Icce 2009 publications format

T. Hirashima et al. (Eds.) (2011). Proceedings of the 19th International Conference on Computers in Education. Chiang Mai, Thailand: Asia-Pacific Society for Computers in Education Collaboration by Choice: Youth Online
Creative Collabs in Scratch
Yasmin KAFAIa, Ricarose ROQUEb, Deborah FIELDSa, Andres
MONROY-HERNANDEZb
aGraduate School of Education, University of Pennsylvania, United States bMedia Lab, Massachusetts Institute of Technology, United States Abstract: Online creative production has received considerable attention for its success in
creating Wikipedia and Free and Open Source Software yet few youth participate in such
voluntary online collaborations, in particular in programming contexts. In this paper, we
describe how youth programmers organized collaborative groups or collabs in response to a
design challenge in the Scratch Online Community. We report on participation in the
“Collab Challenge” in the Scratch community at large and with particular groups, designers’
efforts in recruiting and organizing collab groups, and the role of community feedback. In
the discussion, we address what we learned about youth’s informal collaborative skills,
fostering community participation, and the design of online communities supportive of
creative collaboration, and open issues for further research.
Keywords: Collaboration, online community, programming

1. Introduction

Observations of social network and gaming communities suggest that collaboration among
hundreds, if not thousands, of members can be productive contexts for learning [4, 5]. A
number of studies have shed light on the nature and dynamics of online creative
collaboration, examining knowledge communities such as Wikipedia [2] or smaller
collaboratives or collabs for producing digital media [8]. However, the literature on how to
organize voluntary learning-centric online creative collaborations is limited. As a starting
point for our investigation, we turned to the Scratch Online Community where we organized
an open, collaborative design challenge, the Collab Challenge, with the goals of stimulating
youth to collaborate on programming projects and studying their collaboration in
self-organized groups. With over 850,000 users primarily between 11 to 17 years old and 2
million projects, Scratch is by far the largest youth online programming community [9]. In
this paper we report on participation in the Collab Challenge and youth programmers’
efforts in forming and organizing groups. We discuss the implications for design of what we
learned from observing hundreds of participants, the roles that participants assume, and the
performances or artifacts that result from such collaborations.
2. Background

Most of the existing literature on collaborative learning has examined the factors and
arrangements of structured collaborations [3] but we know little about what youth would
need to learn in order to collaborate effectively in such situations when the choice of
T. Hirashima et al. (Eds.) (2011). Proceedings of the 19th International Conference on Computers in Education. Chiang Mai, Thailand: Asia-Pacific Society for Computers in Education collaboration, partners and topic lies on the students [7]. Two recent studies on online
creative collaborations among adults [8] and youth [1] found that collabs were successful
when they had leaders that assumed collective responsibility [10] for coordinating work and
had a high degree of communication among members. The goal of our study was to better
understand the types of self-organized collaboration in the Scratch community. By looking
at the breadth of participation in the Challenge, we hope to map out this new territory of
collaboration by choice. This paper addresses the following research question: What were
collaborative styles of groups participating in the challenge?
3. Context

