Procedimientos de muestreo representativo de las personas sin hogar (PSH): Avance de un estudio descriptivo en Madrid

Abstract

The present paper is fi-amed within the Homeless Project (Muñoz y Vázquez, 1996). The main aim of this project is the study of economic, psychosocial and health factors afíecting homeless people in Madrid. In this study we show the sampling strategies used to obtain a representative sample of data on homelessness. The sampling strategies included are very innovative in the Spanish context. Besides the sampling method characteristics, we include the main descriptive results on sociodemographic variables.The sampling strategy began by coUecting an exhaustive list of centers for homeless people and a so-called S-Night (i.e., street sampling in a given night) to determine the frequency of use of resources for homeless people. To avoid sampling biases, the sampling methodology also considered the influence of both the size and center type (e.g., shelter vs. soup kitchen) and the day of the week in which the contact was done. The data obtained were statistically weighted as a function of differences betwen the estimated and the empine sample size and as a function as well of the frequency of use of services. Results show descriptive statistics and percentages on the more important sociodemographic variables included in the study. Weighted and unweighted data and their differences are analyzed.