All entries for Friday 21 October 2005

October 21, 2005

Research Plan: ideas for researching networking, narrowcasting and broadcasting by bloggers

Follow-up to Presentation of blogs at Oxford Brookes from Transversality - Robert O'Toole

For a while I have been searching for approaches to doing research that will uncover some of the patterns of behaviour that we see in the use of Warwick Blogs (and more importantly, why some of the expected patterns of usage seem to be so rare). Now I have a few ideas.

During my recent presentation at Oxford Brookes about Warwick Blogs (during which there was an excellent debate), someone made the observation that from the perspective of some outside person browsing through the latest entries in the "showall" aggregation page, it would be hard to understand why anyone would read our blogs, and consequently (given that most people write with the expectation that someone else will read their entries), why would anyone therefore ever write anything?

This wasn't a crude condemnation of the standard of writing within Warwick Blogs. Rather, the observation was that it was hard for an outsider to understand what is interesting in most of the entries (although some are genuinely interesting to the outside world).

Having just read Mark Buchanan's book on small world networks, I could see the error in this observation.

The assumption behind this attitude is that the entries within Warwick Blogs are on the whole written so that any random person may come along and find them interesting and worthwhile. The question assumed that writing a blog is a form of "broadcasting". Indeed this is understandable, as it tends to be such "broadcast" style blogs, aimed at a general and unknown audience, that have caught the interest of the media. The reason for this bias may well be that the traditional media somehow sense that they are being challenged in their dominance of broadcasting.

I suspect that, on the contrary, many blog entries are written with relatively narrow audiences in mind (narrowcasting), but with the open possibility that a wider range of people may end up reading them. This is entirely consistent with the "small world network" model. Meaningful activity occurs on a local and limited scale, determined by the individuals in close proximity. In the case of blogs, that is likely to be other friends who blog, and probably in many cases the friends who introduced the author to blogging. That makes for lots of very small and localized activity, meaningful within isolated cliques.

But there's more to a small world network than that. Operating within a clique presents the risk of being disconnected from the outside world. One may lose track of how to connect to new and unknown people. For a student, that is potentially disasterous, as at some point reality will bite and the clique will dissolve (reaching the end of the course for example). So it makes sense for members of the clique to make some consideration of the outside, to open themselves up to possible extra-clique connections. The clique makes sense of the activity, but the outside is needed to give a small reality check, to add a little affirmation that the activity has value beyond the clique, and to safeguard that value should the clique suddenly dissolve. To cope with this, small world networks emerge, with "hub nodes" that provide a connection between the clique and the outside (or other cliques). These hubs take many forms. In some cases a single key individual may play the role. In other cases, a broadcasting channel may do the trick.

What is really interesting, and perhaps not yet thoroughly researched, is how different the arrangements of different types of clique and hub may be of great significance. For example, in designing a new interpersonal communication system that relies both on cliques and on hubs, if we get the type of hub wrong, then the development of the system may be restricted.

My question then is this: what is the arrangement of cliques and hubs in Warwick Blogs? What kind of small world network is it? Could it work better with a different arrangement of cliques and hubs? Fortunately, I think that it should be easy for us to get empirical data on this, so long as we ask the right questions. Our questions will be based on a model of the relevant types of clique and hub. To start creating this, I have listed some of the forms of "casting" (broad and narrow):

  1. broadcasting – every entry appears on a "showall" aggregation page that lists the latest publications from the whole system;
  2. segmented broadcasting, with entries appearing only to readers with certain membership status (eg student only, university member only);
  3. directory focussed broadcasting – an entry appears in "showall", but the author is writing with consideration of one of the groups represented by an aggregation page in the blog directory;
  4. tags based broadcasting – author targets their entry to people who are likely to be viewing all entries with a known shared tag;
  5. Google targeted broadcasting;
  6. narrowcasting to a limited known predefined group, controlled by privacy controls.
  7. narrowcasting to a limited group, controlled by privacy controls and "favourites" subscription.
  8. narrowcasting in which the privacy permissions are set to allow anyone, but the author tells specific people (often family and friends) to look at the entry (often by email) – added following Graham's suggestion.

And then we will need some questions that we can use to work out which of these are being used, and to what effect. For example:

  1. did the author consider the mode of casting – the audience?
  2. who did they consider the audience to be?
  3. did this affect what they wrote?
  4. were there instances in which they changed what they planned to write because of the mode of casting?
  5. were there instances in which they changed the mode of casting because of what they wrote.

If you are interested in joining in with this research, please contact me