My theoretical work focuses on tackling the inevitable people problems of nested clinical trials, where participants are members of intact groups (e.g., communities, workplaces, schools) or placed into groups during the study (e.g., therapy groups, training classes). How do we appropriately account for the correlation within groups when we analyze data from these designs, and how do we handle missing data in these scenarios? I demonstrated the strength of a logistic mixed-effects model for binary outcome data from partially nested designs (manuscript in progress) and developed a predictive mean matching approach for multiple imputation of missing continuous data in cluster randomized trials (manuscript under review).
In practice, I am interested in understanding the effects of stress on health and psychological well-being, and working with experts to develop broadly accessible interventions to mitigate the effect of stress on health. To this end, I spent several years as a collaborating biostatistician in the Stress and Health Lab of the Institute for Behavioral Medicine Research, training directly with Dr. Janice Kiecolt-Glaser, a leader in the field of psychoneuroimmunology. In the realm of cognitive psychology, I am currently working with Dr. Sarah Bunnell and two Amherst College students to investigate meaning-making in the public narrative of childhood sexual abuse.