[drupal-docs] Wrap up from Day 3 of Human Computer Interaction 2005

[Kieran Lal]

Hi, today was a more fun day.

Special Interest Group on End Users Creating Effective Software
This was an interesting topic because even though many people who use 
CivicSpace do not directly write software they often have to do complex 
configuration and even create things like regular expressions for 
blocks, or php code for blocks, which do involve an element of 
programming.  Many of the people in the room were academics and focused 
on end user programming research topics.  But there were a number of 
people from industry as well.   Two people that I thought were 
interesting from a CivicSpace perspective were Michael Toomim, and 
undergrad student from UC Berkeley who is working on tools to help 
reduce code repetition.   This is very common in websites, Content 
management systems in particular.  It would be interesting to see if 
his tools could be applied to the theme editor or a php editor to help 
people extend existing themes and modules.

The second person who was doing work I found applicable to CivicSpace 
was Prof. Mary Beth Rosson, formerly from Virginia tech, and now Penn 
State.  She was working on three topics.  The first was a tool called 
Click for developing lightweight web applications.  Her second project 
was working with non-profits to help them with web development, and 
information systems.   Her third project was a web survey to 
characterize web development for people who don't want to be be web 
developers but are (I might have that wrong).   I think Drupal has 
shown a great example in partnering with academics and we should 
continue to expand our reach to other academics who can provide not 
only research insight but a talent base for recruiting students to work 
on our projects.

Can usability scale up?
The speakers were two well know personalities in the field.  One was 
Eric Shaffer, found and CEO of the largest HCI consultancy based 
primarily on running a large HCI shop in India with an international 
set of practitioners who work closely with clients.  Jared Spool runs a 
well known User Experience research boutique firm.   Eric argued that 
usability can scale up and cited his large practice with a solid 
methodology to show that it works.   Jared Spool cited positive 
usability examples like Google, Yahoo, and Amazon that had in the past 
very small usability teams versus HP, IBM, and Microsoft who 
collectively have around 1000 usability employees.  The debate was 
definitely meant to be entertaining and the debaters took extreme 
positions to this end.

I however, was very interested to know if usability could scale up so 
that the dozens and potentially hundreds of modules and third party 
applications that CivicSpace will be working with could in fact 
successfully use the practices of usability to improve the user 
experience.   Here is what I concluded from the debate and other 
conversations.  User Experience is a craft.   What will make usability 
scale up is having good tools for practitioners.  In software, 
programming has scaled up because we have better languages, compilers, 
and better integrated development environments.  Likewise, when card 
sorting, log analysis tools, rapid user design tools and many other 
tools are widely available, then I believe usability will scale up.  I 
ended up having lunch with Jared and a few other folks and was 
impressed with his focus on craftsmanship that delivers business 
results.  I also got a chance to speak with Prof. Simone Barbosa from 
PUC-rio.br.  I am hoping to recruit some Brazilian's on to this project 
since they have a strong focus on social software and their government 
is extremely friendly towards open source.

Understanding Users and Usage patterns
The first talk was on Patterns in Media Use for collaboration.  This is 
very important for the Drupal/CS community because so far most usage 
has focused on textual information.  I came late to the talk but the 
summary was that there were four types of use:  communicating, 
exchanging mixed objects, coordinating,  (e.g., of status reports), and 
semi-archival filing.  They observed various uses of chat, and informal 
uses of files.  Having sat in #drupal and #civicspace for a while I 
would say that was consistent with my experiences.  For example, I 
randomly asked UnConed to change his graphic for theme garden to have a 
cocktail in the picture and he did it.  On another occasion I was 
debugging the install of Bugzilla and worked with their developers to 
submit a patch in IRC.  Both experience were collaborative for me and I 
do expect to have this kind of support in the work that I do.

The next project was on analyzing usenet usage patterns.  What they 
found was that power users were interested in people, and novice users 
were interested in newsgroup summary information.   I think this an 
important observation and could be particularly interesting during 
political campaigns.  Power users would know who they want to listen 
to, and novices would want to see how this campaign was doing overall.  
As we build mailing list functionality we should realize these 
different needs and incorporate this functionality into our designs.  
http://netscan.research.microsoft.com

The last paper was on the role of media in diary studies.  The take 
away was that you need to combine media capture with structured 
question and answer sessions.   Future work will involve tightening the 
loop around capturing memory cues and doing interviews sooner before 
memory starts to degrade.

IDEOs design of mlife
I ducked out of the last talk and caught part of the IDEO talk on their 
involvement in the design of phones for use with ATT's mlife.   IDEO is 
one of the most famous design firms in the world and it seems that all 
usability paths lead to  doing user experience design as early as 
possible.   Design is a creative process and you have to do wacky 
things to get users and designers out of their normal patterns so that 
you can learn from them in context.   For IDEO this involved user 
studies where they paired the wrong team members in a car with their 
phone and told them to go somewhere.   My personal favorite use of 
Google is when I call my wife and tell her I forgot something and I ask 
her to look up some information on Google.  IDEOs user studies 
validated that this was a common and practical use of the phone.   They 
also developed prototype costumes and had users wear phone costumes 
while others push their buttons.  I validated this as an effective 
technique with a designer from Yahoo photos on the trip to the airport. 
   They felt that a good design had solid principles and that you had to 
stick to them.  For mlife, they were social, relevant, and time slice.  
I'd be curious to know what Drupal's principles are.

Dan Gilmore from IDEO concluded with 3 points.  Design for beyond the 
moments of use and reach out to other stake holders.  A multiplicity of 
research methods is often the most effective strategy for design.  Time 
pressure of commercial design is too much and can lead to bad design, 
so stick it out.  In the question period, he had a great comment.  He 
said that usability testing is for folks who haven't been paying 
attention to users during the design.

Summary ideas from CHI2005
I am walking away from CHI with one observation, one main learning, and 
one plan for action.   My observation, is that 1/3 of the conference 
was students and it seemed the majority were young women.  If Open 
Source is going to have a successful user experience it's going to have 
to recruit and create valuable roles for these young women.   My one 
learning was that social context is necessary for content.  From the 
talks on email usage, to uses of media, to analysis of conversations, 
what I learned is that it matters who is involved in the information.  
I have been studying how Drupal is going to integrate with the new 
Community Relationship Managment system and it's increasingly obvious 
that a good design that allows for a social context of content is where 
value is going to come from.   My plan of action is to focus on getting 
more usability tools released in open source.  I am not a better 
programmer because I understand the intricacies of C++ memory 
management. I am better programmer because I use a dynamic programming 
language(PHP) and I don't care about memory management anymore.   We 
are not going to be a better user experience designers because we 
understand more about usability.  We are going to be better because we 
all do usability design because it's free and easy to do.  Better tools 
will make us better crafts people and we need to make those tools 
together.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: text/enriched
Size: 8675 bytes
Desc: not available
Url : http://drupal3.drupal.org/pipermail/documentation/attachments/20050408/50aacdff/attachment.bin