Ordering Effects in a Role-Based Scaffolding Intervention for Asynchronous Online Discussions

Abstract

A common scaffolding approach in computer-supported collaborative learning is the assignment of specific roles to the participants in online asynchronous discussions. Previous work has demonstrated how this type of scaffolding can result in student contributions of greater depth and quality. However, since students necessarily experience the roles in varying orders, it is important to consider whether the ordering impacts the outcome. This paper addresses the issue by examining a scaffolding intervention that was deployed in an asynchronous online discussion forum, where students were assigned to lead the discussion in one thread as the ‘expert’ and to participate in other threads by asking questions. A network analytic approach was used to visualise and quantify several potential ordering effects within the intervention. The constructs of cognitive presence and cognitive engagement, from the Community of Inquiry and the ICAP frameworks, were used together to measure the depth and quality of the discussion contribution expressed in each message. The analysis confirmed that the contributions made while the student was in the ‘expert’ role scored significantly higher for both constructs, but found that the order in which students took on each role had little impact on the quality of their contributions to other threads. This result contrasts with earlier work on single-duty roles that found an advantage in being assigned certain roles early in the discussion, and suggests that instructors should feel confident in rotating more complex user roles between students.

This work was supported in part by the EPSRC Centre for Doctoral Training in Data Science, funded by the UK Engineering and Physical Sciences Research Council (grant EP/L016427/1) and the University of Edinburgh.