Learning Through Reflection at the Tabletop: A Case Study with Digital Mysteries Ahmed N.S. Kharrufa Patrick Olivier Culture Lab, School of Computing Science Newcastle University, Newcastle upon Tyne {ahmed.sulaiman, p.l.olivier}@ncl.ac.uk

David Leat Research Centre for learning and Teaching Newcastle University [email protected]

Abstract: In this paper we present our work in investigating digital tabletop technology as tools for supporting reflection in collaborative learning environments through the design of a digital version of the mysteries paper-based learning technique. We identified the different types of reflection as supported by computer technology and, through an iterative design process, finalized the design of digital mysteries and evaluated its support for reflection through 12 trials with six groups of students in a school environment. By having two of the groups use the application for four trials, we were able to show clear evidence of reflective interactions that positively influenced subsequent trials. This work aims at demonstrating the unique advantages of digital tabletop technology that allows it to combine the benefits of traditional face-to-face collaborative settings and the support of digital technology.

(a) (b) Figure 1: Mysteries application. (a) a number of students using the application. (b) A capture of the completed solution showing a branched sequence that reflects the students answer to the mystery.

Introduction The benefits of reflection for learning are well established (Boyle, 1997; Collins & Brown, 1988, Baker & Lund, 1997). It can improve the learning experience of the students by making them think back on how they solved the problem, and consider their mistakes and alternative problem solving strategies. In fact, some researchers go as far as considering reflection to be ``the ultimate expression of education" (Boyle, 1997). Computers, and due to their ability to record, process, and replay learning sessions, are potentially powerful tools for supporting reflection that can focus students’ attention directly on their thinking process (Collins & Brown, 1988). Moreover, when students collaborate in solving a problem, they engage in certain reflective group interactions that do not occur in individual learning, such as explanation, justification and disagreement (Dillenbourg, 1999; Baker & Lund, 1997). Research on computer supported collaborative learning (CSCL) has mostly been limited to networked environments where the learners collaborate co-located or distant, synchronously or asynchronously, but rarely in a face-to-face setting (Hilliges et al., 2007; Lehtinen, 2003). Comparisons between CSCL and face-to-face collaboration are often presented as mutually exclusive (Resta & Laferriere, 2007). Consequently, investigating CSCL’s support for reflection has only been conducted using networked environments (Baker & Lund, 1997). Baker and Lund justified

this by arguing that face-to-face collaboration uses verbal interactions that are hard to structure, unlike the textual interactions usually used in a networked environment. We describe our work in using computers as tools for reflection in a collaborative, yet face-to-face environment. This is made possible through the use of digital tabletop technology rather than traditional desktops. Digital tabletop (or simply tabletop) technology has recently attracted widespread interest in the research community as it allows us to bring computer-based support to familiar collaborative settings around traditional tables. Tabletops can be defined as large-horizontal surfaces that act as computer displays which allow multi-synchronous interaction (mostly pen- or finger-based) in a face-to-face setting. This distinguishes tabletops from other computer-based collaborative settings because they combine both the social affordances of traditional tables, and the digital affordances of computers. We investigated the support tabletops can provide for reflection through the development of digital mysteries, a digital adaptation of a paper-based learning tool called mysteries. A thorough iterative design process was used to develop a collaborative learning application which is specifically tailored to utilize the unique affordances of digital tabletops in supporting reflection (the focus of this paper) in addition to encouraging effective collaboration, making thinking visible, and bringing the students’ attention to their problem solving strategy. We first summarize the types of reflection that the use of computers can support, as identified by different researchers (Collins & Brown, 1988; Nunes et al., 2003; Baker & Lund, 1997). After explaining the paper mysteries learning tool, we describe the design of digital mysteries, how different types of reflection are supported, the trials conducted to evaluate the final design, and conclude with a discussion of the work and proposals for future development.

