Application of Respondent-Drive Sampling to Evaluate a Youth Street Outreach Program

Background: The United Teen Equality Center (UTEC) in Lowell, Massachusetts provides a safe place of belonging for young people, considered at-risk, ages 13-23. UTEC uses Street Outreach Workers (SOW) to connect with these youth to encourage positive development, provide alternatives to violence and gang-life, and mediate disputes. As part of a larger evaluation of the SOW intervention, we are using an innovative sampling approach to survey youth in the target populations.

Method: This evaluation uses respondent-driven sampling (RDS), a modified form of chain-referral sampling, to identify study subjects. RDS has been used successfully to recruit hard-to-reach populations, but to our knowledge has not been applied to injury. RDS utilizes direct recruitment of peers by peers, recruitment quotas, and a double incentive system. With fidelity to the method, valid inferences about the population can be made from the sample.

Results: After formative assessment, youth from distinct social networks will be selected to recruit up to three of their peers. Youth will receive compensation for each successful recruit and for completing a survey. By using RDS, we expect to obtain surveys from a representative sample, thereby providing generalizable results from the target youth populations on the SOW intervention.

Implications: We explored RDS in the context of violence prevention research and concluded that there are many potential applications of this method to the field of injury prevention and control. RDS can be considered when recruiting individuals for injury studies of a sensitive nature or among hidden populations who may be challenging to recruit.