A vaping survey recruits through peers: follow the sampling chain
Separate recruitment waves, observed responses and assumption-dependent population estimates in respondent-driven sampling.
No Vape · Updated

Locate the procedure behind the name
Find how the first participants entered and how later people were invited. The term respondent-driven is sometimes used loosely; retain the actual procedure rather than supplying missing coupon limits, network questions or statistical adjustments. The label alone cannot establish a complete sampling design. The public methods are the starting point, not the promotional description of a large sample.
CDC’s NHBS documentation describes seeds and later peer recruitment waves in its own surveillance program. It is a primary example of the design’s recruitment structure, not a vaping dataset or a US quit-service recommendation. Columbia’s methods resource additionally describes assumption-dependent estimation and debates about generalizability. No numerical findings from either source are imported here.
US Centers for Disease Control and Prevention: National HIV Behavioral Surveillance (NHBS): How information is collected · Columbia University Mailman School of Public Health: Respondent-Driven Sampling
Draw the chain without identifying people
A small abstract diagram can show initial recruiters, later recruitment waves and the final analyzed responses. Use the report’s labels and counts if supplied. Do not try to identify recruiters or reconstruct private friendship networks. A recruitment wave records a place in the sampling chain; it is not a follow-up visit or evidence of continued quitting.
In a fictional illustration, several initial participants each pass permitted invitations to acquaintances, who may recruit another wave. This can reach people beyond an ordinary public advertisement. It can also leave disconnected adults outside the observed network. The invented chain demonstrates the recruitment route without claiming any actual survey is representative or proposing an estimator.
Separate raw responses from estimation
If the report presents a population estimate, locate the target and the stated analysis. Network size, recruitment patterns and other assumptions can matter. You do not need to calculate specialized weights to preserve that dependence. Record which estimate is reported and which methodological limits accompany it instead of treating a weighted label as proof that all recruitment differences disappeared.
Likewise, do not call the share within the recruited sample a population prevalence simply because peers found the respondents. Participants are actual people observed; a modeled or adjusted estimate answers a further question under a method. Keep raw counts available when supplied, but do not invent missing counts by reversing rounded percentages.
Write a conclusion that retains coverage
Your reading note can say who the survey sought, how peer recruitment operated and which group the authors intend their estimate to describe. Include unresolved network coverage or estimation details. A useful method question is more specific than either “all surveys are biased” or “friends recruited friends, so everybody is represented.”
The design does not make the respondents’ experiences unimportant. It limits which broader claims the collected accounts and analysis can support. Do not turn a survey proportion into your personal quitting probability or treat recruitment incentives as a support service. If the estimate matters for a clinical decision, obtain appropriate interpretation of the actual study and the wider evidence.
Sources and scope
Sources inform the guidance. The worksheets and examples are No Vape editorial suggestions, not tested treatments.
Written with AI assistance and checked against the cited sources. No clinical review is claimed. This guide covers practical planning, not individual medical treatment. How we research and write