This paper reports work regarding the design, development and evaluation of a surface computing application to support collaborative decision making.. The domainindependent application, so called Ideas Mapping, builds on the principle of Affinity Diagramming to allow participants to analyze a problem and brainstorm around possible solutions while they actively construct a consensus artifact - a taxonomy of their ideas. During idea generation, Ideas Mapping replicates physical post-it notes on a multi-touch tabletop. Additional functionality supports student collaboration and interaction around the organization of ideas into thematic categories associated with the problem at hand. We report on the functionality and user experience while interacting with the application which was designed and developed using a user-centered approach. We also report initial findings regarding the affordances of surface computing for collaborative decision making.
INTRODUCTION
A multi-touch interactive tabletop can support
collaboration, allowing different patterns of turn taking,
negotiation and interaction [
EXISTING KNOWLEDGE The work reported in this paper draws from literature in the areas of Human Computer Interaction (HCI) methods (Affinity Diagramming) and Multi-touch interactive surfaces. The current state of the art in these areas is briefly summarized in this section.
Affinity Diagramming
HCI techniques exist to facilitate discussion in groups and
to extract ideas from users’ initial conceptual models. For
example, the Kawakita Jiro diagrammatic method [
Multi-touch Interactive Tabletops
Multi-touch interactive tabletops have recently attracted the
attention of the HCI and Computer Supported Collaborative
Learning (CSCL) communities. Based on preliminary
evidence from the education and computer-science
literature, Higgins et al. [
Multi-touch tabletops have been used with disabled user
groups to promote development of social skills. SIDES, for
example, is a four-player cooperative computer game
designed to support adolescents with Asperger’s syndrome
to practice social skills and effective group work during
their group therapy sessions [
With regards to using tabletops in formal learning settings,
a series of studies are currently being conducted as part of
the SynergyNet project [
DESIGN METHODOLOGY We adopted a strongly user-centered approach, emphasizing the engagement of students and instructors in all phases of the design process. Four university students and three instructors were involved, contributing to design elements of the application.
First, through low-fidelity paper-based prototypes, we simulated a collaborative activity with four students around a (turned-off) tabletop using paper and pencil. The scenario involved “the creation of a computer games industry in Cyprus and the factors involved.” First, students generated ideas individually for 10 minutes. They wrote a (physical) post-it note for each new idea. Next, the ideas appeared one-by-one on the table and became subject to discussion, after a brief explanation from their originator, in an effort to categorize them in thematic units. Students revisited and changed ideas, rejected less promising ones, and generated new ideas during a collaborative decision making process leading to their thematic categorization. Finally, the activity concluded with a consensus of the main factors (i.e., resulting thematic categories) involved in the creation of a computer games industry in Cyprus. After the completion of the activity, instructors (who observed and kept records of all interactions during the activity) and students discussed the potential surface computing application and contributed to elements of the design from their own viewpoints.
Following the low-fidelity design discussions and analysis of user needs, a prototype Beta version application was developed in Action Script 3.0, for a multi-touch tabletop, the MagixTable. The application, so called Ideas Mapping, was designed to be domainindependent with a mild learnability curve. Our participants were called back to collaborate on different scenarios using Ideas Mapping and provide feedback on its user experience and further suggestions for improvement. Evaluation sessions took place in a fully equipped usability lab and all sessions were video recorded and analyzed. Ideas Mapping was optimized and finalized in three major iterative cycles of design, development and evaluation. OVERVIEW OF THE APPLICATION Overall, Ideas Mapping is designed to support idea generation, collaborative decision making and group artifact construction. The application builds on the principle of Affinity Diagramming to allow participants to analyze a problem and brainstorm around possible solutions while they actively construct a consensus artifact; namely, a taxonomy of their ideas. This is done in three stages: Stage 1: With a scenario at hand, each collaborator generates new ideas. Ideas are typed into a web application (producing an XML file associated with Ideas Mapping) through the use of a mobile device (laptop, tablet, smartphone connected to the Internet). The need for the integration of mobile devices and a web application emerged from a constraint imposed by the MagixTable (also true for other platforms such as the MS Surface) -that text entry can be done from one pre-existing keyboard at a time. For the kind of activity we sought, this constraint would be significant. To resolve this problem, we developed four virtual keyboards on the tabletop (one for each user). However, users experienced difficulties typing extended ideas on the virtual keyboard during stage 1; the keyboard interaction suffered from input latency and mistyping issues. Thus, the use of mobile devices for input via a web application was considered as a practical solution to this problem for stage 1. This problem demonstrates both the still existing technical limitations of tabletops but also the importance of user input in developing applications for such technologies. educational aspects.” The session was video recorded and analyzed.
