Pipelines into Biostatistics Annual Symposium

Thursday, July 20th
DFCI, Smith Building, rooms 308/309
Check-in @ 8:30am
Program 9am-4pm

Knashawn Morales – Keynote Speaker
Dept Alum, Sc.D. ’01
Associate Professor of Biostatistics in Biostatistics and Epidemiology at the Hospital of the University of Pennsylvania, Biostatistics and Epidemiology departments

Latent variable models: Applications in aging and obesity prevention studies

Longitudinal studies, common in public health and medical research, are intended to provide information on changes in health or disease states across time. The analysis of longitudinal data can be complicated by a hypothesis of multiple disease processes over time or if there are multiple measures used to quantify the same disease process. Standard modelling approaches make some simplifying assumptions. The most common are to: (1) assume a single disease process; or (2) summarize, combine, or collapse multiple measures so they fit into a traditional longitudinal analysis framework. In both settings, potentially important information about the progression of the disease process is ignored. Latent variable modeling provides a flexible framework to accommodate the complexities of the data and simultaneously utilize all available information.

In this presentation, an introduction to latent variable methods will be provided by way of three examples: 1) examining the stability of preferences for life-sustaining treatment in older adults; 2) examining the course of depression in late life; and 3) exploring weight loss patterns during an obesity prevention study. In addition, Dr. Morales will also share her personal background and career experience.

The Pipelines Into Biostatistics Symposium is the culminating event for the Department’s Summer Program in Biostatistics and Computational Biology. Attendees are current and past summer program participants, faculty, fellows, graduate students, visiting faculty from minority serving institutions, the external advisory board, and members of the Longwood Medical community.