I agree with Howard Nobel’s assessment of the issues with spreading your assets across multiple non-interoperable economically potentially non-viable solutions:
“While I see these applications as being very successful in terms of building communities, well designed interfaces, great features, cost-effective hosting etc I think we are benefiting from being at the start of a growth in this type of system where an ecology of systems is yet to evolve. It strikes me that as more and more of these systems are built then we’ll be struck with the same interoperability problem that we face today within the information environment landscape e.g. not being able to search across repositories/ service providers with sufficient accuracy, needing many different login details, no easy access to full text or full movie, high resolution image etc.”
While I see the login hurdles slowly being brought down with OpenID and other distributed authentication systems, another potential issue is the different semantic frameworks underlying each data storage service, which will affect search and retrieve by the users.
One big limitation of common commercial and open source VLEs is their object oriented approach for peer assessment and more generally for peer activities.
VLEs typically allow interaction between students only when designed in by the course owner. It encompasses usually text-based communication and some form of sharing: students can discuss or answer polls, they can look at shared objects (such as uploaded assignments). This is fine for many common teaching and learning tasks. However, peer assessment and peer learning need to include learning process observation, which is one of the most common teaching strategies for competence development.
Here’s an example: A group of young doctors walk through the ward and in turn look at patient records and talk to patients while the others observe. This allows them to instantly give feedback to the actor and also instantly apply it in their own behaviour. Support questions can be: what did s/he do well? where could s/he have done better? what went wrong?
Although such peer assessment strategies can be captured in formal IMS LD and QTI (Miao & Koper), learning by observing the behaviour and actions of others is not easily accommodated in an online environment and poses serious challenges even to expert learning designers and technology developers. Maybe 3D worlds and VR are more suitable for this type of peer interaction, but one still needs a descriptive design of the pedagogic scenario, a practitioner-friendly authoring environment and a player environment to run it in.
*Miao, Y. and Koper, R. (forthcoming): A Domain-Specific Modelling Approach to the Development of Online Peer Assessment