Interview with Mary Czerwinski
by Juan C. Dürsteler
[message nº 58]
|We had the opportunity to hold a very interesting interview with Mary Czerwinski from Microsoft Research at the conference IV2001.
| || |
Mary Czerwinski is a researcher of the Adaptive Systems and Interaction Group at Microsoft. Mary's research focuses primarily on the usefulness and usability of novel information visualization, the effect of notifications while multitasking, and reminder systems.
Her background is in visual attention and user interface design. She holds a Ph.D. in Cognitive Psychology from Indiana University.
You can find further information at her personal page.
During the Conference IV2001 held last July in London we had the opportunity to interview him. What follows is an excerpt of the interview.
It appears to be some overlapping and confusion between Information Visualisation, Information Design and Information Architecture. What do you think about this.
Well, I think that [Information Architecture] is the process of distilling semantic concepts down into reasonably navigable structures, so I agree with Information architects when they say that this process has to be carried out for any website.
Then, once you have the semantic structure of the web site, you have to design it and, as a part of that information design, one needs to visualise information for the user in order for meaningful patterns of the data to "pop out" or be easily detectable.
These are very different ideas that perhaps all belong as part of one web site’s design. I’m not sure about whom is confused with regard to the three concepts, they are quite well defined and all pertain to the design of a web site.
What are the main challenges in Information Visualisation from the standpoint of Cognitive Psychology?.
One of the biggest challenges, in my personal opinion, is moving the field from a field of researchers that make point designs based on what the technology happens to currently afford , to a field of researchers that have a principled, scientific approach.
Our field needs to develop the ability to measure and evaluate the effectiveness of the resulting information visualisations in order to move the field forward as a science. […] What we need to do is become more scientific in our approach, starting with point designs built around technology but then moving towards principled designs based on real people and real tasks. We also need to evaluate emerging principles of design, and potentially develop theories, in order to move the field forward. […].
This is not to say that there is no place for creative design, there will still be plenty room for that, but the best designers have an understanding of human perception and cognition, and this knowledge needs to be captured, documented and improved upon
Do you think that Cognitive psychology is mature enough to provide these principles?.
Yes the field is mature enough, but we just can't pull up principles from Cognitive psychology, but from Perceptual psychology, from attention research, decision making research, sociology, anthropology graphic design and computer science. So what we need to do is to work together as multidisciplinary teams to create these designs and to move the field forward as a science. You can see already that approach in the human-computer interaction domains of attention and navigation design.
If we just use the findings of perhaps, psychology, certainly we may not ever get where we want to be because the technology is always changing and we never will have that perfect experiment that proves that this process is correct, what we need to do is work together iteratively as a community, in developing our own techniques for measurement and evaluation and our own theories born from multiple disciplines.
Finally, how do you envisage the future of Information Visualisation? What techniques are the most promising ones.
That is a risky question! But I like science fiction. I think that the next 5 years is a little bit easier to predict because I don't think a lot will change.
However, I am predicting-- I would even bet money on it-- that you will see more 3D environments, you will certainly be seeing more 3D animation to help you understand complex changes as data are manipulated, and I think you will see a better use of transparency in the user interface regardless of whether or not the UI is 2D or 3D.
I think you will see an appreciation for the user and the user's tasks. We'll see more visualisation of network data just because of the advance of the XML language. Those things are easy to predict since they are almost here.
One last thing that I would include is that the AI (Artificial Intelligence) community has been contributing quite a bit to the Information Visualisation community in task prediction, context prediction and I think it's the right time to bring these two communities together.
* We thank Mary Czerwinski for allowing us to use her photograph and for the correction of the original transcription of the interview.
Links of this issue: