Research shows that learning in groups improves students' achievement of learning objectives. Some indicate that a facilitator will smooth the process of collaboration (Johnson et al, 1987; Hooper, 1992; Moore and Kearsley, 1996; Brandon & Hollingshead, 1999; Bernard et al, 2000). Advocates for peer-controlled collaboration claim that learners usually feel more comfortable discussing without the presence of the instructor. This may be explained by research which indicates that when participants have more active control of the learning process, learning increases (Jensen, 1996). This previous research lays a foundation for hypotheses about online peer collaboration using different types of moderators and strategies; however, there is a lack of empirical evidence for making strong recommendations. Which approach is more effective for group collaboration, the peer-controlled online collaboration forum or the structured and moderated online collaboration forum? This question is a difficult one to answer, but this study begins to answer the question by investigating the quality of peer online discourse. This study discusses possible activities to enhance Web-based peer collaboration activities as well as the roles that instructors and instructional designers can take in the creation of effective online peer collaboration activities. In this study, complete electronic discourse transcripts of college students in a collaborative problem solving assignment will be analyzed to measure the quality of peer online collaboration with the two different methods, to help understand the group dynamics and types and patterns of social interaction which occur online. (Author/AEF)