Skip to Content

Snowball sampling

OverviewThe ethical challengesWhen snowball sampling may be appropriateEthical review requirementsDesigning snowball sampling ethicallyHigher-risk populationsRelated resources

This article is part of the Iris Ethics Knowledge Base; you can find a directory of all articles here.

Overview

Snowball sampling is a recruitment method in which existing participants refer potential participants from their own networks. It is an efficient way to reach hard-to-access populations and can materially reduce recruitment costs, but it carries distinct risks to privacy and voluntary participation that must be addressed before it is used.

Snowball sampling can be consistent with the National Statement, but its use requires justification and careful design; it is not a default or low-scrutiny recruitment choice.

The ethical challenges

Traditional snowball sampling asks current participants to provide contact details, such as names, phone numbers or email addresses, of people in their networks who might be eligible and interested. The research or evaluation team then contacts these referred individuals directly. This raises three related concerns.

  • Privacy and unsolicited contact: referred individuals are unlikely to have consented to being contacted or to having their personal information shared with the research team. They receive unsolicited contact, have had no prior opportunity to decline being approached, and have had their personal information disclosed to a third party without their authorisation. This is particularly significant where the research involves sensitive topics or populations at higher risk of discomfort or harm (NS 3.1.17(c)).
  • Autonomy and voluntary participation: respect for persons extends beyond the moment of consent to include autonomy over whether a person wishes to be approached about research at all (NS 1.10–1.12). When a participant refers someone in their network, the referred individual may feel a social obligation to participate because of the relationship with the referrer. This is relationship-based influence that falls short of overt coercion but can still compromise the freedom of a person's decision (NS 2.2.9).
  • Information asymmetry: referred individuals typically receive contact with little context about how their information was obtained, who provided it, or what was shared. This undermines the trust that supports valid consent.

These concerns are heightened where snowball sampling is used to reach populations who may already be at increased risk of harm through participation, such as people affected by stigma, marginalisation or power imbalances within their community (NS Chapter 4.1). The same factors that make a population hard to reach by other means can make privacy breaches and relationship-based pressure more damaging when they occur.

When snowball sampling may be appropriate

Snowball sampling can be justified where other recruitment methods are not feasible for reaching the target population and the research merit justifies the methodology; where trust-based referral within a close-knit or marginalised community is the only realistic way to recruit; or where legitimate resource constraints make more privacy-protective methods impracticable and the design otherwise incorporates safeguards to protect privacy and voluntary participation. Convenience or cost-effectiveness on its own does not justify the use of snowball sampling where it compromises participant privacy and autonomy; applicants must show that alternatives were properly considered (NS 3.1.12–3.1.13).

Ethical review requirements

Snowball sampling does not meet the criteria for exemption from ethics review. Exemption is only available for research that meets specific conditions, such as use of information from which all personal identifiers have been removed before researchers receive it, or surveys and observation of public behaviour that are highly unlikely to cause distress (NS 5.1.17). Traditional snowball sampling involves the collection of identifiable contact details of people who have not yet consented, which places it outside these criteria; research using personal information without consent cannot be exempted (NS 5.1.16).

Applications proposing snowball sampling should:

  • describe the recruitment methodology fully, including how referrals will be solicited, what information will be collected, and how referred individuals will be contacted;
  • submit all communication materials for review, including scripts for soliciting referrals, initial contact templates, and opt-out procedures (NS 3.1.19; NS 5.2.15);
  • be submitted for full HREC or lower risk review, as appropriate to the overall risk profile (see Lower risk vs full HREC review); and
  • address snowball-specific risks explicitly in the risk assessment, including privacy risks and the potential for relationship-based pressure (NS 2.1.1).

Where snowball sampling is one component of a multi-method recruitment strategy, all recruitment approaches should be included in the scope of review so the committee can assess participant recruitment as a whole. See Recruitment strategies.

Designing snowball sampling ethically

Where snowball sampling is justified, the following design choices reduce the ethical risk:

  • Minimise information collected from referrers: typically a name and one form of contact detail. Additional personal information about the referred individual, or information about the relationship between referrer and referred, is rarely justified.
  • Clear, transparent first contact: tell the referred individual how their contact details were obtained, what information was shared, that participation is voluntary, and how to opt out immediately.
  • Immediate opt-out: provide a simple mechanism to decline further contact, with a commitment to delete contact details promptly and no requirement to explain the decision.
  • Prefer opt-in variants where feasible: a referral-card or information-sharing approach, where current participants pass on study information rather than contact details, or a facilitated-introduction approach, where a participant only shares contact details after the referred person has expressed interest, both preserve the network-based benefit of snowball sampling while substantially reducing privacy intrusion. These approaches may yield fewer referrals but better protect autonomy.

Higher-risk populations

Additional safeguards apply where snowball sampling is proposed for participants who may be at increased risk of harm because of their circumstances (NS Chapter 4.1). Consider whether a referred individual's connection to the research topic could itself pose a risk, such as revealing membership of a stigmatised group; assess whether social hierarchies within the community could create pressure to participate or to share contact details; and ensure the referral approach and communication materials are culturally appropriate. For research involving Aboriginal and Torres Strait Islander peoples, researchers should refer to Special considerations: Aboriginal and Torres Strait Islander peoples (NS 4.7) and ensure that any proposed snowball sampling approach has been discussed with, and endorsed by, relevant communities or community-controlled organisations.

Related resources

This page was last updated on 16 September 2026.

Was this helpful? We'd appreciate your feedback!

This article makes reference to the National Statement on Ethical Conduct in Human Research 2025. It is intended as practical guidance and does not constitute legal advice.

AI Disclosure: Resources in this library were prepared using Large Language Models with input from Iris Ethics staff who guided the scope and design. Subsequent revisions and final versions were developed and approved by Iris Ethics staff.