To observe and encourage creative collaborations in the Scratch Online Community [9], we
designed the Collab Challenge, which we announced on the Scratch website in January
2011. We issued a call for Scratch users to form teams or collabs and create a Scratch project
using three pre-defined images as a constraint. Halfway through the Challenge, participants
must submit a project draft for the Scratch Team to review and provide feedback. At the end
of the Challenge, members of the Scratch Team reviewed the final projects based on the
following criteria: originality of the project; creativity of the art, music, and animation; and
elegance and sophistication of the programming code.
We collected data on participants’ length of time in the community and self-reported gender, age, and location. We also collected multiple versions of projects, project comments,
relevant discussions in the online forum, and statistics about the projects that included
number of views, “love-its” (a measure of how much people like the project), and remixes.
During the review process, we asked questions about their team formation and collaborative
process. We also analyzed participants’ overall Scratch website activity, including their
entire history of projects and social interaction. In addition, six groups participated in a local,
face to face Scratch workshop hosted by some of the authors, and these were observed
closely.
4. Findings
A total of 137 collabs (341 participants) registered to participate in the Collab Challenge in
January 2011. Of these, 52 collabs with 139 participants (groups with 3-7 members, an
average of 2.6 members) submitted at least a first draft of their project for review.
Twenty-five of those collabs submitted a final version. Of the 139 participants listed in the
registration, 125 users had Scratch accounts, 34 of them self-reported as female (27%) and
mean self-reported age was 17 (std=9.6), median age at 15, and mode of 13, compared to the
average age of 21 (std=17), median age of 17, and mode of 14 of the rest of the community.
Their projects came from a variety of genres from games to stories. The Collab Challenge
seemed to attract new members to the Scratch Online Community. There were 22
participants in the challenge (17.6%) who joined the online community either within a week
of the Challenge’s announcement or during the Challenge. It appears these users joined
solely for the purpose of participating in the challenge, either recruited by friends who
already participated in the Scratch community or through local clubs and workshops that
were focusing on the Challenge.
The Challenge also appears to have re-engaged more experienced, long-time members of the online community. Sixty-seven (53.6%) of the Challenge participants had been community members for more than three months, and most of these, 62 (49.6%), had been on for six months or more. Comments made by some of these more experienced T. Hirashima et al. (Eds.) (2011). Proceedings of the 19th International Conference on Computers in Education. Chiang Mai, Thailand: Asia-Pacific Society for Computers in Education Scratch users suggest that the Challenge provided an opportunity for them to participate anew in the online community. One experienced Scratch user said that the Challenge pushed him/her to go deeper into Scratch: Yeah, this is the first time I saw the Scratch Team create a contest so I went all out. Usually, I don’t have to put too much thinking into my projects – the scripts are already in my head and I just code it up in Scratch. However, this project really pushed me. And it’s the only collab project I’ve ever finished. Of the 52 collabs that submitted a project, we know that at least 15 collabs were face-to-face interactions consisting either of family members or friends. Other collabs found each other online, either from previous relationships or from recruiting on the forum. At least five groups (23 participants) found each other on the forum and submitted projects. We present here three of these successful collabs and their organizational characteristic. 4.1.1 A Benevolent Dictatorship TheWizard was for two years one of the most prominent members of the Scratch community who had announced he was going to scale down his participation on the website until the release of the new Scratch version. However, a few days after the Collab Challenge was announced, TheWizard came back and posted an invitation for people to join “TheWizard's Coolio Collab” where he described that he got inspired by a “truly good game idea,” a mix of an RPG and a fighting game. TheWizard enticed potential collaborators with fame by association, suggesting that teammates would gain popularity because most of his projects “get to the top ranked lists” on the front page. Once he approved applicants to join the group, including two relatively new members (< 3 months on the Scratch site) he gave them several options of different tasks they could work on. Five people formed the group and they went through more than 30 versions before finishing the game. Their collaborative project was well received by the community at large and it was arguably one of the most sophisticated projects submitted to the challenge. While every member of the collab contributed in visible ways, TheWizard carefully orchestrated the whole effort. Almost half of the comments on the “Coolio Collab” forum thread were posted by TheWizard. Overall, the group's organizational model was primarily centralized and dependent on a strong leader but members seemed to have enjoyed participating in it as they were most likely honored to be part of TheWizard's team. 4.1.2 A Team Effort Like TheWizard’s Coolio Collab, “The Angelic Collab” formed in the website discussion forums initiated by another experienced Scratch user Archangel. Its seven members were spread across three countries. While Archangel facilitated the group collaboration, ideas and decisions were negotiated by the group based on the feasibility and likelihood of advancing the team in the Challenge. Occasionally, a member would summarize the ideas and ask members to vote and reach a consensus to move forward. To develop the project, members split themselves up based on their interests and skills into graphic artists and programmers. Programmers would remix each other’s projects to add their update to the ongoing development, while graphic artists would share their assets such as images or animations through Scratch projects that programmers would later integrate. They coordinated their exchanges and remixing through the website discussion forums, regularly summarizing their efforts and tasks to keep everyone in sync. In the end, they produced one of the T. Hirashima et al. (Eds.) (2011). Proceedings of the 19th International Conference on Computers in Education. Chiang Mai, Thailand: Asia-Pacific Society for Computers in Education Challenge projects that received the most attention from the community. The collaborative
processes of the Angelic Collab demonstrated how one collab achieved success through a
shared leadership model that spread the responsibilities of organization, decision-making,
and development across its members.

