I'm getting toward the first draft for what is proving a rather demanding E-Primer: "04: Online discourse". What is making it difficult is the incredibly rich literature. I have over 300 peer-reviewed articles (from the last ten years only) and a stack of 18 authoritative titles to search through. So far I have 51 pages of draft; I'm picking at least ten or so more before the initial editing and polishing takes place. I aim to (at least) halve that so that the final product will consist of 30 quality pages!
Anyways, there is a hot topic at present just crying out for some PhD-level activity (or some other form of intensive research project). There are several methods for analysing asynchronous online discourse, and comparative studies between them are few. The results of studies into the nature and dynamics of asynchronous discourse are many, yet spread across different methodologies. As a researcher and practitioner, I would value an analysis of the same set of raw threaded discussion data analysed using different methods. Meyer (2004, PDF) compares four different frames of analysis in the Journal of Asynchronous Learning Networks however various popular techniques were not selected as the basis for comparison. The effect of different categorisations for discourse (such as those of Oriogun 2003 in AJET and Han & Hill 2006 in JALN) have not (to my knowledge) been assessed in terms of how they might influence outcomes.
Wow. Lots of potential to add to knowledge in this area, and establish a study that will be widely cited. If you'd like to go lead author, count on my support!
Ah well, back to the draft...