A Serious Game for Children with Autism Spectrum Disorder ... 3 - Games... · A Serious Game for Children with Autism Spectrum Disorder as a Tool for Play Therapy Alejandra Ornelas

A Serious Game for Children with Autism Spectrum Disorder as a Tool for Play Therapy

Alejandra Ornelas Barajas, Hussein Al Osman, and Shervin Shirmohammadi Distributed and Collaborative Virtual Environment Research laboratory (DISCOVER Lab)

University of Ottawa, Ottawa, Canada aorne025 | [email protected], [email protected]

Abstract—We propose a Serious Game (SG) for children on the Autism Spectrum Disorder (ASD) composed of a Tangible User Interface (TUI) and a Graphical User Interface (GUI). The TUI is built from physical Lego-like building blocks augmented with electronic modules. The proposed SG is envisaged as a play therapy tool aimed at improving autistic children’s social and cognitive skills. We investigate the effects of using our SG in a play therapy exercise, by running an empirical study that compares conventional clinical non-computer block-games with the proposed SG. In our preliminary experimental study, the proposed system showed an improvement in social interaction, collaborative play and exercise performance, and a decrease in solitary play. Our results suggest that the proposed system can be a useful tool for play therapy aimed for young children with ASD.

Keywords: Autism Spectrum Disorder; Tangible User Interface; Play Therapy; Special Education; Serious Game.

I. INTRODUCTION

Autism Spectrum Disorder (ASD) is defined as a developmental condition characterized by a marked impairment in social interaction and communication skills [1]. Findings of the CDC’s (Centers for Disease Control and Prevention) Autism and Developmental Disabilities Monitoring Network, show that about 1 in 68 children in the United States were identified with ASD [2]. Some of the most common signs of ASD include differences in the way a person communicates, learns, interacts, and behaves [1]. These signs usually appear during early childhood. The sooner the individual receives an intervention, the more likely he/she can live an independent life as an adult [3].

People with ASD typically present limitations in adaptive behaviour that engender Special Education Needs (SEN) [4]. During the past few years, researchers have reported positive results for the use of technology in special education [5][6][7]. With the growing popularity of mobile devices such as smartphones and tablets, video games are becoming even more prominent tools for education [8][9]. The term SG was first introduced by Abt in 1970 [10]. He devised games to improve education inside and outside the classroom. Ever since then, serious gaming has been redefined to encompass any game that presents a utilitarian purpose in education, health care, work training, military, among many others. More recent definitions of serious gaming refer exclusively to computer games (i.e. video games) that exhibit another goal beyond mere entertainment [12]. For this work, we adopt the latter definition, since we are solely interested in computer games.

TUIs, when used as a gaming interaction style, have proven to be useful for the education of children with special needs [13]. TUIs take advantage of the human ability to manipulate physical objects and allow users to create a connection between the physical and virtual worlds. Hence, children tend to find it easier to interact with a TUI compared to a GUI [13].

In this paper, we propose a SG targeting three basic skills: conceptual, practical, and social. We use MEGA BLOKS® as a TUI integrated electrically with a 3D graphics application running on a computer to create a cognitively stimulating and interactive game. We probe the following research questions:

1. Does the combination of a tangible and a graphical interface lead to an effective educational tool that enhances the basic skills of children with ASD (i.e. conceptual, practical and social)?

2. Is the SG more appealing to children with ASD compared to the non-computer games used for play therapy?

3. Does the SG increase the level of social interaction between children with ASD compared to the conventional non-computer approaches?

In this paper, we do not aim to conclusively answer the above questions through our preliminary and short-term experiment; however, we intend to shed some light on these queries. Hence, this paper serves as a preface for a deeper study of these questions.

The rest of this paper is organized as follows. Section II presents a review of related work on serious games and play therapy for children with special needs. Section III describes the characteristics of the proposed system along with the technical details of the implementation. Section I details the method used to evaluate the proposed SG. Section V presents and discusses the results obtained. Finally, section VI summarizes the conclusions of this work.

