Integrated Science Lab
Decision making and operational management in health care, business, finance, public safety, and environmental protection depend upon information systems that enable human cognitive processes -- learning and discovery, insight, creativity, reasoning and problem solving, coordination, and communication -- to be informed by data and computational analysis of information. Increasingly these data environments are "big" in the sense of volume, variety, and velocity. Frequently they are also uncertain in origin, relevance or veracity, and must be integrated with deep understanding of organizational capabilities and processes. Perhaps most critically, the decision made and proceedures followed must be ethically grounded, consistent with law and policies, and justified in relation to human beliefs and values. This last characteristic necesssarily requires technological support for "due dilligence" on the part of the person or people responsible for the decision.
Design of these information systems must be informed by both psychological science and computational analytics. To do this we are integrating cognitive science and design science to create information systems for data-intensive analysis, decision-making, and operational management. This integrated science must be precise enough to effectively guide technology builders and interaction designers. It should also guide organizations to adopt new processes and support new ways of training data scientists and their collaborators and managers. Our work often begins when a cognitive ethnographer is embedded with the decision-makers in the environment in which they work— in the field, office, etc.-- taking field notes, and recording audio/video These papers are reviewed by experts in the application area to confirm their accuracy and sometimes in applied social science venues to confirm methodologies.
Based on those reports we use a cognitive engineering approach to develop information systems that better support knowledge workers s to use data more effectively in their organizations. These systems are evaluated "in the wild" with field experiments that record interaction between a visual analytics trained data scientist and one or more subject matter experts as they conduct a data analysis, make an informed decision, and come to an agreement about a course of action. Video and keystroke analysis of this joint activity is mined to build a design language of cognitive processes and methods by which they can be shaped by dynamic and interactive visual information systems. These papers are reviewed by experts in information system and computer technology.
We may also do laboratory studies and field experiment testing on specific aspect of the interface to determine how analysts process information contained in those visualizations and how they concieve of and execute queries using the interface. These studies extend distributed cognition and visual cognition theory and methods to focus on how interactive visual information systems shape thinking. The papers are reviewed for psychological/cognitive science and HCI conferences.
For more information about the methods we use, please look at Pair Analytics, Visualization Literacy, Personal Equation of Interaction, and Distributed Collaborative Analytics in the methods section of the menu.
For more information about specific applications, please look at the projects that can be found under Health analytics, Aircraft Safety, Emergency management, and Public Safety submenu.