EMILY E. DAVIS, Ph.D.
Postdoctoral Fellow, University of TorontoDepartment of Psychology
ABOUT
I am a postdoctoral fellow at the University of Toronto under the supervision of Dr. Morgan Barense. I previously completed my PhD at Brock University under the supervision of Dr. Karen Campbell.By using both behavioural (including eye-tracking) and neuroimaging techniques, I study the role of attentional control in shaping how and what we remember. In particular, I study how these processes develop in children and change in aging. I am also interested in how atypical neurodevelopement or other health disorders may affect these processes.In my spare time: I like to hike, paddle board, cook, and chase my toddler around.I'm on the academic job market looking for tenure track professorships
My research program investigates how attentional control shapes memory across development, aging, and in individuals with autism. Across these domains, I test the idea that apparent memory differences arise not only from memory mechanisms themselves, but from how information is selected and integrated over time. To address this, I combine behavioral paradigms, naturalistic stimuli, and neuroimaging methods to better capture cognition in real-world and hard-to-measure populations.
HOW DOES ATTENTION SHAPE LONG-TERM MEMORY ACROSS THE LIFESPAN ?Many previous studies have argued that children and older adults do worse on memory tasks because the memory mechanisms responsible for encoding and retrieving those memories underdeveloped or degraded with age. My research challenges the traditional view that age-related memory decline reflects failures of memory mechanisms alone; instead, I propose that the critical factor lies in how attention affects memory.In this line of work, I explore how attentional factors influence performance on implicit and explicit memory tasks, the extent to which attention results in the erroneous formation of associations between targets and distractors ('hyper-binding'), and how sustained attention at encoding might influence memory for targets and distractors later.
HOW CAN NATUALISTIC STIMULI ADD TO OUR KNOWLEDGE OF MEMORY-ATTENTION INTERACTIONS IN AGING?Much of what we know about memory-attention interactions and their development has been learned through experiments with tightly-controlled stimuli (e.g., carefully crafted word lists). These stimuli are beneficial in that they allow for experimental rigour. However, to advance our understanding of these cognitive processes, there is an advantage to using stimuli that are more ecologically valid such as natural language and/or videos (e.g., movies). This may be particularly important in studying children and older adults given that these stimuli are more attention capturing and may result in better memory performance than what we normally see in typical lab experiments.My research extends traditional laboratory paradigms by asking whether age-related memory differences persist when attention is engaged by naturalistic, meaningful stimuli. In doing so, we've shown using eye-tracking that when using naturalistic stimuli, older adults attend to the same things during movie watching, recall the same number of details and perceive event boundaries at the same time points as young adults.
CAN WE MEASURE LANGUAGE PROCESSING IN THE ABSENCE OF SPEECH IN INDIVIUDALS WITH AUTISM?Individuals with non-speaking autism are often assumed to have limited language comprehension, in part because traditional assessments rely on verbal or motor responses. This creates a critical gap: we lack reliable tools to measure language processing independent of output.In this line of work, I use functional MRI to examine language processing in individuals with non-speaking autism. By leveraging a well-validated auditory language localizer, I have successfully identified canonical language-selective brain networks in non-speaking individuals.These findings provide some of the first neural evidence that language-selective processing can be detected in this population.More broadly, this work asks how we can separate language comprehension from language production, and highlights the potential of neuroimaging to provide more inclusive and accurate assessments of cognition.
CAN WE LEVERAGE MARKERLESS PROSPECTIVE MOTION CORRECTION IN FUNCTIONAL MAGNETIC RESONANCE IMAGING RESEARCH FOR USE IN SPECIAL POPULATIONS?In parallel, I investigate whether markerless prospective motion correction (PMC) can reduce one of the largest barriers in neuroimaging: head motion. This is especially critical for studying populations such as children and individuals with autism.Moving beyond proof-of-concept, I evaluate whether these systems are suitable for large-scale psychological research, examining both structural and functional imaging. To better understand the limits of PMC, I integrate convolutional neural networks to detect subtle facial movements (e.g., eye blinks, facial shifts) during scanning and relate these signals to motion correction performance.This work aims to identify when PMC succeeds, when it fails, and how it can be refined for real-world research applications.
HOW DOES A NEW FACE BECOME FAMILIAR?We recognize people we know quickly and efficiently, but struggle to recognize people that we don't know. The question that drives this work is what are the underlying mechanisms of face learning. In this line of work, I've explored one potential mechanism: ensemble coding (the rapid extraction of summary statistics). The visual system is remarkably efficient: it can rapidly extract the “average” of a group of faces, forming an ensemble representation of identity. This raises a key question: Can these summary representations support learning new individuals?My work tests this directly by examining how ensemble coding interacts with face identity learning. Across multiple experiments with both young adults and children, I find that although ensemble representations of faces are robust, they are not sufficient for learning new identities. Instead, successful learning depends on access to individual exemplars, suggesting that ensemble and exemplar representations are not independent, but fundamentally intertwined.This work challenges the idea that the brain can rely on compressed, summary representations for identity learning, and instead points to a system that must balance generalization across variability with precision at the level of the individual.
