We've previously explored the overarching changes to Section 4 of the National Statement in the 2025 edition. This deep-dive focuses specifically on Chapter 4.4, which addresses research and evaluation involving people in dependent or unequal relationships. Unlike other chapters, the changes are more subtle but add some useful practical steps to take in designing projects.

Introduction

Chapter 4.4 addresses research and evaluation involving people in dependent or unequal relationships. These are contexts where pre-existing power imbalances could compromise voluntary participation or informed consent. This includes relationships between health care professionals and patients, teachers and students, employers and employees, carers and people with disabilities, prison authorities and prisoners, and service providers and service users. However, it also extends to relationships between evaluator/researcher and participant, depending on the context and nature of the relationship.

The 2025 edition makes subtle but significant changes. As for other parts of Section 4, the automatic requirement for full HREC review has been removed, applying instead the proportionate, risk-based framework from Chapter 2.1. 

What has changed?

The automatic HREC review requirement has been removed

Under the 2023 edition, any research or evaluation involving people in dependent or unequal relationships automatically required full HREC review, regardless of the actual risk level of the research activities.

This is no longer the case.

The 2025 edition applies the risk framework outlined in Chapter 2.1. This means that lower risk research and evaluation involving people in dependent or unequal relationships can now proceed through proportionate review pathways, rather than automatically requiring full HREC review. The determining factor is whether the power imbalance magnifies the risks inherent in the research activities. If it does not, a lower risk review pathway is appropriate.

Acknowledgement of both over- and under-representation

The updated chapter builds on the 2023 statement in recognising that people in dependent or unequal relationships can be both over-represented and under-represented in research and evaluation. This dual reality is articulated in the introduction and elaborated in Guideline 4.4.1. 

Over-representation occurs because researchers/evaluators have relatively easy access to certain populations once they've gained entry to a setting as part of a project. The temptation to recruit from captive audiences is real, and the chapter challenges researchers/evaluators to consider whether they are seeking out people for legitimate research reasons or simply because access is convenient.

Under-representation occurs through gatekeeping by those in positions of power or authority. When researchers/evaluators must navigate institutional hierarchies to reach potential participants, those with authority can block access, ostensibly for protective reasons but often resulting in the exclusion of entire groups from research that could benefit them. As a HREC we are acutely aware of the role we play in this, and it is a key consideration in our decision-making and process design.

Power imbalances are contextual, not categorical

The 2025 edition reinforces that power imbalances may be relevant to a specific aspect of a person's life rather than defining their entire existence. A person may be in an unequal relationship with their employer but not with their doctor. A student may be dependent on their teacher for grades but retain full autonomy about research participation.

This contextual understanding prevents researchers/evaluators from treating people as uniformly "vulnerable" across all domains. The relevant consideration is not whether someone is in any dependent relationship, but whether the specific relationship at issue could compromise their participation in the proposed research or evaluation.

The chapter also acknowledges that power imbalances may be linked with economic, political, social, or cultural factors that generate systemic prejudice, stigmatisation, poverty, illiteracy, or victimisation. This recognises that dependency and inequality operate at multiple levels simultaneously.

Reordered and clarified guidelines

The guidelines (4.4.1 to 4.4.8) have been rewritten to reflect more practical language and a different ordering than the 2023 edition. Previously, the guidelines were aligned with the principles of the National Statement. The 2025 version reorganises them to follow a more logical sequence for researchers/evaluators working through the design, consent, and conduct phases of their projects.

Notably, guideline 4.4.4 now provides specific clarity on what researchers/evaluators should consider in the consent process, including:

• The possibility that potential participants may be compromised in their ability to decline
• Whether the approach is being made by someone the participant depends on for care, services, or other interactions
• Whether there is any direct, indirect, or perceived benefit to those in positions of power arising from participation
• Whether there is potential for participants to be adversely affected by their decision to participate or decline
• Whether there is reason to believe participation is contrary to the participant's best interests

This checklist makes explicit what should be running through a researcher's/evaluator's mind during consent planning, rather than leaving these considerations implicit.

Removal of the term "vulnerable"

Consistent with changes across the 2025 National Statement, the word "vulnerable" has been removed from Chapter 4.4. This reflects a deliberate move away from labelling entire groups as inherently vulnerable and toward recognising that risks of harm and discomfort are situational, contextual, and often created by research and evaluation design rather than located within individuals. 

