Enumerating Homeless People
Alternative Strategies and Their Consequences
Published in: Evaluation Review, v. 20, no. 4, Aug. 1996, p. 378-403
Posted on RAND.org on January 01, 1996
Decisions about how inclusive one should be in attempting to sample and/or enumerate homeless populations are best guided by information regarding what is sacrificed when different sampling choices are made. This article draws upon data from the Course of Homeless Study to explore how three progressively less inclusive sampling frames affect understandings of the size and characteristics of homeless populations in two Los Angeles sites. Findings suggest that less inclusive sampling frames substantially affect population estimates, but do not consistently produce biased estimates of population characteristics. Whether, and the extent to which, such bias is introduced varies by site, by, gender, and by the population characteristics in question.