Types of Reflection Reflection is important to learning because it helps make students aware of their thinking process (metacognition) and problem solving strategies. With reflection students can derive abstractions about their thinking process and compare it with their earlier performances or to the performances of others. This enables them to identify weaknesses and areas for improvement that can greatly increase the benefits gained from any problem solving process. If mistakes and weaknesses in problem solving are not highlighted and discussed, students are likely to repeat them in later attempts. It is important to highlight that when different researchers talk about supporting reflection, they can in practice be talking about different types of reflection. We have identified three distinct types that are relevant to collaborative learning activities around digital tabletops: post-activity, inter-activity, and part-ofactivity reflection. Post-activity reflection: this refers to the provision of a “playback” of a certain problem solving activity (or even an athletic activity like serving a tennis ball), in a number of ways after the completion of the task, in order to identify strengths and weaknesses in the process (Collins & Brown, 1988). Collins & Brown categorized four types of postactivity-reflection: 1. Imitation: The teacher imitates what the student did highlighting the correct and incorrect aspects and the critical moments in the process. An obvious weakness in this approach is that it depends on how accurate the teacher’s imitation is, and how confident the student is about the teacher’s imitation. 2. Replay: A conventional video playback of the process. If done with the supervision of the teacher, and as with imitation, correct and incorrect aspect and critical moments in the process can be highlighted. 3. Abstracted replay: Recordings of only critical aspects of the process, and possibly from different perspectives. This type is particularly useful when too much data is involved, and thus it helps in keeping the student’s focus on the important aspects of the process. The important issue here is in finding out the right level and type of abstraction. 4. Spatial reification: This is a form of a static visualization of the process that displays time spatially allowing the student and the teacher to identify critical aspects and moments of the process quickly. It is clear that while a teacher, with the use of video playback, can provide the first two types of reflection support, this is difficult to integrate into the classroom context. Furthermore, digital technologies have the potential to play a very useful role in providing the latter two types of support. Inter-activity reflection: this refers to the induction of reflection moments, that is a chance to pause and think about a certain (probably incorrect) action or decision, during the activity (Nunes et al., 2003). Nunes et al. defined the types of activities that allow the integration of reflection support to be those that constitute problems that do not have a direct path leading to the solution, that is, those that involve multiple steps and strategies. This is basically because

problem solving inherently requires reflection to make decisions about which steps to take as well as why and when they must be employed. The benefit of these multi-stage, mutli-path types of problems is that they admit the observation of students solving the problem and construction of a database of common error patterns. This makes it possible for a computer to identify problems and intervene at appropriate times, usually at the boundaries between steps. Nunes et al. (2003) suggested that such tasks should be followed with activities that encourage students to think about (reflect on) their overall problem solving process to maximize the benefit from the session, which can be any combination of the types of reflection that Collins & Brown (1988) identified. Part-of-activity reflection: Designing tasks to include reflective interactions as part of the activity. In collaborative learning such interactions include explanation, justification, and evaluation which are more likely to occur in collaborative tasks than in individual learning (Baker & Lund, 1997; Dillenbourg, 1999). The type of problems specified by Nunes et al. can still apply to collaborative problem solving tasks, with the additional requirement that the task be chosen to allow a space for the above types of reflective interactions to occur. In their attempt to introduce computer-based support for such collaborative tasks, Baker & Lund (1997) used a networked-based environment which allows for the provision of different levels of structuring to the interaction between students and justified this by arguing that face-to-face collaboration, even around computers, uses verbal interactions that are hard to structure. Each of these classes (and sub-classes) of reflection can be useful according to the learning context, and the greatest benefit is likely to be achieved by providing more than one type to the student. Without computer support, it is only possible to provide a limited type and level of these classes, and this requires designing collaborative learning activities that require reflective interactions, video taping the process, providing continuous scaffolding on the part of the teacher to make sure that the reflective interaction does in fact occur, inducing reflection during the process, and finally showing the students a simple video playback of their session to highlight critical aspects. While this can be (and has been) applied in case studies that investigate the benefits of reflection, it is not a practical addition to standard learning activities.. Interactive digital technologies have the potential to play an important role in supporting the first two types of reflection activities. But if reflection is to be supported for collaborative learning tasks, it must either be for networked tasks where each student has his own computer, although this will make it difficult for the teacher to provide scaffolding on an individual basis; or if students collaborate in a shoulder-to-shoulder setting around one machine with a single point of access, which also makes it difficult to identify the student initiating actions (and place students in less familiar collaborative setting). We argue that digital tabletop technologies provide educators with the unique advantage of being able to provide all types of reflection support for face-to-face collaborative tasks and demonstrate our approach through the develop of digital mysteries.

