paper discusses creating integrated visualizations that support reasoning under uncertainty and adding visualization support for uncertain (methods)
- lots of points, some below:
- representing the reasoning process
- making self aware of biases that might otherwise not recognize
- visualizing an argument
- can help resolve ill-structured problems
- causality perceptually linked to temporality
- estimation introduces uncertainty
- this fact should be highlighted
- evidence strength and weight
- associated with work in belief systems (?)
- strength == how much; weight == how reliable
- visualization will trigger associations that may bias the reasoning process
- [Goodchild] animations can show auto-correlations in geospatial data
- ignorance of rules
- visualizations can lead to different statistical understandings
- uncertainty and reasoning time
- uncertainty can impair ability to make decision in a timely manner
- may result in better decisions
- [Bill] We’ve talked about this a a couple of times
- hypothesis: correct paradigm might be to take info away rather than add it
- remove the sense of certainty
- slow down reasoning process
- [Lace] NPR reference: hard to read font slowed people down and resulted in better decisions
- [from editor] possible source?: Alter, Al. et al. Overcoming intuition: metacognitive difficulty activates analytic reasoning. J. of Exp. Psych.-Gen. 2007. – link
- goes back to the fact that reasoning with uncertainty is inherently different from quantifying uncertainty
- Goldilocks Principle: you don’t want representations that are either too easy or too hard to understand
- too simple can result in less comprehension
- [Bill] explanation only from 3-min discussion, need to followup
quantifying uncertainty throughout the process very different from displaying uncertainty throughout the process in order to aid the reasoning process
we should pay attention to communicating the process to decision makers