II. RELATED WORK

SGs are not necessarily video games. Hence, in this section, we will first describe relevant non-technological SGs intended for children with special needs. In particular, we will describe the concept of play therapy which forms the basis of this work. Second, since this paper presents a Computer-based SG for children with special needs, we will survey the latter type of applications. Third, we will describe the use of TUIs as a

mechanism for interacting with such applications. Finally, we will present an analysis that highlights the gap in existing work.

A. Non-technological Serious Games for Special Needs

As mentioned previously, SGs are not necessarily video games. In this section, we will present some of the non-technological SGs designed for children with special needs. More specifically, we will discuss the use of SGs during play therapy. Play therapy is a type of occupational therapy commonly used for children with developmental conditions such as ASD and other SEN. This technique allows children to improve their social and communication skills while playing in their own way [14].

Building blocks such as LEGO® are a popular tool used for play therapy. Several studies have shown the advantages of these toys in improving social skills. For example, LeGoff used LEGO® play, in both group and individual therapy, to improve the social competences of children with ASD [15]. He also demonstrated the long term benefits of this technique through a 3-year study that showed that “the LEGO® groups made twice the gains demonstrated by the control group on the VABS–SD (Vineland Adaptive Behavior Scales–Socialization Domain)” [16]. And more recently, LeGoff et al. released a book on LEGO®-Based Therapy where they provide an evidence-based manual for professionals working with children with autism and related conditions [17].

B. Serious Videogames for Special Needs

Previous work has shown positive results in the use of technology for children with SEN [6][9][18]. Particularly, researchers have developed numerous SGs that aim to improve the social and communication skills of children with ASD [3] [5][19][20][21]. ComFiM [5] and Open Autism Software apps [3] are examples of collaborative games dedicated to encourage positive social interaction among children with ASD using a tablet with a touch screen. Similarly, GameBook [19] is an interactive storyteller mobile application that aims to improve social and cognitive skills while promoting motivation and imagination.

It has also been proven that it is possible to develop a SG using a participatory approach [18][22]. Bossavit and Parsons presented a work where high functioning autistic teens co-designed a SG to improve their academic skills in Geography [18]. Fletcher-Watson et al. worked closely with pre-school children with ASD to develop an iPad application where the children collaborated in the design process [22].

Other studies have shown the influence of avatars and virtual characters on children with ASD [20][23][24][25]. LIFEisGAME [20] and emot-iCan [23] are SGs aimed to help children recognize and express emotion through facial expressions. Furthermore, Carter et al. documented the interaction between these children and avatars [25]. Their results show an improvement in nonverbal social behaviors when avatars with exaggerated facial motions are used.

C. Tangible User Interfaces for Children with Special Needs

While most videogames are digital, TUIs offer an interesting way of coupling the physical and virtual worlds. Sitdhisanguan

et al. demonstrated through their experiments that a TUI-based system offers improved ease of use compared to mouse-based systems. In terms of learning efficacy, TUI-based systems show better results over touch-based systems [26]. Moreover, a TUI with an embedded digital system has shown social benefits that are promising in special education [27].

Block-based toys have been used as educational tools targeting different types of skills. Zuckerman et al. introduced the Flow Blocks, a set of components that typically developing children use to build causal systems [28]. Farr et al. introduced Topobo, a construction toy with programmable movement of plastic blocks [27]. The game was used with autistic children to improve their social skills; their findings suggest that toys with embedded digital technology may lead to an increase in social interaction. Humanoid robots are another example of TUIs used during therapy and special education for children with ASD [29][30][31]. The ROBOSKIN project based on tactile feedback is an example of the latter [29].

D. Gap Analysis

There is no doubt that serious gaming has proven to be a motivating educational method [8][7][32]. But, purely GUI based videogames lack the interaction with physical objects, while TUIs have shown to be engaging and have improved collaborative play in previous cases [13][27].