Paper Mysteries Mysteries (Leat & Nicholas, 2000; Wright & Taverner, 2008) is a collaborative learning tool for the development and assessment of students’ higher level thinking skills. Its design aims to make evident students’ cognitive processes through the physical manipulation of data slips to solve a given mystery, which is usually an open question in mathematics, history, geography, or any other subject (a sample question is “Why is the village shop in Hensford closing?”). The data slips represent clues about the mystery in the form of facts, background information, abstract ideas, or even red herrings. When solving a mystery, it was observed that higher achieving students usually go through three stages: (1) reading all the slips; (2) categorizing them in groups; and finally (3) building a sequence of slips that embodies their answer. During this process, students engage in collaborative activities such as discussions, disagreements, and justification, thereby creating their zones of proximal development (Wright & Taverner, 2008). That is, students with different levels of knowledge about the subject of the mystery start to benefit from each other through discussions and the sharing of ideas. Teachers can assess the level of thinking of the students through careful observation. A stimulated recall session involving questions from the teacher about specific actions in the process has been suggested to increase the effectiveness of the learning experience and induce reflection, but this requires the session to be video recorded and then played back (the full session) to the students. This recording and play back process, other than for research purposes, is impractical and time consuming. Mysteries, with its open question, and multi-stage collaborative design that requires students to discuss and justify their actions to others, satisfies the criteria for the type of activity that allows all types of reflection support discussed above. Moreover, since it is a task originally designed to be carried out in a face-to-face setting around a table, it is a natural candidate for a digital tabletop learning application.

Work related to Tabletops and Learning Research on tabletops started with the pioneering work of Wellner (1993). However, to date there have been few tabletop interaction studies designed to focus on comprehension and higher level thinking skills (and conducted in realistic setups). Indeed, most previous research has been exploratory in nature, focusing on layout design and physical manipulation of virtual objects or on performing comparative studies between tabletops and other settings (Piper & Hollan, 2009; Do-Lenh et al., 2009). Previous research has, however, emphasized the motivating nature of tabletops and consequently higher level of engagement as compared to other media. Rick and Rogers (Rick & Rogers, 2008) focused on the adaptation process of a desktop learning application, DigiQuilt, to a multi-touch surface, DigiTile. In their work, which focused on the design and not evaluation, they used learning theory to motivate the transformation, guide the design process, and inform the evaluation. SIDES (Piper et al., 2006) targeted people with social skill difficulties and aimed to help them practice effective group working skills like negotiation, turn-taking, active listening, and perspective taking. Piper and Hollan (2009) focused on tabletops as a studying tool for pairs of students (undergraduates in this case) compared to studying on paper. Their findings showed that students with paper material made more detailed notes and worked serially, while students with the tabletop display repeated activities more and performed better on exams. Morgan and Butler (Morgan & Butler, 2009) considered a range of theories in the design of collaborative learning tools, and proposed a number of (unimplemented) design features for concept mapping and storytelling applications based on notions of division of labor and role assignment. Do-Lehn et al. (2009) investigated the effect of tabletops with tangible interfaces on collaborative learning. They compared between a concept mapping task performed around a traditional computer display with one mouse/keyboard input, and a tabletop based one with some tangible characteristics, and measured the differences between individual and group learning gains. Their results showed no significant effect on individual learning gains, but also showed that groups in the traditional computer setting, due to the single point of access, learned significantly more from their partners than in tabletop setting.

Digital Mysteries The design process Digital mysteries was developed using an iterative design process that consisted of building three versions of the application, each tried using a small number of groups in a school and submitted to detailed evaluation which led to the next version. In total we conducted 22 trials in a school over a period of 18 months with each trial carried out by a group of three students (age range 11-14 years). The design process started by observing how students solved paper based mysteries and compared their use with a first version of digital mysteries that represented a very basic digital translation of the paper version without any meaningful additional functionality. This was followed by a second digital version that implemented the first concepts of the new design; and ended with a third revised version which we will discuss here. Observing how students solved paper mysteries and the two first iterations of the digital mystery allowed us to identify common error patterns of the students, as recommended by Nunes et al. (2003), and therefore made it possible for the final version to identify these patterns and provide appropriate intervention as will be explained in later sections. While we focus only on the design aspects related to reflection support, the design of digital mysteries targeted a number of learning outcomes which included encouraging collaboration, externalization of thought, and scaffolding. While all these play an important role in making reflection support successful, due to space limitations we cannot discuss them further at this point. A more detailed description of the iterative design process, the theoretical background of the design, and the support for externalization can be found in Kharrufa et al. (2009) Structuring the task When higher achieving students solve paper mysteries, they usually passed through three distinct stages: (1) reading the slips, (2) grouping them, and then (3) building a branched sequence that reflects their solution (Leat & Nicholas, 2000). From observing students solving paper mysteries and the first (unstructured) iteration of the digital mysteries, we found that not all groups went through such clearly delineated stages, and lower achieving groups might overlap stages (1) and (2), and even skip stage (3) altogether. We also observed that this lead to lower quality answers, and consequently, taking into consideration the recommendation of Nunes et al. (2003), we subdivided the problem into three steps, enforcing a structure to the task of solving digital mysteries into the corresponding reading, grouping, and sequencing stages. These are followed by a dedicated reflection stage (Nunes et al., 2003; Collins & Brown, 1988). At the beginning of each stage, the application displays a set of clear and simple instructions about what is required from the students, and how to proceed to the next stage.

