Snowball sampling: a cautionary tale involving a study of older drug users

Snowball sampling is generally seen as a highly effective sampling technique that allows for the study of difficult to reach or ‘hidden’ populations. It is also seen as a valuable tool for the study of particularly sensitive or private matters. As a result, it was chosen as the sampling method for a research study into illegal drug users over the age of 40 who are not in contact with the criminal justice authorities or treatment agencies regarding their use. However, there proved to be great difficulty in locating and interviewing potential participants using snowball sampling. Four reasons for this are proposed, and it is suggested that snowball sampling is prone to failure in particular circumstances. Thus, caution should be exercised when considering its adoption as a sample-building technique.