For example, poorly designed research, inadequate consent processes, failure to provide independence in recruitment, or lack of genuine assurance that declining participation will have no consequences can all increase risks to participants by amplifying power imbalances. The 2025 edition places responsibility where it belongs: on researchers/evaluators and reviewers to identify and mitigate risks, rather than on participants to overcome their supposed vulnerability.

Key provisions for social researchers, market researchers, and evaluators

The consent process requires heightened scrutiny

Guideline 4.4.4 establishes that being in a dependent or unequal relationship may influence or compromise a person's decision to participate in research or evaluation. Importantly, the chapter clarifies that this "does not necessarily invalidate the decision" but "constitutes a reason to pay particular attention to the process through which consent is negotiated."

This is a measured approach. It acknowledges the real risk that consent may be compromised without presuming it inevitably is. The burden is on researchers/evaluators to demonstrate they have considered the specific ways power dynamics could affect the consent process in their project and have implemented appropriate safeguards.

The chapter is particularly concerned with situations where the approach to potential participants is made by someone they depend on for care, daily services, or other interactions. This is the clearest route to compromised consent, and researchers/evaluators must address this directly in their design.

Conflicts of interest can be managed through independent support 

Guideline 4.4.7 states that where the researcher/evaluator has a pre-existing relationship with potential participants that could be considered a potential conflict of interest, it may be appropriate for an independent person to engage in discussions about participation and/or obtain consent.

The use of "may be appropriate" rather than "must" is deliberate. Not all pre-existing relationships create conflicts of interest severe enough to require independent consent processes. For example, a teacher conducting research with their current students about their learning experiences, a doctor conducting a clinical trial with their patients, or an employer evaluating workplace satisfaction among their employees (a topic we'll be covering soon) would have conflicts of interest that require a means of independent support to be applied, such as a third party acting as a mediator between researcher and participant. 

By contrast, a teacher conducting research with former students years after graduation, or a healthcare professional conducting research with patients from a different service, may not require the same level of independent support. The determining factor is whether the relationship could reasonably be perceived to compromise voluntary participation.

Researchers/evaluators should err on the side of independence when in doubt, and reviewers should be prepared to require it when the application does not adequately justify direct recruitment by someone in a position of power.

No negative consequences for declining or withdrawing

Guidelines 4.4.7 and its cross-references to Guidelines 2.2.19 and 2.2.20 establish unequivocally that participants who decline to participate or decide to withdraw must not suffer any negative consequences. This includes unfair discrimination, reduction in level of care, dismissal from employment, or any other disadvantage.

This protection must be more than a statement in a participant information sheet. Researchers/evaluators must actively design their projects to ensure that decisions about participation are not visible to those in positions of power (where possible), that withdrawal can occur without explanation, and that there are no mechanisms through which non-participation could be detected or punished.

In evaluation contexts, this is particularly challenging. If an evaluator is assessing a service and service users decline to participate, service providers may notice low participation rates and draw conclusions about individual decisions. Researchers/evaluators must consider how to structure recruitment and data collection to protect individual decisions while still achieving evaluation objectives.

Privacy and confidentiality in shared spaces

Guideline 4.4.8 requires researchers/evaluators to respect privacy and ensure confidentiality in settings such as workplaces, hospital rooms, or residential care facilities. These are environments where privacy is already compromised, and research or evaluation activities can further erode it.

This means:

• Not conducting interviews or data collection in spaces where others can overhear
• Not leaving research/evaluation materials visible to staff or other residents/patients/colleagues
• Ensuring where possible that participation status is not disclosed to those in authority
• Protecting data in ways that prevent institutional access

This last point is crucial. If research or evaluation data is stored on institutional servers, accessed through institutional accounts, or otherwise within the reach of those in positions of power, confidentiality may be compromised.

The evaluation challenge: positively biased responses

While Chapter 4.4 applies to both research and evaluation, evaluation contexts present a particular challenge that deserves explicit attention. Evaluation is fundamentally about assessing whether something works, whether it should continue, and how it might be improved. The stakes of these questions create powerful incentives for participants to provide positively biased information.

