Nathan Szeto ’23

University of Michigan, School of Public Health, Ann Arbor, MI

This summer I worked with the University of Michigan School of Public Health’s (UMSPH) Department of Biostatistics in Ann Arbor. UMSPH performs research in topics relevant to the public health of the United States and globally. The school’s key values are compassion, innovation, impact and inclusion. It works toward health equity for all people as well as greater inclusion in research, education and service. It also believes in improving the health of people around the world through equitable partnerships with individuals, communities and practitioners. More specifically, I worked with Professor Peter Song and one of his students, Margaret Banker.

Nathan Szeto with the internship advisor.My work mostly involved sleep, physical activity and data about aging. My primary project involved building an R package that implements a proprietary algorithm that estimates sleep duration based on wearable accelerometer data, specifically ActiGraph data. This package is in the process of being made publicly available via the Comprehensive R Archive Network (CRAN) for use by other researchers needing to measure sleep duration. I learned aspects of the development process, such as managing package dependencies and good R programming practices. This algorithm will potentially prove to be an improvement over existing algorithms for measuring such data.

My other project was surveying and reporting on biomedical databases like NHANES and the UK Biobank for the purposes of informing future research projects. As part of this assignment, I compiled a local database containing data of interest. I was also tasked with communicating with database management teams to determine data availability and policies on appropriate usage of their databases.

From a professional standpoint, my internship experience has helped me realize that I am interested in a career in research and that I would like to pursue graduate school post-graduation. I felt that the work I did had more potential to help others than the work I had performed previously. I was also able to establish relationships with practitioners in my field of interest through interactions with colleagues and outreach to Williams alumni. Furthermore, I realized that I enjoy working in cooperative environments where my performance is not compared to that of my colleagues. In this type of environment, I am more willing to admit my mistakes and share my ideas.

From an academic standpoint, I came to better understand the value of my education. For example, I was able to quickly learn R because of previous computer science courses I have taken at Williams. I also was able to adapt quickly to an unfamiliar field of study because I was used to doing so through my previous coursework. With that said, I also realized that my background in statistics is relatively lacking, and going into the next semester I will need to take advanced statistics courses to ensure that I am prepared for graduate school.

I would like to conclude by thanking the alumni who support the Alumni Sponsored Internship Program and the ’68 Center for Career Exploration for their logistical and professional support.