Wikis, known as collaborative knowledge building environments, are being used to facilitate computer mediated collaboration among undergraduate students working in project-based settings. As educational technology becomes an integral part of the instructional engineering landscape and distance education becomes accepted practice, understanding how students learn engineering becomes intertwined with understanding how computer mediated communication helps and hinders learning.
The Pedagogical Wiki project is creating a novel framework for assessing and scaffolding collaborative learning within Wiki environments. Research being undertaken includes (a) studying student adoption and interaction in new and ongoing Wiki-based engineering courses, and comparing Wiki adoption in non-engineering (education) courses;(b) developing instructional assessment tools based on discourse analysis and course topic ontology for qualitatively evaluating student Wiki interactions; and (c) identifying scaffolding opportunities, such as topic-based material sharing, to promote student engagement and communication. Activity Theory is being used as a framework to study Wiki adoption.
Ultimately, a key component of this work lies in the new discourse and topic-based instruments that enable instructors and educational researchers to better assess student learning within Wiki environments. Qualitative as well as quantitative metrics are being developed, some with respect to the particular domains being studied. The combined use of new natural language processing techniques and traditional instruments to study Wiki adoption in an engineering context contributes to an understanding of how students learn engineering in a collaborative, "real-world" medium. Codifying best practices for computer mediated collaboration is impacting the way engineering is being taught.
The "Pedagogical Wikis" project began as a study of undergraduate student collaboration in a joint sophomore and junior year computer science course programming assignment. Approximately 10 teams of 12-15 students each were studied in each of the two years. Although the students initially used Wikis to collaborate, data collection immediately expanded to encompass data from a wide range of sources favored by students including Google Docs and Google Groups; Moodle Docs, Wikis and Discussions; and SVN, a code version control system. Activities were analyzed using NLP, Data Mining and Machine Learning approaches based on the structure of the course assignment, documents produced by the team, and team interactions. Results were displayed by team and within teams, and made available to team members, team managers and instructors. The research produced, for the first time, a real-time activity profile that allowed instructors insight into team dynamics and team progress during the eight weeks of the project. The analysis identified the type of document students were working on - e.g., research, meeting, design, coding - when it was edited and by whom. Team managers from an upper level computer science course especially like the SVN analytics. Summative analysis linked associated project grades (outcomes) to activity patterns of team production and interaction, allowing researchers to infer best practices for team success. The creation of a robust Machine Learning pipeline based on project expectations was specific to the course assignment. Other aspects of the research, such as the SVN analysis, was not domain specific and is more easily replicable. Project research produced the following published papers and posters, presented at conferences and workshops. Notably, the principle authors of three of these were graduate students. Sen Lui, Jihie Kim, Sofus A. Macskassy, Erin Shaw, Predicting Group Programming Project Performance using SVN Activity Traces. Educational Data Mining. Memphis, Tennessee, 2013 Jihie Kim, Erin Shaw, Hao Xu, Adarsh G V, Assisting Instructional Assessment of Undergraduate Collaborative Wiki and SVN Activities, Proceedings of the International Conference on Educational Data Mining, 2012. Chitra Ganapathy and Jihie Kim, Analysis of Student Interaction aspects of Collaborative Programming Projects using SVN, Proceedings of the ITS workshop on Intelligent Support for Learning in Groups , 2012. Chitrabharathi Ganapathy, Erin Shaw and Jihie Kim, Assessing Collaborative Undergraduate Student Wikis and SVN with Technology-based Instrumentation: Relating Participation Patterns to Learning, Proceedings of the American Society of Engineering Education Conference, 2011. Chitrabharathi Ganapathy, Jeon-Hyung Kang, Erin Shaw and Jihie Kim, Classification Techniques for Assessing Student Collaboration in Shared Wiki Spaces , Proceedings of the AI in Education Conference (poster), 2011.
