UC San Diego

Through navigating and interacting with the world, we develop knowledge that can be used to help guide future interactions or process similar information more efficiently. For example, as someone proceeds through medical school and residency to become a radiologist, they develop a deep understanding of medical images. This learned knowledge influences several cognitive processes, which allows them to find abnormalities quickly and efficiently. My dissertation examines factors that may drive how prior knowledge in perceptual domains influence cognitive processes such as memory and perception. I use a range of stimuli and methods that mimic real-world scenarios and require significant perceptual prior knowledge. In particular, I use medical imaging and face perception as two unique but strong indices of perceptual expertise. In chapter 1, I establish that radiologists have enhanced memory for abnormal compared to normal mammograms. I show evidence that this memory benefit appears to be driven by a unique role of distinctiveness, which emerges with significant prior knowledge in the expert’s domain. In chapter 2, I demonstrate that additional knowledge of an abnormality in a medical image not only shapes memory but also enhances experts’ perception of medical images themselves, a form of bias called visual hindsight bias. In chapter 3, I examine the impact of occlusions and feature learning on our ability to extract summary statistics of complex facial information. Overall, I highlight how learned knowledge shapes a range of cognitive processes including memory, visual perception, and ensemble perception, and touch on several avenues for future direction.