(1) Methods for discovering and measuring cognitive biases. TU Graz researchers based their studies around the Selective Exposure Experiment to measure biases such as the Clustering Illusion. The team investigated how cognitive biases could be expressed so that mitigation strategies could be specified in a system implementable manner. The team also investigated how cognitive biases could be identified based on user observation and automatic logging of user interactions with a visual analytics system.
(2) Methods and guidelines for mitigating cognitive biases were developed. The team adapted guidelines already reported in the literature the challenges presented when interacting with visual analytics systems. One approach adopted in the design of the VALCRI system was a visual design strategy that makes relevant data visible and persistent in order to create data awareness as an approach to mitigate biases such as confirmation and availability biases.
Addressing target audiences and expressing needs
- Collaboration
- Fellowship to advance my/our research
Together with researchers at Middlesex University, TU Graz seeks collaborations and funding to study how to design of visual representations of automated data analysis techniques used to augment human intelligence to mitigate cognitive biases in more natural, non-intrusive ways. Having a “MS Paperclip” pop up periodically to remind a user of cognitive biases, is considered intrusive, and will interrupt the cognitive momentum of the user during investigative sense making activities. We prefer to use visual design methods such as principle of focus and context.
- Research and Technology Organisations
- Academia/ Universities
