[drupal-docs] Hello from Computer Human Interaction 2005

[Kieran Lal]

> Hi, I am attending the CHI 2005 conference in Portland.  I thought I'd 
> share what I am learning briefly and how I think it might apply to 
> CivicSpace/Drupal.
>
> The first session I attended this morning was on large communities and 
> had 3 papers.
> The first talk involved modelling the efficacy of communities.  This 
> was an exciting idea because I'd love to have a way to model 
> communities and understand how effective they are.  There was some 
> good math in the model an he used concepts like path analysis and 
> factor analysis in creating the models.   He started off with 13 
> factors, community beliefs and behaviours, for measuring community 
> efficacy and then reduced it substantially to measure one particular 
> community which he measured twice during this communities 13 year 
> life.  Modelling communities is interesting but I doubt we can draw 
> any results until the models start getting applied widely to more 
> communities.  He did mention that he was to interested in applying his 
> models to Open Source and so we will see if he can get some attention.
>
> The second paper was basically restating obvious stuff.  Moderated 
> communities perform better and more reliably in completing tasks than 
> unmoderated ones.   Nice to know, but I drew the conclusion that 
> bosses are useful in online communities.
>
> The third paper validated the use of chat as very important in 
> distance interaction.  As someone who sits in IRC a reasonable amount 
> I wasn't surprised.  This project did analysis on collaboratories for 
> earthquake engineers and found chat was more useful than visual and 
> auditory interactions.  No surprise there.
>
> In the afternoon, I attended a session on document interaction.  The 
> first paper asked the question was there a need for representing the 
> concept of paper clipping in the online world.  The answer was yes.  
> No surprise, we see this commonly in blogging with the use of 
> trackbacks.  41% of users clipped stuff from magazines to share, and 
> 28% of people clipped stuff for reference later.  What was interesting 
> is that most people don't consider themselves clippers but they did 
> clip any way.  The take away is that we should consider having a way 
> for a CivicSpace/Drupal user to have the ability to clip digitally to 
> share and reference.
>
> The next paper was a very good overview of studies of paper use in the 
> office.   Unfortunately, this topic has been studied to death and the 
> results are well known among the elite in the field.  At least that's 
> what they told me when I bumped in to some famous friends from IBM and 
> MIT when I ducked out early.  Let me share with you since, you 
> probably aren't an expert.  Paper is good for reorganization.   
> Retrieval isn't good enough to replace a document management system.  
> Email is very good for organizing, but it is overloaded.  The 
> hierarchical structure and lack of transparency in file systems make 
> them bad for the 'At hand' management of information.   If you want a 
> screen shot of my desktop that will help explain this concept better.  
>  What I liked was her two recommendations, and I'll extract her 
> results to CivicSpace/Drupal.  CS should have it's design focus on 
> accessibility and decreasing functionality.  Also we should focus on 
> flexibility and on unobtrusiveness.  We have been having this debate 
> on the Drupal Docs mailing list recently.  Books, nodes, and 
> organizing content is too difficult in Drupal. It's nice to have some 
> external results to point to anyway.
>
> The final talk was on the use of categorization in searching.  No 
> surprise that Google doesn't use it.  Basically, you should use 
> categorization(Think topic buckets from your search results) when you 
> want lots of results(>3) or when you don't find any results you want, 
> in the results from your search.
>
> Then I went to the Usability ROI talk.  I think what summed it up was 
> the pronouncement that we were attending the usability ROI funeral.   
> Most of the discussion focused on how usability should be valued as a 
> risk mitigation factor or a strategic ROI factor.  In short there were 
> few convincing arguments.  The two insightful comments during the 
> panel were that usability designers were brought in too late,  and 
> that usability should be focused on the people with the power.   I 
> asked what the value of usability was for open source developers and 
> got various answers to the wrong question.  Hopefully things will be 
> more insightful tomorrow.
>
> Cheers,
> Kieran