Group Interactivity Patterns Within Classrooms and Student Teams in a Constructionist Program of Game Design Learning

• In Event: Connecting the Moving Dots: Comparing Approaches to Coordinating Temporal Analyses of Groups and Individuals

Abstract

Introduction
The Globaloria-WV project is a 7-year state-wide network of game design learning and social media engagement, being piloted in WV schools. Middle, high school and community college students enroll in a blended learning game design elective course offered daily, for 1-2 semesters, credit and a grade. N=534 students for Pilot Year 3 (2009/2010 school year).

A non-profit provides students and teachers digital learning supports via a wiki-based social media platform called MyGLife.org. Individual students interact with each other, teachers, and content resources on the wiki and in class, following a blended learning curriculum. They build games in Flash, constructing a portfolio of project-based work on wiki “Team Pages,” layering design documents, game assets, links, embedded game presentation videos, intra-group communications and final game files. Student activities follow the instructor’s course structure; our data show substantial variability across teams and schools.

Goggins and collaborators have applied group interaction analysis methods described in Goggins, Laffey, Amelung & Gallagher (2010) and Goggins, Galyen & Laffey (2010) to nine different data sets in domains of software engineering, online education, disaster relief, math education and online political discourse. For example, in studies of online graduate education, Goggins, Galyen & Laffey (2010) found that betweenness centrality indicates power brokering, which is the classic definition, and, “constructive lurking.” The authors applied this finding in a software engineering context, and developed a new estimated measure of developer coordination requirements: “proximity” (Blincoe, Valetto & Goggins, under review). Goggins, Laffey & Amelung (2011) describe social network interactions of persons and artifacts in collaborative learning tools.

Most group collaboration occurring in this project occurs within teams of 2-4. In some cases, certain individuals take on a programming expert role in the class overall, serving as peer mentors. We address three main questions:
1. What patterns of cross-time team-level game design activity are evidenced by youth in this context?
2. What patterns of cross-time class-level game design activity are evidenced by youth and teachers in this context?
3. In what ways may findings relate to those described above?

Method
There are three levels of analysis in our data: classes, teams, and individuals. Student interactions and productive game design work are recorded as editing and uploading wiki traces. Our data corpus includes wiki log files for Pilot Year 3, containing user ID, page ID, the text/content that was edited, file upload type, and timestamp. By ordering logs by page and time we constructed two sequences of edit interactions organized by user, and by page. Our analysis will construct social networks by teams and classes.

Results
Results are in progress and use three primary visualization techniques: 1. Graphed comparisons of network centralization and participation over time; 2. Sociogram of weighted networks by team over time demonstrating connections within and between teams during different course sessions and modules; 3. Graph of network statistics by user and team organized through qualitative analysis of wiki content changes. Future work includes testing contribution of group interactions to student knowledge outcomes, as measured by Reynolds (2011).

Authors

• Rebecca Reynolds, Rutgers University

• Sean Goggins, University of Missouri - Columbia