Stage 2: Next, the ideas are presented one-by-one, as digital
post-it notes in the middle of the tabletop surface and
become subject to discussion amongst the collaborators. For
each idea, collaborators make an effort to categorize it in a
thematic unit. Functionalities include:
• Each post-it note must be categorized before the next
one appears. If controversy exists, an idea can be
placed in the “Decide Later” depository to be revisited
upon the categorization of other ideas. Post-it notes are
automatically oriented to face their contributor, which
encourages them to elaborate on the idea. This
functionality was implemented as a result of users’
feedback and is consistent with previous work by [
Stage 3: In this last stage, more flexibility is given to the participants to finalize their taxonomy. In addition to the above, users can now edit ideas or generate new ones, delete ideas or thematic units that are less promising, and reallocate ideas into thematic units for a better fit. Overall, students engage in a collaborative decision making process, leading to the construction of a group artifact -- a taxonomy of their ideas.
STUDIES WITH IDEAS MAPPING To examine the affordances of surface computing for collaborative decision making two studies were conducted with groups of university students: a small pilot study and a larger scale investigation.
THE PILOT STUDY Participants and Setting: Four university students, aged between 22-27 years old, were recruited to participate in a short activity around the tabletop. The scenario involved the “creation of an action plan that can improve university students’ experiences at the Cyprus University of Technology, including social and
An exploratory approach was used to trace the kinds of interactions amongst the collaborators and the technology and to better understand the role of tabletops in supporting learning. General research questions guided our video analysis such as: what kinds of interactions take place around the tabletop? and what evidence is present regarding the value of multitouch interactive tabletops for collaborative decision making? One of the researchers considered the video corpus in its entirety – a total of 57 minutes. Most interaction occurred during the 2nd and 3rd stages of Ideas Mapping, which became the focus of the analysis. The researcher repeatedly watched the video, marked segments of interest, and created transcripts, in an effort to categorize the types of discourse and gestures used by the group members around the tabletop. A preliminary coding scheme is presented in Table 1. This coding scheme will be further refined as more studies are conducted in this context. Understanding collaborative decision making around tabletops is currently limited. It is thus important to establish a coding scheme of the interactions evident around this technology (particularly, the synergetic dialog and physical gestures) to be able to examine the phenomenon further. Ultimately, the coding scheme should help us examine interesting patterns of collaborative decision making around multitouch interactive tabletops.
Spoken Contributions • Information Sharing – Defining/describing/identifying the problem • Proposing – Proposing a thematic unit/new idea • Elaborating – Building on previous statements, Clarifying • Negotiating meaning – Evaluation of proposal, Questioning/ answering, Expressing agreement/disagreement, Providing arguments for/against • Stating consensus – Summarizing ideas, Metacognitive reflections • Other talk – Tool-related talk, Social talk, Laughter Gesture Contributions • Communicative Gestures – Show on the table without touching, Dominating/blocking gestures • Touch Gestures – Resize, Rotate, Type, Move something across, Random touching or touching to explore Overall, the pilot study provided initial evidence that the CSCL setting encouraged and stimulated discussion and physical interaction around shared artifacts.
