Computational models of human decision-marking

A fundamental question in the study of perception is how information from multiple stimulus dimensions is processed in basic tasks such as perceptual learning, detection, recognition, visual search, and categorization. For example, a baggage screener at the airport will look for multiple specific features in order to classify whether an object is dangerous. A key question is how the screener combines all information to identify a dangerous object. A decision for each feature can be made either in a serial order or simultaneously in parallel. Understanding the architecture of perceptual processing provides a critical link between perception and decision-making. I applied sequential sampling models of response time (RT) to identify processing architecture. These innovative models describe the time course of decision-making and permit strong tests of competing theories that might otherwise be difficult to distinguish using conventional measures like mean accuracy and RT. I helped to develop a computational framework that combines decision-making with models of categorization and mental architecture to describe how people integrate multiple sources of perceptual information Example 1, Example 2. Importantly, this framework provides theoretical and empirical definitions to differentiate serial from parallel processing. I have been used this framework to challenge long-held assumptions about perceptual processing.