4.1.3 A Friendly Partnership
When the Collab Challenge was announced, experienced Scratch user Sunday from the
United Kingdom asked her Scratch friend fashionista519, who was from the United States
and who she “met” 5 months before in the Scratch Online Community, to collaborate on a
project. When they began exchanging ideas for the Collab Challenge, they both expressed a
mutual interest in making a 3D game and converged on a storyline involving a Samurai
Warrior. Fashionista519 worked on developing graphics for the game while Sunday led the
programming. Both showed equal dedication to their project development, meeting often
online and sharing the responsibilities of the project making. To develop their project, they
took turns adding code and media assets to the main project, which the other would later
remix. Unlike “TheWizard’s Coolio Collab” and “The Angelic Collab”, they communicated
and coordinated through their project comments and in a Scratch website gallery that held
all their project versions. Whenever they spoke about their project progress, they also used
that time to converse about their lives and their other interests. While both were excited at
the prospect of having their project featured, working together on their project became a
social activity for the two friends.
5. Discussion
The observations in three collabs we highlighted in the findings indicated that members
from successful collabs exhibited qualities valuable in collaboration such as clear
communication, solid leadership, and social skills to develop relationships – qualities that
also emerged in previous studies of collabs [1, 8]. These findings also suggest that
participating youth seem to have quite a repertoire of informal collaborative skills to handle
these types of opportunistic collaboration. Analysis of failed collabs (those that did not
create any projects) and the reasons behind their failure could also provide deeper insight.
Beyond strategies in how to best use the Scratch website to work together, these collabs may
have also needed support in how to collaborate effectively. The design of online
communities to support creative collaboration must not only consider the emergent needs of
their creators but also the qualities that creators must cultivate to collaborate successfully.
Ultimately, the Collab Challenge became not only a context to initiate and study collaboration by choice but also a way to engage Scratch members more deeply in their community: oldtimers came back to join collaborations while newcomers joined to do the Challenge. Some groups had members of mixed experience, allowing peer mentorship [6]. Having Challenge projects accessible for the community to view in one gallery, and especially featured on the home page, brought not only attention (views and love-its) but also many comments from the Scratch Team and broader community members, facilitating conversations amongst Scratch users at large. This finding might explain the larger appeal that online creative collaborations have for participants. Amidst all the academic benefits of collaborative work that have dominated research and practice discussions for so long, it points to a motivating dimension of collaboration that has been neglected in research. On a surface level, the presence of audience for work might provide a simple answer, but perhaps on a more profound level the striving for affinity might be a better explanation [4] on why T. Hirashima et al. (Eds.) (2011). Proceedings of the 19th International Conference on Computers in Education. Chiang Mai, Thailand: Asia-Pacific Society for Computers in Education participants are willing to contribute, share their work, and help others. These are aspects
that deserve further investigation.
Acknowledgements
The writing of this paper was supported by a grant from the National Science Foundation
(NSF-CDI-1027736) to Mitchel Resnick, Yasmin Kafai and Yochai Benkler. The views
expressed are those of the authors and do not necessarily represent the views of the
Foundation or the University of Pennsylvania or MIT.
References

[1] Aragon, C., Poon, S., Monroy-Hernandez, A. & Aragon, D. (2009). A tale of two online communities: Fostering Collaboration and Creativity in Scientists and Children. In Proceedings of the Creativity and Cognition Conference, Berkeley, CA. New York: ACM Press. [2] Benkler, Y. (2006). The Wealth of Networks: How Social Production Transforms Markets and Freedom. New Haven and London: Yale University Press. [3] Cohen, E. G. (1994). Restructuring the classroom: Conditions for productive small groups. Review of [4] Gee, J. P. (2003). What video games have to teach us about learning and literacy. New York: Palgrave [5] Greenhow, C., Robelia, E., & Hughes, J. (2009). Web 2.0 and Classroom Research: What path should we take now? Educational Researcher, 38 (4), 246-259 [6] Kafai, Y. B., Fields, D. A. & Burke, W. Q. (2010). Entering the clubhouse: Case studies of young programmers joining the online Scratch communities. Journal of Organizational and End-User Computing, 22(2), 21-35. [7] Kafai, Y. B. & Peppler, K. A. (2011). Beyond Small Groups: New Opportunities for Research in Computer-Supported Collective Learning. In Proceedings of the 9th Conference of Computer Supported Collaborative Learning. Hong Kong, China: International Society of the Learning Sciences. [8] Luther, K., Caine, K., Ziegler, K. & Bruckman, A. (2010). Why It Works (When It Works): Success Factors in Online Creative Collaboration. In GROUP ’10: Proceedings of the ACM Conference on Supporting Group Work (pp. 1–10). New York: ACM Press. [9] Resnick, M., Maloney, J., Monroy-Hernandez, A., Rusk, N., Eastmond, E., Brennan, K., Millner, A., Rosenbaum, E., Silver, J., Silverman, B., & Kafai, Y. B. (2009). Scratch: Programming for all. Communication of the ACM, 52(11), 60-67. [10] Zhang, J., Scardamalia, M., Reeve, R., & Messina, R. (2009). Designs for Collective Cognitive Responsibility in Knowledge-Building Communities. The Journal of the Learning Sciences, 18(1), 7-44.

Source: http://www.apsce.net:8080/icce2011/program/Proceedings/pdf/C2_S1_19S.pdf