Another motivation for the structuring of the task was also taken from the theory of distributed cognition (Perry, 2003) and its notion of presentation states. The approach used by distributed cognition (DC) in understanding interactions between people is through the study of the transformation of representation states during the process. In this regard, students are instructed in stage (2) to classify the slips in categories and in stage (3) to layout the slips in sequences, and in terms of cause-and-effect relations, to encourage them to represent the same problem in two different presentation forms, which by itself has its own cognitive benefits for the students. DC theory focuses not only on the presentation states but on the tools used in the transformation process and this was one of the main concepts that drove the design of the tools and interaction aspect of digital mysteries – this is another aspect of the design that is discussed in detail in Kharrufa et al. (2009). Moreover, classifying and sequencing data are fundamental cognitive skills that provide opportunities for reasoning and negotiation (Wright & Taverner, 2008) which themselves play part in the part-of-activity reflection support (Baker & Lund, 1997). Stage 1: Reading The first stage shows the mysteries’ slips as a number of iconified images randomly distributed on the table surface. Each slip has three display modes: a small mode that shows iconic representations of the slips’ contents (usually used when the contents are internalized by the students), a normal mode that shows the textual content clearly and an enlarged mode that is used to attract attention to a certain slip. At this stage, the application only allows for moving the slips, rotating them, and switch between these three display modes using a form of a crossing-based polar menu specifically designed for tabletop interaction (Sulaiman & Olivier, 2008). The slips are first shown in iconified mode, and when the application detects that all the slips have been enlarged at least once, it concludes that all the slips have been attended (read) and only then moves to the next stage. Stage 2: Grouping In this stage, students are provided with a grouping tool in addition to a note tool and a relation tool (Figure 2). These tools aim at making the students’ thinking visible to themselves and to an external observer in addition to creating a space for discussion, explanation, and disagreement around the act of creating and using them. The relation tool allows a Figure 2: Slip manipulation commands with a student to mark tightly related slips. When selected, a small slip showing in normal mode (left), stage two sticky-tape-like shape is created on the table. If the sticky tape commands (middle), and stage three commands is placed on two adjacent slips, they will move together making it clear that these two slips are related. The post-it note tool (right). aims to encourage students to record their thoughts for themselves and for others. Notes can be manipulated (rotated, re-sized, grouped, or sticky taped to another slip) just like normal slips. The grouping tool creates a semi-transparent, re-sizable, rectangular areas and a slip is made part of a group by dragging it inside the group’s area. From the initial trials, we noted two undesirable behavior patterns during grouping. The first of these we also observed in paper mysteries (and the first digital version) where students just piled up slips without thinking explicitly of what the piles represented. The second we observed in the second digital version in which we provided a tool that allowed the creation of named group. Using this tool, the number of groups created was observed to be closely related to the level of the students, with lower achieving groups creating as few as two groups with names that were not very descriptive (e.g. “g1” and “g2” groups in some cases) and higher achieving student groups creating as many as six groups with highly descriptive names. These two patterns of behavior of behavior were a target of the final design of stage 2. In the final design when selecting the grouping tool, the application explicitly asks for a name for the group. When the group dialog box appears, a soft keyboard is maximized, and the application stops all other interaction so as to focus all the students’ attention on the activity of creating a new group. If the group name entered is less than three letters, the application instructs the students to think more of a descriptive name for the group. The grouping stage ends when all the slips are made part of a group, However, if the students put all the slips in less than four groups, the application shows a dialog box informing the students that their grouping is not good enough (to induce reflection). The application also provides tips, in the form of a post-it note, as to other possible groups giving examples of tightly related slips that are not grouped together by the students based on meta-data associated with the mystery.

