18 Snowball Sampling Examples (Plus Strengths & Weaknesses)

Snowball sampling is a type of non-probability sampling method in which the new participants for the study are recruited with the help of current participants in the study.

The sample group expands like a rolling snowball; hence the name “snowball” is used for this sampling method.

It can be thought of as chain referral as the researcher initially choose a handful of respondents as primary data sources and use them to discover similar respondents.

Snowball Sampling Definition and Uses

Non-probability snowball sampling refers to the practice of gathering a sample based on referrals from other sample participants. This is opposed to probability sampling where the sample is chosen at random, meaning that not every member of the population has an equal chance of being chosen for the study.

This approach is best suited for cases where it is difficult to get in contact with respondents through traditional sampling methods due to reasons like the rarity of the participants, the sensitivity of the topic, or insufficient information on the characteristics of the desired population.

Consider the scenario where you wish to get responses from cancer patients who have an unusual form of the disease: you cannot simply stroll into a hospital and ask for patients’ contact information or medical data making traditional sampling strategies impossible.

What you can do, though, is issue a call to action to meet with one or two individuals who have the ailment, and then request that they recommend you to more possible study participants who would be interested in taking part. Patients can continue to be referred in a chain until your sample frame is large enough.

Examples of Snowball Sampling

1. Studying Underground Art Movements: To understand the dynamics of underground or countercultural art scenes, researchers might rely on initial contacts within the community to introduce them to other key players.
2. Exploring Secret Societies: Investigating the rituals and beliefs of secret societies can be challenging, but with a trusted insider’s introduction, researchers can gain deeper insights.
3. Understanding Informal Economies: In cities where informal economies thrive, researchers might start with a few known street vendors or artisans and then branch out to understand the broader network.
4. Research on Stigmatized Professions: For studies on professions that face societal stigma, like sex work, initial contacts can help researchers connect with others in the community to gather comprehensive data.
5. Post-Conflict Zones: In areas recovering from conflict, trust is often low. Researchers studying post-conflict rehabilitation might rely on local community leaders to introduce them to affected families.
6. Exploring Digital Subcultures: When diving into niche online communities or subreddits, initial members can guide researchers to other significant contributors or moderators.
7. Studying Nomadic Tribes: For tribes that move frequently and have limited contact with the outside world, one tribe member’s introduction can be crucial for researchers to understand their way of life.
8. Investigating Underground Music Scenes: In cities with vibrant but hidden music scenes, initial contacts within the community can introduce researchers to artists, producers, and fans, offering a holistic view.
9. When sample groups live in the shadows: Estimating the size of the problem of trafficking of human beings (Zhang and Larsen, 2021). The researchers used referrals to gain access to sample participants who would usually remain hidden from view and unobtainable.
10. Social network referrals: Researching non‐heterosexual women using social networks – in this study, social networks were used to find otherwise unobtainable subjects (Browne, 2003).
11. Using immigrant social connections: Identifying Argentinian immigrant entrepreneurs in Spain by leveraging the close-knit social interactions of expat groups because the groups were administratively ‘off the map’ (Baltar & Brunet, 2011)
12. Sensitive topics: A study on the prevalence of discrimination towards immigrant students by professors in a university may use snowball sampling due to the sensitive nature of the study.
13. Access to elite communities: If you are studying the level of customer satisfaction among elite Private club members, you will find it increasingly difficult to find primary data sources unless a member is willing to provide you with contacts of other members.
14. Rare subject matter: A study conducted on individuals with rare diseases. If researchers are conducting a study of individuals with rare diseases, it may be difficult to find these individuals through a traditional approach.
15. Populations without contact details: It may be difficult to obtain a list of homeless individuals in a city. However, researchers could find a few homeless individuals and then ask them to recruit more individuals they know who are homeless to be involved in the study.
16. Qualitative studies of close-knit populations: Handcock & Gile (2001) argue that snowball sampling is beneficial for conducting in-depth and thick research on close-knit social groups, where this sampling method can help reveal details of social connections between participant members.
17. Using sensitive data: Researchers are often unable to obtain information about people’s sensitive data, such as health records, so instead, they rely on social networks to find people.
18. Ethnography: Often, ethnographic research needs to rely on snowball sampling because it’s purpose is to uncover social relationships within populations.

Detailed Examples

1. A study on the psychological well-being of ex-convicts.

A researcher trying to evaluate the psychological impact of prison on ex-convicts may decide to use the snowball sampling technique.

It will be difficult to get detailed data on the psychology of ex-convicts as the legal system may not allow providing the sensitive information to the researchers. Furthermore, not every ex-convict will be willing to share personal information making it hard to implement a random sampling/probability sampling approach such as stratified sampling.

But if researchers can find just a few ex-convicts to be in the study, they could ask each of them to recruit additional people they may know who are also ex-convicts.

2. Changes in Romantic relationships over time

A psychology student wants to determine if there is any pattern in behavioral changes among couples in a long-term romantic relationship over time.

As the student is new to the city, they are finding it hard to gather enough participants for a relevant study. So, the student decides to go with the snowball sampling method where they first try to find couples within or through their friend circle.

The length of the relationship, its current status, and the location of the couple can all be some of the selection factors for the study.

