Respondent-Driven Sampling: un nuevo método de muestreo para el estudio de poblaciones visibles y ocultas
Loading...
Identifiers
Publication date
Advisors
Tutors
Editors
Journal Title
Journal ISSN
Volume Title
Publisher
Socidrogalcohol
Abstract
Este artículo presenta una variante de muestreo en cadena:
el respondent-driven sampling (RDS). Este método de muestreo
prueba que las posibilidades ofrecidas por los métodos basados
en análisis de redes se pueden combinar con la validez estadística
de los métodos estándar de muestreo probabilístico. En este
sentido, el RDS se presenta como una mejora matemática del
muestreo por bola de nieve orientada al estudio de poblaciones
ocultas. Sin embargo, aquí tratamos de probar su validez para
ser utilizado con aquellas poblaciones que no están registradas
en marcos muestrales pero que, sin embargo, no ofrecen especiales
dificultades para ser contactadas. En este trabajo explicamos
el funcionamiento básico de RDS a partir de investigaciones
sobre los jóvenes (de entre 14 y 25 años) que salen a divertirse
con frecuencia los fines de semana, consumen alcohol y otras
drogas y tienen relaciones sexuales. La investigación de campo
se realizó entre mayo y julio de 2007 en Baleares, Galicia y
Comunidad Valenciana. La presentación que se hace del estudio
demuestra la utilidad de este tipo de muestreo cuando la población
es accesible pero hay una dificultad que viene generada por
la inexistencia de un marco muestral. No obstante, la muestra
conseguida no es una muestra aleatoria representativa en términos
estadísticos de la población objetivo. Ha de reconocerse
que la muestra final obtenida es representativa de una “pseudopoblación”
que se aproxima, aunque dista de ser idéntica, a la
población objetivo
The paper introduces a variant of chain-referral sampling: respondent-driven sampling (RDS). This sampling method shows that methods based on network analysis can be combined with the statistical validity of standard probability sampling methods. In this sense, RDS appears to be a mathematical improvement of snowball sampling oriented to the study of hidden populations. However, we try to prove its validity with populations that are not within a sampling frame but can nonetheless be contacted without difficulty. The basics of RDS are explained through our research on young people (aged 14 to 25) who go clubbing, consume alcohol and other drugs, and have sex. Fieldwork was carried out between May and July 2007 in three Spanish regions: Baleares, Galicia and Comunidad Valenciana. The presentation of the study shows the utility of this type of sampling when the population is accessible but there is a difficulty deriving from the lack of a sampling frame. However, the sample obtained is not a random representative one in statistical terms of the target population. It must be acknowledged that the final sample is representative of a “pseudo-population” that approximates to the target population but is not identical to it
The paper introduces a variant of chain-referral sampling: respondent-driven sampling (RDS). This sampling method shows that methods based on network analysis can be combined with the statistical validity of standard probability sampling methods. In this sense, RDS appears to be a mathematical improvement of snowball sampling oriented to the study of hidden populations. However, we try to prove its validity with populations that are not within a sampling frame but can nonetheless be contacted without difficulty. The basics of RDS are explained through our research on young people (aged 14 to 25) who go clubbing, consume alcohol and other drugs, and have sex. Fieldwork was carried out between May and July 2007 in three Spanish regions: Baleares, Galicia and Comunidad Valenciana. The presentation of the study shows the utility of this type of sampling when the population is accessible but there is a difficulty deriving from the lack of a sampling frame. However, the sample obtained is not a random representative one in statistical terms of the target population. It must be acknowledged that the final sample is representative of a “pseudo-population” that approximates to the target population but is not identical to it
Description
Bibliographic citation
Mantecón, A., Juan, M., Calafat, A., Becoña, E., & Román, E. (2008). Respondent-Driven Sampling: un nuevo método de muestreo para el estudio de poblaciones visibles y ocultas. Adicciones, 20(2), 161-170
Relation
Has part
Has version
Is based on
Is part of
Is referenced by
Is version of
Requires
Publisher version
http://dx.doi.org/10.20882/adicciones.280Sponsors
Este estudio se ha realizado en parte con financiación de la Delegación del Gobierno para el Plan Nacional sobre Drogas según la Orden de convocatoria SCO/269/2007
Rights
Copyright © 2008 Adicciones