Stage 3: Webbing and sequencing The goal at this stage, which is made clear by the introductory dialog, is to arrange the slips in a branched sequence layout that reflects the reasoning of the students; showing the sequence of events and cause-and-effect relations. In addition to the normal sticky tape tool (used for marking strongly related slips), we added a directional sticky tape tool to be used for cause-effect or sequencing to make the thinking of students in terms of relations more explicitly and to open up more space for the discussion on the type of relation tape to use. This stage was the most challenging to students and a re-occurring error pattern observed in solving paper-based and the first two iterations of digital mysteries is that students end this stage with either a number of piled groups, or a simple, linear, non-branching sequence. Our design response to this problem was to ensure that when students select the finish command, they are presented with a sequence evaluation dialog. The addition of this dialog seeks to leave the task of evaluating the quality of the layout for this stage to the students themselves, rather than have the application detect the type of layout employed. This encourages the students to reflect on their work and opens a space for discussing the layout amongst themselves. This is done by displaying three images with different layouts each correspond to one of the identified patterns: a number of piles, a linear sequence of slips, and a properly branched sequence (Figure 3). Students are asked to select the layout that most resembles their own, and each student has to confirm the selection independently to prevent a single student from taking the decision alone. If the layout chosen is not the branched sequence, hints on how to improve the sequence, based on mysteries meta-data, are provided in the form of a post-it note and the application resumes in the sequencing stage. If the students select the branched sequence, stage 3 ends and the application asks the students to write down a single answer and confirm it independently.

Figure 3: Sequence evaluation dialog Reflection stage The reflection stage starts after the students agree on their final answer. This stage is designed to be conducted under the supervision of the teacher. Logging, and consequently playback of the session is designed to provide a certain level of abstraction to the task. A recording of the session is made by logging only the initial and final stages of the actions. That is, when a slip is moved from one point to another, for example, only the initial and final locations are logged and not the whole movement. This allows a very quick playback of the session showing only actions without intermediate movements or idle states (which constitute a major part of the process). This makes it possible to playback a full one hour session in 5 to 10 minutes depending on the selected playback speed.

Figure 4: The reflection dialog

The reflection stage is controlled using a reflection dialog. The reflection dialog (Figure 4) is designed to provide a simple static visualization of the whole session. The progress bar is divided into three stages, each with a length in proportion to the time taken for each stage, and it displays the time taken for each stage and of the whole session. The bar also shows bookmarks indicating critical moments in the process (Collins & Brown, 1988) where the application provided hints to the students during grouping or sequences stages. The dialog also shows four screen captures of the layout at the beginning of the process and at the end of each stage so that it is possible to quickly get a general feeling on how the students progressed. Clicking on any of the images, or the bookmarks, moves the progress bar to that point in the session so the teacher can quickly identify and facilitate discussion about key moments in the session. An important feature of this stage is that it is possible to pause the playback at any point in time and manipulate the slips as in a regular session. This allows the teacher and the students to discuss and actually explore different scenarios such as creating an additional group, or modifying a certain sequence. Being able to watch a quick playback of the whole session increases students' awareness of their problem solving strategy and with proper guidance from the teacher, students can realize their mistakes and work on improving their strategy in later sessions (Collins and Brown, 1988; Nunes et al., 2003).

The trials We built digital mysteries using a pre-production prototype of the multi-pen horizontal Promethean Activboard (Figure 1-a). The use of this board allowed the application to distinguish between the students and therefore provide participation pie charts and other features that required addressing each student individually. We performed a total of 12 trials using the final design with six groups of students (three students per group). The ages of students ranged from 11 to 14 years. We prepared four new mysteries of increasing difficulty where all six groups attempted the simplest mystery, and two of the groups went on to attempt the three remaining mysteries (in order of increasing difficulty). For these repeated trials, we chose one of the high achieving groups (as identified by their class teacher), and one of the low achieving groups to allow us to contrast the repeated use of digital mysteries by students of different abilities. The repeated trials were conducted within a two week period, with a 1-3 day period separating each trial. All the sessions were video recorded and we focused our analysis on key interactions which span only a couple of minutes and provide good evidence as to the desired outcomes. Our investigation showed strong evidence that the design lead to effective collaboration which was made clear by the amounts of discussions around the acts of creating groups and notes. Out of the 60 groups created in the 12 sessions, 51 of them lead to discussion about creating the group or naming it, and out of the 28 notes created, 21 of them lead to discussions about their contents. Figure 1-(a) shows the final layout of the higher achieving group in their second trial. The figure shows that the students built a proper branched sequence with clear distinction between the use of normal and arrow sticky tapes. It also shows the use of the post-it notes to record thoughts and making them visible to others. In the following, we present sample cases on the effect of the support of reflection in the application for each of the stages for the less achieving group that performed four trials. Grouping stage In solving a mystery with the question “Should Annie leave Windy Creek or should she stay? And why?”, the students and in their first trial of mysteries created two groups only, “stay” and “leave”, and put all the slips into them. With the placement of the last slip in one of these two groups, the application provided a hint about two sets of slips that are tightly related bur are placed in separate groups. The students took the slips mentioned in the hint out of their current groups, read them aloud, put sticky tapes on them and thought about the relationships between them. After that they created two new groups “Arizona” and “no money”

