Sunday, September 23, 2007

Day 1 of Oz-IA 2007

After an alarmingly early start to the conference (8:15!!), the day settled down to the rhythm of presentation-break-presentation-break.

We kicked off with a presentation by Liz & Andy from Fairfax Digital talking through the user experience design process for the recent redesign of the and Web sites. This was a great insight into the process of uxd at a large organisation, and the quality and candor of their commentary was exceptional.

Iain Barker took up the baton from Liz & Andy, running through an overview of recent research into the issue of visitor scrolling behaviour, browser window height, and the implications for interaction designers and information architects. Some really interesting insights into the issue of 'the fold' and how IA thinking might be adjusted to accommodate the disappearance of this myth.

Hural Inan from Bienalto Consulting took a tour through the use of web analytics to optimise landing page layout and content. ( I'll admit to not paying as much attention as I might have, as I was preparing my own presenation, which was up next...)

Next up your's truly talked through a number of statistical analysis concepts, examples & bad practice. And then lunch!! (Photo by Ruth Ellison)

ROI in Information Design was the post-lunch talk by David Sless from CRI. Aside from a very interesting look at the practice of information design, there was a very clear message to the conference to measure what we do. To paraphrase: if you don't measure it, you have no right to espouse the benefits of your work. There was also a bit of semantics about the distinction between usability testing and diagnostic testing, which I'm sure others in the room took more seriously than I did.

Stephen Collins' talk Love in an elevator was up next, looking at the role of IA and UxD in large organisations, which paired nicely with Faruk Avdi's introspective look at a history of projects within the Department of Education in NSW, and the role of external agencies & consultants in those projects.

Patrick Kennedy from StepTwo gave us a very interesting case study in coaching & mentoring an inexperienced UxD/IA team as they journey through a large-scale organisational site re-design. The presentation walked us through the process, challenges, tactics & learnings for the team and for Pat as the mentor.

And last up we had the opportunity to listen to Rashmi from talking about their ninja-style of iteration, release, iteration, release. An anti-process talk was a nice way to end the day, and very Web 2.0 :)

That was it for the serious stuff of the first day of the conference, and so it was off for a few drinks to wind down.


Anonymous said...


It was sooooo good to see a fellow-stats-o-phile try to return everyone the IA community to really-real analysis.

An excellent presentation

When can we expect to see multiple-linear-regression??? Cluster analysis of tags?? Ordinal logistic regression??


Steve 'Doc' Baty said...

M: if I wrote the book version, would you keep a copy on your shelf? I suspect that's the only way to cover topics like multi-variate regression, or clustering analysis, or ordinal logistic regression, or discriminant analysis & a bunch of other topics.

Glad you enjoyed the presentation.