Back to Annual Meeting Page		American Public Health Association 133rd Annual Meeting & Exposition December 10-14, 2005 Philadelphia, PA

Abstract #108449

Peter Connery, BA, Applied Survey Research, 55 Brennan Street, Watsonville, CA 95076, 831 728-1356, connery@appliedsurveyresearch.org

Communities applying for HUD funds to provide homeless services are under new federal mandates to thoroughly enumerate their homeless populations every two years to document need, service gaps, and outcomes. However, the procedure to conduct a comprehensive street enumeration is costly. To effectively and efficiently enumerate Los Angeles County's homeless population, Applied Survey Research (ASR) joined with founding director of UCLA Statistical Consulting Center Richard Berk, Ph. D., to develop a U. S. Census tract-based sampling plan from which countywide projections could be extrapolated.

From more than 1,800 relevant U.S. Census tracts in Los Angeles County, it was determined that 500 tracts would be thoroughly canvassed for visual observations of homeless persons and encampments by trained homeless workers who were familiar with homeless populations living within the assigned tract. It was determined that 200 of the selected tracts would represent those tracts within the county know to hold the greatest concentrations of homeless persons, based on service provider knowledge and outreach activities. The remaining 300 tracts were selected at random from a pool of lower-concentration tracts that had been stratified to reflect a countywide geographic distribution.

From the 500 tracts in which an enumeration was conducted, properly weighted (to take account of the two kinds of tracts) projections were made to the county as a whole. Projections to individual tracts or subset of tracts that were not sampled were made by applying information from the tracts sampled at random to other tracts in the same stratum that were not sampled.

Learning Objectives: