Snowball sampling (also known as chain-referral sampling) is a non-probability (non-random) sampling method used when characteristics to be possessed by samples are rare and difficult to find. For example, if you are studying the level of customer satisfaction among elite Nirvana Bali Golf Club in Bali, you will find it increasingly difficult to find primary data sources unless a member is willing to provide you with contacts of other members.
As the sample builds up, enough data are gathered to be useful for research. As sample members are not selected from a sampling frame, snowball samples are subject to numerous biases. For example, people who have many friends are more likely to be recruited into the sample.
If the topic of the research is not sensitive or personal, it may be acceptable for subjects to provide researchers with names and contact information for people who might be interested in participation. If the topic is sensitive or personal, snowball sampling may be justified, but care should be taken to ensure that the potential subjects' privacy is not violated. For example, studies of networks of drug users or studies tracking sex partners require extreme caution with information gathered from one subject about another.
- Establish a contact with one or two initial cases from the sampling frame. This stage is usually the most difficult one.
- Request the initial cases to identify more cases
- Ask new cases to identify further cases (and so on)
- Stop when:
- a) Your pre-specified sample size has been completed;
- b) There are no further cases left;
- c) Pursuing further cases will make the project unmanageable due to the large size.
- It is possible for the surveyors to include people in the survey that they would not have known.
- There are no lists or other obvious sources for locating members of the population (e.g. the homeless, users of illegal drugs).
- As subjects are used to locate the hidden population, the researcher invest less money and time in sampling.
- Snowball sampling method does not require complex planning and the staffing required is considerably smaller in comparison to other.
- Oversampling a particular network of peers can lead to bias
- Respondents may be hesitant to provide names of peers and asking them to do so may raise ethical concerns
- There is no guarantee about the representativeness of samples. It is not possible to determine the actual pattern of distribution of population.
- It is not possible to determine the sampling error and make statistical inferences from the sample to the population due to the absence of random selection of samples