One of the biggest challenges when developing a SG for children with ASD is to address social interaction and communication problems. There are different known strategies used to improve communication skills during therapy exercises for children with ASD. Some children with ASD respond better to visual instructions, rather than verbal ones. Also, one of the most successful strategies to improve social interaction during exercises is to use visual feedback [11].

Hence, we identify two gaps in existing systems:

• Purely GUI based videogames for autistic children lack the interactive quality that TUIs provide

• Conventional therapy techniques that use tangible objects (e.g. LEGO® Therapy) lack the visual feedback that autistic children often respond to

We propose to add a GUI to the conventional non-computer LEGO®-Based Therapy as a form of visual feedback; we will examine whether the resulting game becomes more appealing and engaging. We will also evaluate whether the proposed game encourages more social interaction during group play.

III. PROPOSED METHOD

We adopt an approach in serious gaming that combines a TUI and a GUI. Our choice of TUI is building blocks due to the proven social benefits and popularity of the toy. Specifically, we use MEGA BLOKS® maxi, with the dimensions of 3 × 3 × 6 cm3. With this interactive game, we intend to create a tool to stimulate the basic skills of children with ASD and improve their adaptive behaviour. The goal of the proposed game is to replicate a model shown on a computer screen with the actual blocks (i.e. TUI). During game execution, the computer will provide real-time feedback as guidance to complete each level.

Fig. 1. Tangible User Interface Diagram

A. The TUI

As mentioned previously, the proposed SG uses building blocks as a TUI. We use big blocks to accommodate ASD children with fine motor skills difficulties. Also, the big blocks offer more space for the integration of electrical elements.

1) The blocks Internally, each one of the blocks has a spring with one end

attached to the top and the other to a circuit; the spring allows the circuit board to move inside the block (see Fig. 1). The circuit board consists of a resistor connected to two terminals. The terminals are concentric circles. The value of the resistance determines the ID of the block (see TABLE I).

TABLE I. RESISTANCE VALUES FOR THE BLOCK ID

Block ID colour Red Blue Green Yellow Resistance (Ω) 0 2.2k 3.3k 4.7k

Fig. 2. System communication

2) The board As shown in Fig. 1, the board is a square constructed using

36 individual blocks (6 by 6 matrix). It consists of all red blocks, except the 4 middle blocks which are yellow, to serve as a

reference point for the child to more easily compare the model shown on the screen to the physical one being built. The game itself uses red, green, blue, and yellow blocks. Inside the board there are four microcontrollers, one master and three slaves communicating though an Inter-Integrated Circuit (I2C). Each stud (head of the block) on the board has two concentric terminals, resembling the ones from the circuit board inside the blocks (see Fig. 1). All interior terminals are connected to the VCC power supply and each external terminal is connected to an ADC channel on one of the microcontrollers. The microcontrollers map the ADC channels to a position on the board. The position is indicated by a different character assigned to each stud (i.e. A-Z, 0-9).

Fig. 3. Microcontroller UML State Machine

Fig. 4. Example of Mini-games

3) Functionality When a block is placed on the board, the resistor gets

connected to the ADC port of one of the microcontrollers (see

Fig. 1). Then, the microcontroller receives the block ID, given by the resistance. The slaves map the ADC channel to a position on the board and then send the information to the master through I2C (see Fig. 2). The TUI reads the position and the ID of the block, and this information gets coded into a three character command (See TABLE II). The master finally sends that command to the computer via serial communication (i.e. UART). Similarly, when a block gets removed from the board, the command gets passed to the computer. When the computer receives the latter information from the TUI, it maps the real blocks to its graphical representation. Fig. 3 shows the details of the process running on the microcontrollers using a UML state machine. The SG software records the performance of the user comparing the information sent by the TUI to the graphical model display on the GUI.

