This important work will lead towards interventions for people to improve self-regulation based on their own personal preferences. This could translate into higher rates of compliance with known strategies for improving health and in the long run lead to better health outcomes.
- Nels Oscar
- http://www.igert.org/profiles/5424
- Graduate Student
- Presenter’s IGERT
- Oregon State University
- Shannon Mejia
- http://www.igert.org/profiles/3793
- Graduate Student
- Presenter’s IGERT
- Oregon State University
Judges’ Queries and Presenter’s Replies
Presentation Discussion
-
-
This work is also interesting in a more general context. Not only can personal preferences influence what should be shown to a person, but also personal abilities. For example, a person’s literacy level (including visual literacy) can be used to determine the best way to present data to them in order to help them make decisions.
-
Further posting is closed as the event has ended.
Diane Cook
Faculty: Project PI
The idea of personalizing visualization is very appealing! How does your algorithm make an automated assessment of the individual’s level of busyness/free time for a given day?
Nels Oscar
For the current version of the algorithm, we’re actually using the amount of time that the participant spent filling out their daily questionnaire. When we were analyzing the PULSE data set we noticed a correlation between the time spent on the questionnaire and the time spent perusing the visualization. There are some problems with this metric such as people leaving their browser open, but it makes a decent approximation.
Julia Hirschberg
Faculty: Project PI
what particular aspects of your visualization do you think will be useful, for whom, and how?
Nels Oscar
Something I would like to point out is that the contribution here is not strictly a visualization, but a construction framework that supports visualization creation. We envision this being useful to any individual interested in monitoring their own data. We’ve drawn from the recent surge in interest in the quantified self movement as motivation. Our approach is intended to target information at individuals who are not interested in interacting with a visualization toolkit to gain insight, but instead to have that information displayed directly from the point at which the frame is first rendered. This will improve awareness for personal health applications in particular, but the metrics we’ve used for categorization are not the only possibilities and this could improve the utility of visualizations in a more broad sense.
Mostafa Bassiouni
Faculty: Project Co-PI
I wonder if you have considered an approach that allows the person to select and control the level of detail for their visualization and data exploration. After the person becomes familiar with the dashboard, he/she is allowed to select a specific level from Fig. 1. The dashboard will always provide the recommended setting but the person is allowed to override this recommended (default) setting. Please comment on the pros/cons of such an approach.
Nels Oscar
One of the things that we didn’t do as good a job as I would have liked is demonstrate the possibilities for interaction that a system like this provides. As I’ve gotten into a little in a previous response, the idea is to give the user as good of a starting point as we can come up with. Giving the user the freedom to select their own level of detail would be a perfectly reasonable thing for the system to allow. It could also help to inform the system’s appraisal of the user, leading over time to more intelligent (better personalized) categorizations. I’m unsure at the moment of what the possible cons of allowing this would be. In the original PULSE study participants were given the option to ‘show more’ details in the data visualization. The most common choice was not to exercise the feature, so in giving more options and more interactivity the greatest detrimental aspect should be that it goes unused.
Mary Kathryn Cowles
Faculty: Project PI
The stereotype of older people is that they are set in their ways and don’t like change. Do you anticipate that some users would be confused or upset if the same data wasn’t displayed in the same place in the same way every time they accessed visualizations of their information?
Nels Oscar
We haven’t yet addressed the issues presented by individuals who are particularly sensitive to change. A flippant answer would be that the system could easily be turned off if they prefer a static representation. I’ll have to think a little more about the ramifications that our system could have for that user story.