LARGER SCALE INVESTIGATION Participants To further examine the value of multitouch interactive tables for collaborative decision making, we recruited postgraduate students in Cyprus to discuss a scenario related to peace. The sample was composed of 17 postgraduate students enrolled in a CSCL/CSCW course at a public university in Cyprus, aged between 22-45 years old (M=30).
The participants were divided into five groups: 3 groups of 3 students and 2 groups of 4 students, suitable for the foursided tabletop. Group members were familiar with working together through other course learning activities. All, but one student, had no prior experience with using a multitouch tabletop. Procedures In this study there was a preparatory phase before students engaged in group work around the tabletop. That is, Stage 1 of Ideas Mapping was completed in distance, during the week before the tabletop investigation. The preparatory week aimed to allow students to research the scenario and think at their own pace. During the preparation week, students were tasked to investigate the topic, think creatively and record at least 10 ideas into the Ideas Mapping web application. The following scenario was presented to the students: “Your team works at a non-governmental organization dealing with global peace. Your project is to create a roadmap of actions to promote global peace using technology.” The specific scenario was chosen for it to be thought-provoking and without obvious answers to it. The goal was to stimulate critical thinking, dialog, and creative problem solving. Cyprus is a country in a long lasting political conflict. Thus, the topic was both personally important for the student participants, but also required their emotional and mental engagement. The next phase involved collocated collaboration around the tabletop. Following the prep week, each group met faceto-face and engaged in collaborative work as described in Stages 2 and 3 of Ideas Mapping. Briefly, the ideas of each group were presented on the tabletop one-by-one. Students engaged in discussion and physical interaction with the tabletop in an effort to categorize the ideas in thematic units (i.e., taxonomy of ideas).
Data collection The sessions of all five groups were video recorded for subsequent utterance coding and analysis. To complement the video data, a questionnaire was administered to all participants soon after the completion of the activity. The questionnaire aimed to assess students’ perceptions of the collaborative learning experience and the usability of the surface computing application.
Video Analysis An extensive video analysis of the data was carried out but due to the scope and space limitations of this paper the results will be presented elsewhere.
The questionnaire included 30 Likert-type items with a 7point agreement response scale (from 1: completely disagree to 7: completely agree). These items measured three constructs of interest: (1) Collaboration Support, assessing the extent to which students thought the technology supported their collaboration such as, “The technology helped me work effectively in my group”, “The technology met my needs as a collaborator”; (2) Learning Experience, assessing the extent to which students were satisfied with their learning experience overall, such as “Overall, my collaborative learning experience was positive”, “I am satisfied with my experience through this activity”, and (3) Usability Satisfaction (adapted from Lewis, 1995), assessing the extent to which students were satisfied with the usability of the system such as, “It was simple to use this system”, “I can effectively complete my work using this system”, “I like using the interface of this system”. A total of 17 students completed the questionnaire. First, the internal consistency for each subscale was assessed using Cronbach’s alpha; all 3 subscales had acceptable internal consistency (Cronbach’s alphas > .80). Then, subscale mean scores were calculated for every participant (i.e., an un-weighted composite score for each participant on each subscale) followed by computation of descriptive statistics. As shown in Table 2, means were well above the midpoint of the 7-point response scale for all three measures, suggesting that the technology was positively endorsed by the participants overall. Specifically, the participants thought the technology supported their collaboration (M=5.53, SD= .22), and were satisfied with their learning experience (M=5.77, SD= .51). With regards to the third measure, participants found the system usable overall (M=4.93, SD= .77), but individual item means pointed to some aspects which may need improvement. The rating average was lower for three particular items in this scale, suggesting that we should improve the way participants recover from mistakes (“The system gives error messages that clearly tell me how to fix problems” M=3:00 and “Whenever I make a mistake using the system, I recover easily and quickly” M=3.36), as well as extend the application to include more functionality (“This system has all the functions and capabilities I expect it to have” M=3.88).