TABLE II. COMMAND EXAMPLES

Command A r n

A (position) character associated with stud on the board r (colour) r = red, b = blue, g = green, y = yellow n (state) n = new block, r = block removed

B. The GUI

The game features a 3D GUI to represent a virtual view of the board and the blocks. From the main menu of the game, the player can choose between different available mini-games. Each mini-game consists of a different model representing a number, letter, or pattern. 0 shows few examples of these mini-games. The player is assigned the task of replicating the model given by the computer onto the real board (i.e. TUI). Each time a block is placed on the board, the TUI reads the position of the block and relays the information to the GUI. Fig. 4.1 and 4.2 show a transparent white model. This means that the user(s) only needs to match the pattern (i.e. the number 2 and letter A). Fig. 4.3 and 4.4 depict colored models and hence the user(s) has to match both colour and pattern.

Fig. 5 shows a sequence of moves to complete a mini-game. If a block is placed on the correct position (Fig. 5.2), the GUI presents a positive visual effect (e.g. glowing particles) and the transparent block becomes opaque. If the block is placed on the wrong position, the GUI outputs instructive feedback to remedy the problem (Fig. 5.4). We minimized the negative feedback to avoid any possible frustration by the users and to encourage them to try again. Once the user(s) completes building the model, the game displays a success message (Fig. 5.6) and asks the user(s) to remove all the blocks to proceed to the next mini-game.

Fig. 5. A mini-game in action. Transparent blocks are the white ones on the monitor

IV. EVALUATION

We conducted an experiment that compares the proposed SG with a non-computer block game, similar to the one described in [17] during a LEGO® Therapy session. Our experiment was approved by the Office of Research Ethics and Integrity at University of Ottawa (File Number: H06-16-09). The participants were recruited from Children at Risk, a community organization in Ottawa that provides services to families of children diagnosed with ASD. Due to the nature of this research on a special population, the number of participants was limited. Nine children between the age of 6 and 15 years old (M=10, SD = 3.20) volunteered to participate. All participants were diagnosed with ASD according to clinical opinion, and all of them were boys. Given that ASD is significantly more common among boys than girls [2], it was not surprising that all volunteers that came forward were male.

The experiment was conducted during two consecutive days; five children were tested on the first day and four children on the second day. The total duration of the study was approximately one hour each day. Parents of the participants were present at all times.

A. Procedure

On Day 1, at the beginning of the experiment, the parents of the participants signed a consent form and answered a questionnaire about their children. The questionnaire, shown in the Appendix, included questions about their children’s experience working with mobile applications and educational games. After that, the children were divided into two teams (Team A and Team B) according to their age. Each team had between two and three members. For the first phase of the experiment, Team A was assigned to play with a non-computer game (conventional method) and Team B was assigned to play with the computer game (i.e. proposed SG). Both teams received the instructions for their assigned game. They were given 15 minutes to play. The difficulty level of the game was adjusted when necessary (i.e. if the teams found it too easy or too hard to complete).

For the non-computer game, participants were asked to play with blocks without the graphical interface. Instead, they received a card depicting an image of the model they need to replicate with the blocks; a similar approach used on LEGO® Therapy [17]. For the computer game, participants were asked to play with the proposed SG. Both teams were instructed to build the same models.

For the second phase of the experiment, the teams switched games and they were given the same time (15 minutes) to play. Finally, at the end of the experiment, participants were asked to choose their preferred game. On Day 2, the same procedure was repeated for Team C and Team D.

B. Measurements

We focused on two aspects in our evaluation of the two games: social interaction and performance. We used an observational grid to record the following five variables:

• Social interaction: Number of initiated social interactions between the participants.

• Solitary play: Time the participant spent playing alone.

• Collaborative play: Time the participant spent playing with other participants.

• Engagement in other activities: Time the participant was not engaged in the game.

• Performance: Number of mini levels (i.e. models) they successfully completed.

V. RESULTS AND DISCUSSION

For the statistical analysis of the results, a paired-samples t-test was used to compare both games for each of the previously described five variables.

There was a significant difference in four out of five variables (see TABLE III). The computer game showed an increase in the number of social interactions over the non-computer game (p = 0.005) with 3.55 interactions on average, compared to 1.11. Participants spent 45.6% of their time playing alone with the non-computer game, but they only spent 25.5% in solitary play time with the computer game. Which means that the solitary play time was reduced (p = 0.044) by more than 20%. In the same way, collaborative play time was significantly increased (p = 0.004) by almost 30%. The level of engagement seemed to be slightly better when the computer game was employed, however the results were not statistically different (p = 0.330). And finally, we observed a better performance during the computer game (p = 0.009), based on the number of mini-games completed going from 2.44 to 6.11.

TABLE III. EXPERIMENT RESULTS

Variable Non-Computer game Computer game

Mean σ Mean σ Social interaction a 1.11 1.27 3.55 1.74

Solitary play b 45.6% 34.55 25.5% 20.19

Collaborative play a 26.07% 42.13 56% 27.68

Engagement in other activities 28.29% 29.01 18.5% 21.37 Performance a 2.44 2.35 6.11 4.70

a. p < 0.005

b. p<0.05

Discussion

As it has been previously shown, serious games involving building blocks can be used to promote social skills for children with ASD [16][17][27]. But also, the results from our experiment suggest that computer games can encourage more social interaction and collaborative play than similar non-computer games. Our conclusions are compatible with existing works on the topic [3][7][9][32]. Having a more interactive game seems to motivate children to work together on a common goal. In the same way, TUIs provided additional help to enhance the social benefits of the game. The children enjoyed manipulating physical objects and seeing the effects on the

computer screen as a result of their actions. The latter finding agrees with Farr et al. [27] and Sitdhisanguan et al. [26].

We also noted in our experiment that the children’s performance was significantly better with the computer game as opposed to the non-computer game. The visual and interactive feedback offered by the computer allowed them to complete the game’s levels faster as they were able to follow instructions more easily and intuitively. However, there was one subject who liked the negative feedback presented on the GUI and consequently committed mistakes on purpose. One possible way to resolve this issue in future iterations of the SG is to increase the visual stimulation for success (e.g. display smiley faces, confetti, and sounds) and reduce it for mistakes. Our results support the notion that visual support is a very important factor to consider when working with ASD children.

We found that the time the participants spent engaging in other activities did not present a significant difference between the two compared games. We think that the way the experiment was set up might have affected this result. There were some distractors in the room that interfered with the dynamics of the games. For instance, since the parents were present at all times, some of the children got distracted interacting with them, and hence had some difficulty following instructions. Also, the children did not have a prior relationship with the researcher before the experiment. This might have made it more difficult for them to trust her and follow her directions. Also, some of the older and higher functioning participants lost interest in both games after a few minutes, as the games were too easy for them. This can be avoided by including more challenging levels and using smaller blocks. Furthermore, because there were two teams playing with different games at the same time, some of the children wanted to play with the computer game while they were supposed to be playing with the non-computer game.

As previously mentioned, the non-computer game we used was based on LEGO® therapy [17]. However, we adapted the game to allow for a comparison with the proposed computer game. For instance, we did not use LEGO® blocks; instead we used MEGA BLOKS® (bigger blocks). And even though we tried to assign the roles of “Engineer and Builder” as proposed by LeGoff, only one team was able to assume such roles. This may have happened because the participants were not familiar with the method and possibly needed more time to process the rules. One of the parents suggested using hats with different colours to help children differentiate between roles. We will adopt this advice in future experiments.

There were some limitations to our experiment. Although we observed statistically significant differences on several key metrics between the non-computer and computer game, the number of subjects employed limits our ability to generalize our results. However, this is a common limitation among work on the topic. Also, the experiment did not address long term behavioral changes achieved with the help of the proposed SG. Our immediate aim is to demonstrate the potential for such games in encouraging social interaction and achieving a higher level of engagement. Further investigation is required to study long term effects of employing the proposed SG.

