Online

Linnell, Kelsey Meegan

Mathematics and Statistics

2022

Ph. D.

Health surveillance and assessment are considered essential components of a functional public health system. The recent ubiquity of mobile devices and social media have created a wealth of behavioral data, and bring into existence new forms of population health monitoring. These new digital sources can provide direct and passive data for more detailed and nuanced health factors, and have expanded the human, spatial, and temporal scales at which these factors can be measured. In this project, I leverage digital trace data from tweets and mobile device location pings to explore population scale sleep loss, and nature exposure through park visitations in the United States. Both sleep and nature exposure are essential contributors to well-being, and have historically relied on either survey data or direct observation of individuals to measure. I begin by demonstrating the ability of Twitter data to passively reflect population-scale sleep loss at the state level. This is followed by an exploration of park visitation measured through mobile device GPS data. Changes in county-scale park visitation behavior at the onset of the COVID-19 pandemic are analyzed and comparisons are made using population density, employment sector, income, and vot- ing records. In the final chapter I investigate the viability of predicting park visitation using demographic information from the surrounding neighborhood. I conclude with a brief discussion of the significance of measuring these behaviors, and the potential for health policy improvement.