Jim Hudson

Biography
Jim Hudson is a Ph.D. Candidate in the College of Computing at the Georgia Institute of Technology. Working with Amy Bruckman in the Electronic Learning Communities (ELC) Research Group, Jim’s research interests in human-computer interaction (HCI) fall at the intersection of two areas: computer-supported collaborative learning (CSCL) and computer-supported cooperative work (CSCW). His research focuses on understanding and promoting effective group collaboration through the design of socio-technical systems. Specifically, his dissertation work examines participation in learning environments. Other research projects include interruption in business and home environments and distributed collaboration in safety-critical environments. Recipient of the IBM Ph.D. Fellowship, Jim has interned at IBM Research with Wendy Kellogg and Pitney Bowes with Austin Henderson, where he continues to consult.

Position Paper
Chatrooms are not like traditional written exercises, nor are they like spoken discussions. Instead, they fall somewhere in between—features of both written and spoken conversations appear in chat environments (Collot and Belmore, 1996). As such, they hold the possibility to engender patterns of participation that might not occur in more traditional environments. In business environments, for example, group chat systems allow the possibility for lightweight discussions (Herbsleb and Handel, 2002) that help alleviate some of the problems of interruption (Hudson et al., 2002).

Research on foreign language education has shown that features of both written and spoken conversation play important roles the pedagogical use of chat environments. Some of the earliest research on chat for language learning focused on deaf students learning English. For this group of students, the English language is typically fixed in written form. They never have the chance to play with the English language in conversation. Using chatroom discussions allows this group of students to experience English as a dynamic, playful language (Bruce et al., 1993). For other students, however, the written aspects of chat environments allows a bit of distance (Hudson and Bruckman, 2004), which helps students feel more comfortable participating in online discussions (e.g., Kern, 1995).

As chat environments play a more prominent role in education and business, we must ask how the quality of conversation changes when it moves online. In my work, I am studying the quality of conversation in a college course on professional ethics. In particular, I am comparing face-to-face, small-group discussions with online, chatroom-based discussions. In talking about this course, instructors stress two important dimensions of “quality”: (1) Students should discuss the assigned topics without veering too much into unrelated topics, and (2) in their discussions, students should encounter and engaged a variety of perspectives on any given topic.

Comparing topic digression in the face-to-face, small-group discussions leads to the unsurprising conclusion that students go off topic a few times during the course of a conversation. Typically, this occurs because students introduce a concrete example to make it easier to converse about the abstract topic. While discussing the concrete example, students begin focusing on features that do not tie in with the more abstract topic. Once this happens, it is easy for conversations to veer off topic. Online, however, off-topic discussion is rare. Students still introduce examples to make the abstract topic concrete, but these examples rarely lead students off topic. Early results suggest that threading patterns help explain much of this difference. In face-to-face settings, there is only one thread of discussion; each comment responds to the previous one. Online, multiple threads easily arise, which allows an abstract thread of discussion to co-exist with conversation about a concrete example.

I believe that better understanding the social and psychological mechanisms that new technology enable can help us to better design computational systems to support the types of interaction that we desire.

Back to Social Computing Symposium 2005