Sunday, September 15, 2013

Dr. David Dai Gives Talk at RPI

On Wednesday September 11, 2013 our department’s Dr. Dai gave a talk as part of the Cognitive Science Colloquium at RPI entitled “When Expert Knowledge Is Not Enough:  From Low Roads to High Roads.”  He talked about the nature of expertise and how the current expert-novice comparison paradigm has limitations.  Instead of viewing expertise as between subjects (i.e. what novices don't have vs. what experts do have) we should think of it also as within subjects (i.e. how experts act in different situations).  Expertise isn't static.  Dr. Dai argues that in order to become a "super-expert" one must be adaptive.  The flexibility of adaptive expertise provides a conceptual advantage due to a variety of exogenous and endogenous factors.  For example, people are faced with a continuum of problems ranging from high regularities to high irregularities, high probability and low probability events, and time pressure--exogenous factors.  They also are affected by the set effect (once something is established it frames our thinking), foreclosure of a problem space (you think there's only 1 pathway when there really are multiple), and bounded rationality and cognitive uncertainty--all endogenous factors.  Adaptive expertise is "slow" and based on controlled processes, self-correction and optimization.  In comparison, routine expertise is "fast," based on well-structuredness of a domain and proceduralized knowledge.  We must find a balance in managing this efficiency (routine expertise)  vs. flexibility (adaptive expertise) trade off.  Dr. Dai then discussed several research ideas on how to use the game of "Go" as a venue for studying adaptive expertise based on previous research by Chase and Simon (1973), Reitman (1976) and Hu (2011).  The proposed designs included priming patterns, eye-movement tracking and think alouds, within subjects (simple vs. complex, time limited vs. time unlimited) and between subjects (skill level) comparisons, and presenting a series of typical and "atypical" positions while measuring event-related potentials and duration of deliberation to detect anomalies and find out what makes experts pause and think.  By studying how people play and make decisions in this game, we can begin to see how adaptive expertise works.  This line of research may ultimately help us understand the nature of expertise as a situated and dynamic interplay of representation and reasoning, of fast pattern-based recognition and intuitions and slow analytic thinking and meta-level control, of knowledge encapsulation and functional flexibility, and of developed expertise and fluid intelligence.  

No comments: