CAREER: Measuring and Modeling the Multi-Modal Networks and Demographics of People Experiencing Homelessness

点击翻译按钮获取中文摘要

This research develops novel methods for counting the homeless that improve upon existing ways of doing so. Collecting rich data on sheltered and unsheltered homeless populations is difficult and represents a major complication for researchers and decision makers. The methods developed in this project are based on newly available data collection strategies made possible by the high prevalence of online access among homeless persons via free Wi-Fi through smartphones, free computers in libraries, and other programs. This study examines the demographics (age, gender, race/ethnicity) of the online and offline homeless populations and their social relationships to understand their impact on the timing and longevity of homelessness. It contributes to research by improving the estimation of hard-to-reach homeless populations and by increasing understanding of the social support mechanisms that affect the duration of homelessness. Findings will inform decision makers and healthcare leaders about best methods for counting people experiencing homelessness for resource allocation, best practices for disseminating information, and new strategies centered around social support networks.This research concentrates on a US region that contains a large portion of people experiencing homelessness, with targeted samples in a few cities. The study integrates recent developments in survey sampling and estimation to obtain the demographics and social network data from online sources to compare with offline samples, including the Housing and Urban Development Point-in-Time population count. Novel uses of new and old strategies for measuring hard-to-reach populations are employed through four methods: two methods for online sampling (generalized network scale-up methods and network sampling) and two methods for in-person sampling (respondent driven sampling and space-based sampling). These are compared against the current federal standard Point in Time count. The project leverages the resulting data by extending spatial network models -- which have been shown to provide insight into the networks of people experiencing homelessness -- to fully understand the effects of geography and demographics on the social network structure of people experiencing homelessness and its resultant impact on the timing and duration of homelessness. Results of this study provide needed information on the spatial dimension of homelessness and new statistical network methods for providing update-able simulation models for diffusion of information (online and offline) and disease (offline) through homeless communities.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.