Although attention plays a ubiquitous role in everyday life, we don’t have a straightforward way to measure a person’s overall attentional abilities. Our work suggests that models based on people’s unique functional connectivity “fingerprints” can provide such a measure by predicting how well they pay attention in a variety of contexts. Critically, these connectivity fingerprints are measured as individuals simply rest in an MRI scanner, meaning that we can predict attentional abilities from data collected while people are not performing any particular task at all. We also use predictive modeling methods to inform taxonomies of attention and characterize interactions between attention and other processes such as reading comprehension and memory.

Our ability to focus varies over time, fluctuating from one minute to the next and changing across the lifespan. We use behavioral and functional MRI experiments to characterize moment-to-moment fluctuations in focus during task performance and real-world experiences such as movie watching and story listening as well as changes in attention over days, weeks, months, and years. We are particularly interested in how attention and cognition differ between individuals, change across childhood and adolescence, and vary with people’s experiences and environments. 

Our attention doesn’t fluctuate in a vacuum. Rather, changes in the degree to which we are focused, mind wandering, or distracted can impact learning and memory. Ongoing work in the lab characterizes the consequences of attentional state changes for other cognitive processes such as memory and learning. Our collaborative research also suggests that people better remember images that they see when they’re focused—regardless of whether those images were memorable or forgettable to begin with.

Although a person’s pattern of functional brain connectivity is as unique as a fingerprint, this pattern is far more informative about their behavior. Our lab collaborates to discover what we can learn about a person from their unique brain connectivity pattern, from their personality traits to their creative thinking ability to their clinical symptoms. In this work, we ask what predictive models can tell us about the mind and brain—both when they successfully generalize to predict behavior and when they fail to do so. We also work with a team that makes resources for building connectome-based predictive models available to the scientific community. Recently, we’ve started to explore what different types of brain features, including inter-subject correlation, connectome stability, typicality, and optimality, and edge fluctuations can tell us about different aspects of behavior.