User:Nils peterson/Wiki4OWL

From wsuwiki

 This subpage to a user page would be where the user would post an item 
for OWL review. The discussion page is where the feedback would come.

Analysis of Conversational Scale and the Design of Online Communication Tools

Experimentation among a group of individual bloggers using trackback to connect posts provided insight into the various niches that electronic communication tools occupy. This article introduces the idea of "conversational scale" to distinguish among a range of tools and to provide further design insight into features needed in a university sponsored blog/ journal tool. Knowing the scale of the conversation is anticipated to help understand the features needed in the software, and to help users distribute their conversations across the correct media. The experiments leading to these observations were done by a group of staff and students connected to Washington State University's Center for Teaching, Learning, and Technology (CTLT). The participants have experience with a range of online conversation tools, including instant messaging, email and email lists, chat and threaded discussion. To varying degrees, these tools have been part of the groups' professional and learning life for multiple years. During the summer of 2003, members of this informal experimental group had undertaken a research exploration to analyze features of online courses (designed by members of CTLT) that were perceived as successful. The exploration included reading threaded discussions from the course archives and reviewing the current literature regarding online collaborative learning. From that work two guiding principles were developed that shaped a decision to implement a new university collaborative writing tool.

1. Writers should not be separated from their writing. Often, students writing within course management systems is archived and becomes inaccessible after the end of the term. The study found evidence of students in a learning cohort attempting to refer one another back to writing from previous terms that they, and or their audience, could no longer access.

2. Conversations should be webbed, not threaded. In the typical threaded discussion, conversations diverge from an original prompt along several branches. It is quite possible that in some of the leaves of the discussion, the same ideas are being considered, but there are typically no mechanisms for webbing leaves with common topics back together.

Following this analysis, and a needs analysis that found use cases for student journals, the design team created a specification for a new collaborative tool. In addition to the principles above, were goals of facilitating the formation of learning communities and having writing within the University community readily connect to writers on the open Internet. As work proceeded, it became obvious that a core of the requirements for this new tool could be met with a blog with trackback to web posts together. During the summer of 2004 we implemented dotText, an open source blog engine, written in .Net. Our implementation is single signon with the University portal and anyone with a portal account can have a blog account automatically created simply by signing in to the the blog application. Our implementation is called PBJ, variously meaning 'portal-based journal' and 'pretty basic journal.' The tool launched in Aug 2004, see notes on adoption elsewhere in my journal. Initially, PBJ had an unauthenticated comment feature, common in many blogs. We disabled this function a couple weeks after launch for three reasons:

• Spam robots place comments into under-monitored blogs and we did not want the university community to be subjected to this.
• Experience at another college with hate speech in a blog caused us to want to avoid anonymous posting. We decided that authenticated hate speech from members of our community could be managed by existing administrative means.
• Our principle of the author retaining control of their texts was violated by the comment mechanism, where the comment became the property of post to which it was attached. When comments were disabled, there were some strong reactions from the small community of early adopters. Some liked and some disliked the change. Steve and Josh provide further insights to differences between PBJ as a medium and a threaded discussion or chat. Previously, I started to outline what I understand to be the useful differences between a PBJ Post and an Article. And Theron Desrosier has a showed me a nice piece just appearing in the Chronicle on wikis. All these ideas fold together for me around the question of conversational scale.

By "conversational scale" I mean the size, rapidity, , and number of referances in an utterance. Conversational scale, as used here, attempts to capture ideas about the rapidity of exchange, the semantic size of the conversational unit, the rapidity of generation of the conversational utterances, the amount of context carried within (or outside) the utterance, the anticipated longevity of utterance, the number of conversational participants over that lifetime, and the inclusion within the individual rhetorical moves of references/citations to other work. At the small end of the scale lies Instant Messaging. Individual elements in the conversation are small; often sentence fragments, with small amount of semantic content. They are rapidly generated and depend heavily on context outside the actual utterance. Longevity is short, the conversation is (typically) not recorded and between two participants. Other works are rarely cited, or referred to, during the conversational exchange. At the extreme large end of the scale is a work like Plato's Republic and the texts that have reflected on it. Many of these rhetorical moves have been of large scale, taking much time to produce and have referenced one another and many other contextualizing documents. The discussion has persisted over a long period of time, involving many readers and reader/authors.

Given those endpoints, I would put typical electronic discussion tools into this rank order: Instant Messaging, Chat room, threaded discussion, blog post, and blog article. A peer (or slight lesser) to threaded discussion would be the comment to a blog post. Email might also be a peer to threaded discusison. I have not included a wiki in this continuum because it is not clear to me that it lies in the same plane. My hesitation stems from the key wiki feature of multiple authors modifying the document over time. The document is, in some sense, always the end-point in the conversation, and not its trajectory. It is a difference between process and product. This analysis helps me understand that our original principle of not separating a writer from their text may be too extreme. Writers, given the choice and understanding, may elect to give some texts to others (in the form of blog comments), and retain other texts (shared in the from of trackback comments). There is a missing section of this analysis that examines the experimental group's long standing experience with, and understanding of threaded discussion. We came to blogging, and comments, from a discussion perspective-- smaller conversational scale. Many of the early posts in PBJ are acting like threaded discussion moves -- they are of a small conversational scale. So, given this analysis, how should we think about comments in a blog, and what additional tool(s) (if any) are needed to support university classes, trans-course learning communities, and other discourse activities within the University?