Funded by the Office of Naval Research
Cognitive and Neural Sciences Division
Program Officer: Dr. Astrid Schmidt-Nielsen
Program Number N000140310088
2002-2006, $408,310.00

PI: Dr. Stephanie Doane
Team Members: Randy Brou, Bailey Stephenson, Leah Winter, Sean Walters, Alex Anderson, Jacob McIntosh

Abstract:
Learning to discriminate between visually displayed objects is an important aspect of human performance. The proposed research uses a cognitive theory-driven approach to examine the role of strategic processes in attending to relevant features while visually discriminating between displayed objects. Previous research suggests that, initially, humans acquire processing strategies that reduce the number of redundant comparisons required for accurate discriminations. With practice, the strategy becomes more efficient by elimination of redundant and unnecessary comparisons (e.g., Haider & Frensch, 1999). Previous research results suggest that initially acquired strategies can be transferred to novel stimiluli (e.g., Doane et. a., 1999). Such findings support the hypothesis that initially acquired strategies are born from processing specific stimuli but are not tied to these stimuli per se; they are stimulus independent and as such can be transferred to processing novel stimuli. However, the transfer findings also suggest that strategies are transferred regardless of their effectiveness for discriminating between novel stimuli and that, once acquired, they are difficult to modify (e.g., Doane e al., 1999). Previous research does not address the precise nature of the strategies acquired. Instead, past research has indirectly inferred the nature of the strategies from accuracy, reaction time, and memory data. In addition, previous research has not addressed the factors that facilitate strategic modifications when changes in discrimination context render previously acquired strategies ineffective. The proposed research uses digital eye-tracking technology to provide direct evidence of the nature of the visual processing strategies acquired and transferred, and to explore task contextual factors that influence their modification. Among the factors to be examined are the frequency and duration o fperformance3 feedback during transfer and how the timing of exposure to variable discrimination task contexts influences the malleability of acquired strategic skills. In addition, a computational model of cognition (ADAPT, Doane, et al., 2000a; 2000b) will be used to model individual eye movements and discrimination performance during strategic skill acquisition and transfer. The ability of the model to describe and predict individual eye scans and discrimination performance will be assessed, and used to inform a more precise theory of strategic skill acquisition. The theoretical contribution of this work is to further our understanding of the role of strategic versus stimulus-specific processes in learning to visually discriminate between objects in the environment. The findings from this research may transition into augmented cognition, where a cognitive system can “monitor” eye fixations and performance and provide information that optimizes operator performance.