We can explain the snowball sampling technique as a technique that is used by the researchers to identify the potential subjects where the possibility of locating the subjects is difficult.
Researchers mostly use this kind of a technique when the sample for study is not very common or is confined to a relatively smaller subgroup of the population. It is more like a chain referral In this kind of sampling, after having finished his observation on one sample, he asks for assistance from the observed sample to help in identifying people who have similar interest. It is pretty much like asking the one subject to nominate the next subject. The process continues like a chain reaction until the target of the sufficient number of subjects is not achieved.
When the subjects want to observe a rare disease, the snowball technique may be a good option by the researcher to identify the sample and get recommendations from the subjects already under study for the next set of referrals. Observing one sample may lead to the next set of samples.
There are different types of snowball sampling. The different types are:
- Linear Snowball Sampling
- Exponential Non-Discriminative Snowball Sampling
- Exponential Discriminative Snowball Sampling
Snowball Sampling comes with its own advantages and disadvantages. The main advantages are that it allows the researcher to reach those populations that are difficult to sample when other sampling methods are being used. To add to it, it is a cheap and cost efficient method as it does not require too much of prior planning and and workforce requirement as compared to other sampling techniques.
Snowball sampling has its own set of disadvantages as well. There is very little or negligible control of the researcher over this kind of method as the new subjects of the study have the dependance on the previous subjects that have been observed. Secondly a generalised representation of the population cannot be assured in this case as the researcher cannot b e sure of the correct distribution of the sample. Another disadvantage is that of the scope of sampling biasad when the existing subjects nominate new subjects thry go on the basis of familiarity and some preconceived thought process. It creates a great possiblity that they would share similar traits and the researcher would only be able to obtain a very small group of the population.