Figure 5: The layout at the end of the grouping stage.

(Figure 5). This is a clear case where the reflective intervention by the application lead the students to rethink their grouping. Moreover, For the same group in the following trial, and at the beginning of the grouping stage one of the students said: “Remember four groups not two.” This shows an example of learning from previous trials. Sequencing stage Figure 6 (left) shows a case where the group indicated the completion of the sequencing stage with two parallel linear sequences without any type of branching or relation between the two sequences. When the students where asked to identify their sequence’s layout, they picked the linear layout, and after reading the hint, they completely re-worked their layout thinking more deeply about the relations between the slips and built the layout shown the right figure.

Figure 6: Moving from a simple linear layout to a branched one based on the feedback from the sequence evaluation dialog. Reflection stage Figure 7 shows the reflection dialog for the group at the end of the session. From the reflection dialog the teacher immediately noted that the students spent very little time in the reading stage, and figure 8 shows the related discussion

Figure 7: The reflection dialog showing the short duration of stage one. Finally, during the reflection stage in the second trial of the higher achieving group, the teacher asked the students about which techniques they have used from a list of learning techniques written on the wall of the room. A student replied that they have used “thinking about thinking, because we’ve thought about how we could have thought about it.” Another student said “we’ve already starting getting strategies and build on them”. These comments from the students show a clear evidence that the application was successful in bringing the students attention to the their own thinking process and making them aware of the concepts of metacognition.

Teacher: It took you no time to get from there (points to beginning) to there (points to the end of the reading stage). Student 1: Is that good? Teacher: You are going to answer that question in a second yourselves. Student 2: Because we didn’t really read them. Teacher: Didn’t really read them? The teacher explained to them that they should have read them together and how to do it next time. Same group in the next trial, the students are about to start: Student 1: Alright let’s start reading them Student 2: One at a time, one at a time. During the reflection stage: Teacher: What did you do differently from yesterday? Student 1: We did better than yesterday Student 2: We read one at a time Student1: And we put it aside Teacher: What were the effects of doing that? was it better, was it worse? (all said better) Student 2: Because we all knew what was going on. Teacher: So did it help you when you got to your grouping stage? Student 1: Yes, we were all working as a team

Figure 8: Discussions during the reflection stage and their effect in subsequent trials

Discussion and Future Work The previous examples of interaction from our trial demonstrate that the design of the application was successful in inducing reflective interactions at different stages during the process and that it lead to improved performance with repetitive use. While the choice of the appropriate collaborative problem solving task, structuring it, including reflective hints at critical moments, and following this with the dedicated reflection stage, played an important role in making the application successful, we believe that the other design aspects of encouraging effective collaboration and encouraging students to make their thinking visible (externalization) played an equally important part. The ability to combine the features of face-to-face collaboration and providing computer-based support that allows the structuring of the task and appropriate interaction were only possible because of the unique affordances of digital tabletop technology. An important aspect in the design of reflection in digital mysteries, is the minimum amount of supervision required from the teacher. Unlike the case of paper mysteries where the teacher needs to closely monitor the session to be able to assess the performance of the students, and if a post-activity reflection is required, which involves a lengthy video playback process, digital mysteries only requires minimal teacher intervention at the reflection stage which in a matter of 5 to 10 minutes provides the teacher with a quick overview of the session allowing him both to assess the performance of the student and provide post-activity reflection support. In future work we hope to introduce a simple planning stage at the beginning of the task and at the beginning of each stage. This, in combination with the reflection support, should encourage students to be more aware of their strategies in solving the problem by being able to clearly compare between their selection of possible strategies and their outcomes. We are also are planning to follow the problem solving session with a story writing session where students are asked to compose a story based on the sequence they have built. This has the advantage of supporting the story telling task with a narrative that the students have build during the task, and on the other hand, when the students know that they will have to compose a story based on their solution, this should serve to motivate them to work on better solutions for the problem

