[Page 3213]The ideal sampling technique in research is probability sampling, which involves random selection of respondents from a population of interest and permits generalization of results to the entire population. In many areas of study, however, investigators cannot enumerate the population exactly, and, even if they can, they may not have the resources to randomly sample from it. One practical alternative is snowball sampling, which involves finding individuals associated with the population of interest and then using those people to gain access to others.

Snowball sampling is sometimes referred to as chain-referral sampling, chain sampling, or referral sampling. A variation of snowball sampling is respondent-driven sampling, which employs snowball sampling techniques with statistical modeling that allow for approximation of probability sampling and greater statistical validity. This ...