Research on Homelessness

Abstract

The difficulties surrounding research on homelessness are numerous and substantial. Using a statewide census of homelessness, this article analyzes and critiques common methodological techniques employed in that study. In addition, using cross-tabulation and loglinear modeling, the relationship between key demographic variables and the cited primary cause of homelessness, as well as the types of housing needed, are assessed. Through a multiphased process, which isolates the effects of the operationalized definition of homelessness, significant changes emerge. The operational definition of homelessness is found to affect the demographic composition of the sample, the estimation of annual incidents of homelessness, and the estimation of the statewide number of homeless individuals. The impact of the uncertainty surrounding research on homelessness on policy formation is addressed.