References Boyle, T. (1997). Design for Multimedia Learning. Prentice Hall. Baker, M., & Lund, K. (1997). Promoting reflective interactions in a CSCL environment. Journal of Computer Assisted Learning, 13, 175--193. Collins, A. and Brown, J. S. 1988. The computer as a tool for learning through reflection. In Learning Issues For intelligent Tutoring Systems, H. Mandl and A. Lesgold, Eds. Springer-Verlag New York, New York, NY, 1-18

Dillenbourg, P. (1999). What do you mean by collaborative learning? In P. Dillenbourg (Ed) Collaborativelearning: Cognitive and Computational Approaches, chapter 1, pages 1-19. Elsevier, Oxford. Do-Lenh, S., Kaplan, F., and Dillenbourg, P. Paper-based Concept Map: the Effects of Tabletop on an Expressive Collaborative Learning Task. In The 23rd BCS conference on Human Computer Interaction (HCI 2009), pages 149-158, Cambridge, UK. ACM. Hilliges, O., Terrenghi, L., Boring, S., Kim, D., Richter, H., and Butz, A. (2007). Designing for collaborative creative problem solving. In C&C '07: Proceedings of the 6th ACM SIGCHI conference on Creativity & cognition, pages 137{146, New York, NY, USA. ACM. 10 Kharrufa, A.Sulaiman, Olivier, P., Leat, D. (2009). Digital Mysteries: Designing for Learning at the Tabletop. Technical report, CS-TR No 1171, School of Computing Science, Newcastle University, Sep 2009. Leat & Nicholas, 2000, D. and Nicholas, A. (2000). Brains on the table: diagnostic and formative assessment through observation. Assessment in Education: Principles, Policy and Practice, 7(1):103-121. Lehtinen, E. (2003). Computer-supported collaborative learning: An approach to powerful learning environments. Powerful learning environments: Unravelling basic components and dimensions, pages 35-53. Morgan, M. and Butler, M. (2009). Considering multi-touch display technology for collaboration in the classroom. In Proceedings of World Conference on Educational Multimedia, Hypermedia and Telecommunications 2009, pages 674{683, Honolulu, HI, USA. AACE. Nunes, C. A. A., Nunes, M. M. R., and Davis, C. (2003). Assessing the inaccessible: metacognition and attitudes. Assessment in Education: Principles, Policy and Practice, 10(3):375-388. 11 Perry, M. HCI Models, Theories, and Frameworks, chapter Chapter 8: Distributed cognition. Morgan Kaufmann, 2003. Piper, A. M. and Hollan, J. D. (2009). Tabletop displays for small group study: affordances of paper and digital materials. In CHI '09: Proceedings of the 27th international conference on Human factors in computing systems, pages 1227-1236, New York, NY, USA. ACM. Piper, A. M., O'Brien, E., Morris, M. R., and Winograd, T. (2006). Sides: a cooperative tabletop computer game for social skills development. In CSCW '06: Proceedings of the 2006 20th anniversary conference on Computer supported cooperative work, pages 1-10, New York, NY, USA. ACM Press. Resta, P. and Laferriere, T. (2007). Technology in support of collaborative learning. Educational Psychology Review, 19(1):65-83. Rick, J. and Rogers, Y. (2008). From digiquilt to digitile: Adapting educational technology to a multi-touch table. In Horizontal Interactive Human Computer Systems, 2008. TABLETOP 2008. 3rd IEEE International Workshop, pages 73-80. Sulaiman, A. N. and Olivier, P. 2008. Attribute gates. In Proceedings of the 21st Annual ACM Symposium on User interface Software and Technology (Monterey, CA, USA, October 19 - 22, 2008). UIST '08. ACM, New York, NY, 57-66. Wellner, P. (1993). Interacting with paper on the digitaldesk. Commun. ACM, 36(7):87-96. Wright, D. and Taverner, S. (2008). Thinking Through Mathematics. Chris Kington Publishing.