VI. CONCLUSIONS

In this paper, we propose a SG for children with ASD using a combination of TUI hardware and GUI software. During our experiments, we observed significant differences between the proposed SG and a conventional non-computer game used for play therapy. Similar to Farr et al. [27], we found that TUIs can stimulate social interaction and collaborative play in children with ASD. We observed that children engaged in less solitary play and achieved better performance with the computer game. Furthermore, we detected some short-term benefits for visual feedback during play therapy. Our results suggest that using a SG, consisting of a blocks-based TUI and a GUI, offers a promising method that could be used in play therapy for young children with ASD.

For future work, we intend to use a bigger sample size to see if our results stand. And also, we will limit the age of participants to 9 years old, since the older children did not show interest in playing with big blocks.

APPENDIX - QUESTIONNAIRE

REFERENCES

[1] American Psychiatric Association, “Autism Spectrum Disorder,” Diagnostic and Statistical Manual of Mental Disorders, 5th ed., Arlington, VA: American Psychiatric Association, 2013, pp. 50-59.

[2] Autism and Developmental Disabilities Monitoring Network, “Community Report on Autism,” Centers for Disease Control and Prevention, United States, 2016.

[3] J.P. Hourcade et al., “Evaluation of tablet apps to encourage social interaction in children with autism spectrum disorders,” Proc. of the SIGCHI Conf. on Human Factors in Computing Systems. ACM, 2013.

[4] G. Baird et al., “Prevalence of disorders of the autism spectrum in a population cohort of children in South Thames: the Special Needs and Autism Project (SNAP),” The lancet, vol. 368, no. 9531, pp. 210-215, July 2006.

[5] P. C. Ribeiro and A. B. Raposo. "ComFiM: a game for multitouch devices to encourage communication between people with autism." Serious Games and Applications for Health (SeGAH), 2014 IEEE 3rd Int. Conf. IEEE, 2014.

[6] C. Afonseca and S. Bermúdez. "Supporting collective learning experiences in special education." Serious Games and Applications for Health (SeGAH), 2013 IEEE 2nd Int. Conf. IEEE, 2013.

[7] J.L. González et al., “Using videogames in special education,” Int. Conf. on Computer Aided Systems Theory. Springer Berlin Heidelberg, 2007. vol. 4739, pp. 360-367.

[8] M. Griffiths, “The Educational Benefits of Videogames,” Education and Health, vol. 20, no. 3, pp. 47-51, 2002.

[9] K. Durkin et al., “Video Games for Children and Adolescents With Special Educational Needs,” Zeitschrift für Psychologie, vol. 221, no. 2, pp. 79-89, 2015.

[10] C.C. Abt, “Serious games: The art and science of games that simulate life,” New Yorks Viking, 1970, p. 6.

[11] L. A. Hodgdon, “Visual Strategies for Improving Communication,” Practical supports for school and home. Troy, MI: QuirkRobertsPublishing, 1995.

[12] S. Egenfeldt-Nielsen et al., “Serious Games- When Entertainment is not Enough,” Understanding video games: The essential introduction, 3th ed., New York, NY: Routledge, 2009, pp. 205-222.

[13] L. Xie, Lesley et al., “Are tangibles more fun?: comparing children's enjoyment and engagement using physical, graphical and tangible user interfaces,” Proc. of the 2nd Int. Conf. on Tangible and Embedded Interaction. ACM, 2008, pp. 191-198.

[14] A.H. Morgenthal, Child-Centered Play Therapy for Children with Autism: A Case Study, Ph.D. dissertation, Clinical Psychology, Antioch University New England, Keene, New Hampshire, 2015.

[15] D.B. LeGoff, “Use of LEGO© as a therapeutic medium for improving social competence,” Journal of Autism and Developmental Disorders, vol. 34, no. 5, pp. 557-571, 2004.

