Comment: On Respondent-Driven Sampling and Snowball Sampling in Hard-to-Reach Populations and Snowball Sampling Not in Hard-to-Reach Populations

In this commentary attention is drawn to the difference between snowball sampling not in hard-to-reach populations and snowball sampling and respondent-driven sampling in hard-to-reach populations. The approach to sampling design and inference called snowball sampling (not in hard-to-reach populations) was introduced in Coleman (1958– 1959) and Goodman (1961); and respondent-driven sampling in hardto-reach populations was introduced more recently in Heckathorn (1997, 2002, 2007). Still more recently, Gile and Handcock (2010) sounded a cautionary note for the users of respondent-driven sampling in hard-to-reach populations. Coleman (1958–1959) notes that snowball sampling in survey research is amenable to the same scientific procedures as ordinary random sampling, and Goodman (1961) introduces statistical methods with snowball sampling for the estimation