[Respondent-Driven Sampling: a new sampling method to study visible and hidden populations]

Area of Science:

• Social Sciences
• Statistics
• Epidemiology

Context:

• Studying hidden or hard-to-reach populations presents significant sampling challenges.
• Traditional probability sampling methods are often infeasible for populations lacking a sampling frame.
• Network-based sampling offers potential solutions for accessing such groups.

Purpose:

• To introduce and evaluate Respondent-Driven Sampling (RDS) as a method for studying populations without a sampling frame.
• To demonstrate the combination of network analysis with probability sampling principles.
• To assess the utility and limitations of RDS for accessible yet unlisted populations.

Summary:

• Respondent-Driven Sampling (RDS) is presented as an advancement over traditional snowball sampling, integrating network analysis with statistical validity.
• The study applies RDS to a population of young people (14-25) engaging in clubbing, alcohol/drug consumption, and sexual activity in Spain.
• Fieldwork was conducted in Baleares, Galicia, and Comunidad Valenciana between May and July 2007.
• RDS proved useful for populations that are accessible but lack a sampling frame.
• The study acknowledges that the RDS sample is not statistically representative of the target population but rather a 'pseudo-population'.

Impact:

• Highlights the applicability of RDS for specific research scenarios where traditional sampling fails.
• Provides empirical evidence on the strengths and weaknesses of RDS in real-world fieldwork.
• Contributes to the methodological toolkit for social and epidemiological research on vulnerable or hidden groups.
• Informs researchers about the representativeness limitations of RDS, emphasizing the need for careful interpretation of findings.