Abstract:

Activities such as collaborative learning are dependent on the participants ability to coordinate their activities in order to strive for a common goal. According to Roschelle & Teasley (1995), collaboration “...is a coordinated, synchronous activity that is the result of a continued attempt to construct and maintain a shared conception of a problem”. Further, coordination can be understood as “the act of working together harmoniously” (Malone & Crowston, 1990), and is often seen as a requirement for collaboration (Gross, 2013). We can also understand coordination as an overhead in a collaboration, and the less time and effort spent on coordination, the more time can be spent on getting "things done".

When collaborating around a table, people have numerous ways to manage the collaboration and coordinate the activity at hand - we see the action of others and we can talk and ask questions. We can pick up more subtle queues such as where others are looking and whether they are nodding or sighing. We can direct the attention of others by pointing and referring to objects within the collaborative area. Moving the collaboration to the web, depending on the type of system, many such cues and actions we need and are used to utilize in order to coordinate the collaboration is lost. We thus need to re-introduce such cues and actions as a form of technological approximations (Ackerman, 2000; Ackerman & Halverson, 2004) into the webbased collaborative system, both enabled as well as restricted by the technology available to us. We argue that when designing such technological approximations, we need to consider the characteristics of the underlying technological foundations in order to support participants' coordination of activities.

We have been developing a realtime web based shared workspace in the form of a whiteboard for a number of years, and in this paper, we will focus on delays in the transmission of data carrying information of value for coordination activities. While technology now enables us to send and receive data at very high speeds through the internet from almost everywhere we are in the world, we always have to deal with a certain delay. Latency and jitter (i.e. delay and variance of said delay over time) are ever present and is conditioned by the network topology and the technology therein, as well as by sheer distance - we can never move data faster than the speed of light. The existence of these two types of delays could potentially pose a great challenge for designers of collaborative systems, especially if we consider collaboration as a synchronous activity as defined by Roschelle & Teasley (1995). What happens to the coordination of activities when synchronicity is just "almost" synchronous, as well as varying over time? And how do we design technological approximations so that we can support coordination and maximize time and energy to learning activities?

During a recent phase in the design of the collaborative system, we focused on documenting the characteristics and subsequent effect of latency and jitter in various real world contexts of use, from high-speed wired connections to low quality mobile network connections in rural areas. This paper outlines a number instances where the observed network characteristics affected the design of the technological approximations used in the system in order to aid coordination.

Keywords:

Collaboration, coordination, shared workspace, CSCL.