[16] D.B. LeGoff and M. Sherman, “Long-Term Outcome of Social Skills Intervention Based on Interactive LEGO ® Play,” SAGE Publications and The National Autistic Society, vol. 10, no 4, pp. 317-329 , 2006.

[17] D.B. LeGoff et al., LEGO®-Based Therapy: How to build social competence through LEGO®-based Clubs for children with autism and related conditions. Jessica Kingsley Publishers, 2014.

[18] B. Bossavit and S. Parsons, “"This is how I want to learn": High Functioning Autistic Teens Co-Designing a Serious Game,” Proc. of the 2016 CHI Conf. on Human Factors in Computing Systems. ACM, 2016, pp. 1294-1299.

[19] P. Cunha et al., “Augmented reality for cognitive and social skills improvement in children with ASD,” 13th Int. Conf. on Remote Engineering and Virtual Instrumentation (REV), Madrid, 2016, pp. 334-335.

[20] B. Abirached et al. "Improving communication skills of children with ASDs through interaction with virtual characters." Serious Games and Applications for Health (SeGAH), 2011 IEEE 1st Int. Conf. IEEE, 2011.

[21] L.E. Boyd et al., “Evaluating a collaborative iPad game's impact on social relationships for children with autism spectrum disorder,” ACM Transactions on Accessible Computing (TACCESS), vol. 7, no 1, pp. 3, 2015.

[22] S. Fletcher-Watson et al., “Designing for young children with autism spectrum disorder: A case study of an iPad app,” International Journal of Child-Computer Interaction, vol. 7, pp. 1-14, 2016.

[23] D. Sturm et al. "eMot-iCan: Design of an assessment game for emotion recognition in players with Autism." Serious Games and Applications for Health (SeGAH), 2016 IEEE Int. Conf. IEEE, 2016.

[24] D. Moore et al., “Collaborative virtual environment technology for people with autism,” Focus on Autism and Other Developmental Disabilities, vol. 20, no 4, pp. 231-243, 2005.

[25] E.J. Carter et al., “Investigating the Influence of Avatar Facial Characteristics on the Social Behaviors of Children with Autism,” Proc. of the 2016 CHI Conf. on Human Factors in Computing Systems. ACM,2016.

[26] K, Sitdhisanguan et al., “Using tangible user interfaces in computer-based training systems for low-functioning autistic children,” Personal and Ubiquitous Computing, vol. 16, no 2, pp. 143-155, 2012.

[27] W. Farr et al., “Social benefits of a tangible user interface for children with Autistic Spectrum Conditions,” Autism OnlineFirst, vol. 14, no. 3, pp. 237-252, 2010.

[28] O. Zuckerman et al.. “Flow blocks as a conceptual bridge between understanding the structure and behavior of a complex causal system,” Proc. of the 7th Int. Conf. on Learning Sciences. International Society of the Learning Sciences, pp. 880-886, 2006.

[29] B. Robins and K. Dautenhahn, “Tactile Interactions with a Humanoid Robot: Novel Play Scenario Implementations with Children with Autism,” International Journal of Social Robotics, vol. 6, no. 3, pp. 397-415, 2014.

[30] B. Robins et al. “Tactile interaction with a humanoid robot for children with autism: A case study analysis involving user requirements and results of an initial implementation." 19th Int. Symp. in Robot and Human Interactive Communication. IEEE, pp. 704-711, 2010.

[31] D.C.M. François, Facilitating play between children with autism and an autonomous robot, Ph.D. dissertation, School of Computer Science, Faculty of Engineering and Information Sciences, University of Hertfordshire, UK, 2008.

[32] A. Hiniker et al., “Go go games: therapeutic video games for children with autism spectrum disorders,” Proc. of the 12th Int. Conf. on Interaction Design and Children. ACM, 2013, pp. 463-466.