The questionnaire also included an open-ended question concerning the pros and cons of using tabletops for collaborative learning activities. We reviewed students’ responses to identify themes. Several students commented on how the tabletop promoted collaboration, helped them maintain attention to the task and was enjoyable to use. For example, one of the participants commented: “The tabletop helped us collaborate and the resulting product was a group effort. It helps you pay attention. I also found it very enjoyable”. Often, students pointed out the capabilities of the system that enabled effective collaboration, such as “It was nice all of us could use the tools at the same time, to rotate a note, to make it larger to read, or to put it in the box to revisit later.” On the negative side, a few participants found the virtual keyboard difficult to use and that the system needed improvement in handling mistakes, which was consistent with the findings from the quantitative data. These results confirmed our views regarding the affordances of multi-touch tabletops to support collaboration activities and also contributed to further refinement of Ideas Mapping.
# Cronbach’s Subscale M (SD)
Items Alpha 1. Collaboration Support 6 .94 5.53 (.22) 2. Learning Experience 5 .96 5.77 (.51) 3. Usability Satisfaction 19 .97 4.93 (.77) Table 2: Subscales statistics and descriptive statistics (N=17) DISCUSSION - CONCLUSION This study reports on the functionality and user experience while interacting with a multitouch application which was designed and developed using a user-centered approach. We also report initial findings regarding the affordances of surface computing for collaborative decision making. Ideas Mapping builds on the principle of Affinity Diagramming to allow participants to analyze a problem and brainstorm around possible solutions while they actively construct a consensus artifact -- a taxonomy of their ideas. We feel Ideas Mapping makes the Affinity Diagramming technique more collaborative. By allowing for an extension sorting activity, it provides a way for participants to negotiate around an emerging group artifact and make sense of challenging problems, such as how to promote world peace using technology.
We further have evidence that the CSCL setting of the study, and surface computing more generally, encouraged and stimulated dialog and collaborative work around an authentic problem . Following the individual generation of ideas, Ideas Mapping supported a 2-stage collaborative activity that promoted ideas sharing, negotiating, sorting and constructing a group artifact while coming to a consensus.
Moreover, we believe that traditional user experience evaluation methods (e.g. questionnaires) were useful for evaluating Ideas Mapping. However qualitative evaluation (e.g. video analysis and the establishment of a coding scheme) is also important; such methods can reveal interesting patterns of interactions amongst the participants and with the technology beyond what is self-reported. Below, we identify some implications of this work for future research and practice in the fields of HCI and CSCL.
1. 2.
Designers should focus on engaging students and instructors in the design process of educational surfaces computing applications.
Current interactive tabletop technologies come with a lot of user interface limitations. These should be taken into account when designing applications for such surfaces.
The CSCL setting of the study encouraged and stimulated active dialogue with a problem at hand and a multitouch interactive tabletop application to support them. 4. Self-reported measures showed that students positively endorsed the use of multitouch interactive tabletops for small group work.
The proposed coding scheme can be applied and extended to more studies in the area.
New qualitative analysis methodologies evaluating user experience are needed. for The role of surface computing in promoting dialogue around sensitive topics (like peace) is an interesting area for further research.
A framework for using surface computing for collaborative decision making in general (especially related to sensitive issues) can be developed and tested.
This project is funded under the Cyprus Research Promotion Foundation’s Framework Programme for Research, Technological Development and Innovation 2009-2010 (DESMI 2009-2010), co-funded by the Republic of Cyprus and the European Regional Development Fund, and specifically under "Bilateral Cooperation" (ΔΙΑΚΡΑΤΙΚΕΣ/ΚΥ-ΣΛΟ/0411) 6.Hatch, A., Higgins, S., Joyce-Gibbons, A. & Mercier, E. (2011) NumberNet: Using Multi-touch Technology to Support Within and Between Group Mathematics Learning. In Proceedings of the 9th International Conference on Computer Supported Collaborative Learning (pp. 176-183). International Society of the Learning Sciences.