. “Virtual Aquarium” [13] focuses on stimulating users to brush their teeth for 3 minutes. Playful Toothbrush [2] on the other hand focuses on number of brushing strokes and number of unbrushed teeth. The added value of the presently described toothbrush over the two examples mentioned above is that it supports users in learning the appropriate technique. By combining the advancements from all three experiments, one could support tooth brushing even further. Our research-through-design approach allows transforming behavior through product interaction. In the interviews conducted, many students elaborated that they had a hard time to follow up the feedback given by the dentist. While many noticed a form of feedback and did not understand what to do with it, the results indicate that the users did adapt towards the desirable behavior. Although the feed forward was still perceived, it was in the periphery of their perception, and thus did not rely highly on their cognitive skill. This is an aspect that requires further investigation. More subtle feedback could potentially lead to fully unaware interaction, as demonstrated by Bruns Alonso et al. [1] in their pen experiment, making the learning process rely solely on the perceptual motor skills. Researchers in HCI are still exploring interactive computing systems designed to change peoples behavior [4] from a highly cognitive approach, as indicated in the examples on tooth brushing [2, 13, 17, 20]. The toothbrush described in this paper implicitly influences users to transform their behavior by embedding the behavior modification in the interaction. The experiment has demonstrated that our motivation to shift from the cognitive skill approach of learning behaviors to a more perceptualmotor skill based approach is worthy of further exploration. Furthermore, the approach as presented by Stienstra et al [19], enabled a clear process to achieve our goal. The presented research offers a different perspective on how to design persuasive technology that is embodied in the interaction. By applying the concepts of inherent feed forward according to the principles of the Frogger framework [21] and following the approach of interactive materiality [19] – changing elements in the dynamics – we were able to transform a behavior of tooth brushing. By pursuing this method in other products that require skill development, these could become more meaningful, more natural to use and thereby perhaps less complex. Consequently, we suggest to continue exploring the design approach of interactive materiality for behavioral transformation in more diverse contexts. ACKNOWLEDGMENTS We would like to thank Loe Feijs for his support in coaching the project and his comments on the text. Finally, we thank the participants for their time and effort. Part of this work has been supported by the EC within the 7th framework programme through the FET Open scheme under grant agreement no. 309191. REFERENCES 1. Bruns Alonso, M., Hummels, C.C.M., Keyson, D.V., Hekkert, P.M. Measuring and adapting behavior during product interaction to influence affect. Pers. Ubiquit. Comput. (2011), 1-11. 2. Chang, Y.C., Lo, J.L., Huang, C.J., Hsu, N.Y., Chu, H.H., Wang, H.Y., Chi, P.Y., Hsieh, Y.L. Playful toothbrush: ubicomp technology for teaching tooth brushing to kindergarten children. In Proc. CHI2008, ACM Press 2008, 363-372. 3. Colgate brushing instructions. http://www.colgate.com/app/CP/US/EN/OC/Informatio n/Articles/Oral-and-Dental-Health-Basics/OralHygiene/Oral-Hygiene-Basics/article/Taking-Care-ofYour-Teeth.cvsp 4. Fogg. B.J., Persuasive technology: using computers to change what we think and do. Ubiquity (2002) 5. Formicola, A.J., McIntosh, J., Marshall, S., Albert, D., MitchellLewis, D., Zabos, G.P., Garfield, R. Populationbased primary care and dental education: a new role for dental schools. J. Dent. Educ. 63 (1999), 331-338. 6. Frens, J.W., Djajadiningrat, J.P., and Overbeeke, C.J. Form, Interaction and Function, An Exploratorium for Interactive Products. In Proc. of 6thADC (2003). 7. Ganss, C., Schlueter, N., Preiss, S., Klimek, J., Tooth brushing habits in uninstructed adults – frequency, technique, duration and force. Clin. Oral Investig., Springer Berlin (2009), 203-208. 8. Gibson, J.J. The theory of affordances. In R. Shaw & J. Bransford (Eds.), Perceiving, acting, and knowing: Toward an ecological psychology, (1977), 67-82. 9. Kahol K., Leyba M., Deka, M., Deka V., Mayes, S., Smith, M., Ferrara, J., Panchanathan, S., Effect of fatigue on psychomotor and cognitive skills. Am. J. Surgery (2008), 195-204. 10. MacKenzie, I. S., Sellen, A., and Buxton, W. A. A comparison of input devices in element pointing and dragging tasks. In Proc. CHI 1991. ACM Press (1991), 161-166. 11. Merleau-Ponty, M. Trans: Smith, C. Phenomenology of Perception, Routledge, London, UK, 2005. 12. Mills, L.F. Arch width, arch length, and tooth size in young males. Angle Orthod. 34 (1964), 124-129. 13. Nakajima, T., Lehdonvirta, V., Tokunaga, E. and Kimura, H. Reflecting human behavior to motivate desirable lifestyle. In DIS '08. ACM Press (2008), 405414. 14. Overbeeke, C.J., Djajadiningrat, J.P., Wensveen, S.A.G., and Hummels, C.C.M. Experiential and respectful. In Proc. of Useful and Critical (1999). 15. Poyata-Ferrera, M., Segura-Egea, J., Bullón-Fernández, P. Comparison of modified Bass technique with normal toothbrushing practices for efficacy in supragingival placque removal. Int. J. Dent. Hygiene 1 (2003), 110114. 16. Schlueter, N., Klimek, J., Saleschke, G., Ganss, C. Adoption of a toothbrushing technique: a controlled, randomised clinical trial. Clin. Oral Investig. (2010), 99106. 17. Soler, C., Zacarías, A., and Lucero, A. Molarcropolis: a mobile persuasive game to raise oral health and dental hygiene awareness. In Proc. of ACE’09, ACM Press (2009), 388-391. 18. Ståhl, A., Höök, K., Svensson, M., Taylor, A., Combetto, M. Experiencing the Affective Diary, Journal Article, Pers. Ubiquit. Comput. (2009), 365-378. 19. Stienstra, J.T., Bruns Alonso, M, Wensveen, S.A.G., Kuenen, C.D. Designing interactive materiality for transformation of behavior. In Proc. NordiCHI’12. ACM Press (2012), 21-30. 20. Sylla, C., Branco, P., Coutinho, C. and Coquet, E. TUIs vs. GUIs: comparing the learning potential with preschoolers. In Pers. Ubiquit. Comput. (2012), 16(4), 421-432. 21. Wensveen, S. A. G., Djajadiningrat, J. P., C. J. Overbeeke. Interaction Frogger: a design framework to couple action and function through feedback and feedforward. In Proc. DIS 2004, ACM Press (2004), 177-184. 22. Zhai, S., Milgram, P., and Buxton, W. The influence of muscle groups on performance of multiple degree-offreedom input. In Proc. CHI 1996. ACM Press (1996), 308-315. 23. Zimmerman, J., Forlizzi, J., and Evenson, S. Research through design as a method for interaction design research in HCI. In Proc. CHI 2007, ACM Press (2007), 493-502.