After the initial interview, the interviewer might inquire whether the couple knows other couples who might be interested in participating in the study and if they do refer a couple that meets the selection criteria, the interviewer gets their next respondents. After speaking with the second couple, they request a recommendation once more. Depending on the desired sample size, for example, 20-30 couples, the process continues until the sample size is adequate.

3. Snowball sampling for research on non-heterosexual women.

Research conducted by Kath Browne using social media used the snowball sampling technique to recruit non-heterosexual women for focus group interviews to discuss their opinions and experiences.

A total of 28 women were involved in this study, all of whom were recruited using snowball sampling. Snowball sampling employed the researcher’s personal networks as 15 of the 28 women were recruited directly by the researcher using social media while the other 13 were recruited by the participants already involved in the study.

In the research, seven out of 28 women spoke of deliberately transgressing heterosexual norms. The remaining 21 spoke of negotiating heterosexual codes so they did not enact their sexual identities in particular spaces and at specific times.

The researcher found that being rooted in social networks was significant because participants were able to check the researcher both as a researcher and a person. Furthermore, they concluded that using friendship networks in these instances included word-of-mouth assurances which are significant when the research is of a sensitive nature and when participants are wary about revealing details of their personal lives to strangers.

4. Estimating the Size of the People Trafficking Problem

Zhang and Larsen tried to find the best approach to estimate the actual size of trafficking problem from limited data that can be obtained through snowball sampling.

One major obstacle in obtaining valid estimates of trafficking is the general lack of consistent and uniform measures that researchers can use for data collection purposes.

To complicate the issue further, the hidden nature of trafficking of humans makes it difficult to apply conventional probability-based sampling strategies, without which for reference purposes one cannot easily assess the merits of alternative estimation techniques.

This special issue represents the most recent development and applications of one particular method, the multiple systems estimation (MSE) method.

Although the researchers remain biased toward primary data for prevalence estimation, they claim that MSE represents a cost-effective alternative for the purposes of advocacy, policymaking, and victim services.

5. Study to identify Argentinean immigrant entrepreneurs in Spain

Researchers Baltar & Brunet designed a virtual method using Facebook to identify Argentinean immigrant entrepreneurs in Spain (214 cases).

Due to some of them having dual citizenship (non-EU and EU nationality), this community is administratively invisible in national statistics as Argentineans

They were able to observe that the number of instances recognized by Facebook and the virtual response rate is higher than the conventional snowball technique (contact through institutions).

The researchers attributed this increased success to a rise in people’s confidence levels as a result of the researcher sharing personal information on Facebook and engaging in groups relevant to their interests

Strengths of Snowball Sampling

• Researching populations that you would not otherwise have access to is made possible via snowball sampling.
• Because they are afraid of being exposed, members of stigmatized groups, such as homeless individuals, may be reluctant to take part in a study. Since the participants send people they know and trust to the researcher, snowball sampling is helpful in such cases.
• Snowball sampling is simple to use and inexpensive.
• As the initial subjects serve as the recruiters who bring in new subjects, snowball sampling eliminates the need for a research team to hire recruiters for the study.

Weaknesses of Snowball Sampling

• The sample is not representative of the entire population under study because it was not picked at random.
• Reaching your sample may be challenging if you rely on recommendations. People might not want to work with you, be reluctant to expose their identities or have a general mistrust of researchers.
• There is no assurance that the study’s sample will be a representative sample of the whole population.
• Bias in sampling is likely to happen. The fact that initial subjects recruit more subjects increases the likelihood that many of the subsequent subjects may have qualities or characteristics in common that may not be typical of the larger population being studied.

Conclusion

Because snowball sampling is often used to recruit individuals who don’t want to be identified or known, the topic of the research is usually sensitive and personal.

For this reason, researchers must be extra careful to protect the private information of the individuals in the study so that their contact details and information isn’t leaked. The researchers should inform existing subjects and potential future subjects that all of their private information will be kept safe.

Furthermore, since the sample is likely to be biased, it can be hard to draw conclusions about the larger population with any confidence. For this reason, snowball sampling is often used as part of the exploratory analysis – when researchers are simply interested in gaining a better understanding of a certain population and potentially uncovering information that they weren’t aware of.

References

Baltar, Fabiola y Brunet Icart, Ignasi (2011). The use of Facebook in social research. The virtual snowball method applied to the study of immigrants entrepreneurs in Spain. Comunicación presentada en 5 International Technology, Education and Development Conference (INTED), Valencia [ESP], March 7-9, 2011. ISBN 978-84-614-7423-3.

Browne, K. (2005). Snowball sampling: using social networks to research non‐heterosexual women. International journal of social research methodology, 8(1), 47-60. Doi: https://www.tandfonline.com/doi/abs/10.1080/1364557032000081663

Handcock, M. S., & Gile, K. J. (2011). Comment: On the concept of snowball sampling. Sociological methodology, 41(1), 367-371.

Zhang, S. X., & Larsen, J. J. (2021). Estimating the Size of the Human Trafficking Problem: MSE and Other Strategies. Crime & Delinquency, 67(13-14), 2169–2187. doi:10.1177/00111287211029856