Service users may believe that negative feedback about a service will result in that service being reduced or withdrawn, leaving them worse off. They may be correct in this belief. Program managers may believe that poor evaluation results will lead to program termination, loss of funding, or reputational damage. They may also be correct.

These are not irrational concerns, and evaluators cannot simply reassure participants that their honest responses will have no consequences. In many cases, the entire point of the evaluation is that responses will have consequences, ideally, improvements to the service or program.

What evaluators and program/service managers can do to manage this is:

1. Separate evaluation from service delivery. Where possible (and proportionate), evaluators should not be involved in delivering the service being evaluated, and participants should understand this separation.
2. Aggregate and anonymise data. Individual responses should be protected through aggregation wherever feasible, so that service providers or program managers cannot identify who said what.
3. Clarify how findings will be used. Participants should understand the specific decisions that will be informed by the evaluation and have realistic expectations about potential consequences.
4. Use independent evaluators. External evaluators without ongoing relationships with participants or vested interests in evaluation outcomes are more likely to receive honest responses.
5. Create safe channels for negative feedback. This might include anonymous surveys, independent interviewers, or mechanisms for participants to provide feedback outside the formal evaluation process.

Even with these safeguards, researchers/evaluators should acknowledge the inherent limitation: evaluation in contexts of dependency will always be affected by power dynamics, and participants may reasonably choose to moderate their responses to protect their interests. This is not a failure of ethics design; it is a reality of evaluating services and programs that people depend on.

Practical questions for researchers and evaluators

When designing research or evaluation that involves people in dependent or unequal relationships, consider:

Recruitment and access

• Why are we recruiting from this population? Is it because they are the appropriate participants for our research/evaluation question, or because they are easily accessible?
• Are we contributing to over-representation of this group, or are we addressing historical under-representation?
• Who will make the initial approach to potential participants, and does that person have power or authority over them?
• If we require institutional gatekeepers to access participants, how will we ensure gatekeeping does not result in exclusion of particular individuals or subgroups?

Power dynamics and consent

• What is the nature of the pre-existing relationship between researchers/evaluators and potential participants?
• Is this relationship relevant to all aspects of participants' lives or only to specific domains?
• Could the relationship be perceived as a conflict of interest requiring independent involvement in recruitment or consent processes?
• What direct, indirect, or perceived benefits might accrue to those in positions of power from participant involvement?
• Have we considered whether potential participants might have unrealistic expectations of benefits from participation?

Consent process design

• How will we ensure participants understand that declining or withdrawing will have no negative consequences?
• Can we demonstrate that non-participation is genuinely consequence-free, or are we simply stating it?
• Should we invite potential participants to discuss their participation with someone who can support them in decision-making?
• Should we appoint a participant advocate to safeguard participants' interests?
• In evaluation contexts, have we addressed the specific ways power dynamics might bias responses?

Privacy and confidentiality

• Where will recruitment, consent, and data collection occur?
• Can others (including those in positions of power) observe or overhear these activities?
• How will we protect participation status from disclosure to institutional authorities?
• Where will data be stored, and who has access to storage systems?
• Have we ensured that institutional authorities cannot access identifiable data?

Minimising detrimental effects

• What are the potential detrimental effects of the dependent or unequal relationship on the conduct of the research or evaluation?
• What are the potential detrimental effects of the research or evaluation on participants involved in that relationship?
• What steps have we taken to minimise, mitigate, or manage these effects?

Ethics review pathway

• Given the nature and risk level of our research or evaluation, what is the appropriate ethics review pathway?
• Does the power imbalance magnify the risks inherent in our research activities?
• If we are advocating for a lower risk review pathway, what is our justification?

While providing explicit answers to these questions is not a requirement in ethics applications, researchers/evaluators should ensure these considerations are clearly addressed in their application materials, such as project plans, consent forms, and participant information sheets.

Summary

Chapter 4.4 represents a shift toward proportionate, risk-based approaches to research and evaluation involving people in dependent or unequal relationships. The removal of automatic full HREC review requirements reflects recognition that power imbalances do not uniformly compromise consent or magnify risk, and that blanket requirements can inadvertently contribute to the under-representation of important populations in research.

What has not changed is the fundamental ethical responsibility to ensure participation is voluntary, informed, and free from coercion or undue influence. For researchers/evaluators working with people in dependent or unequal relationships, this means thinking about power, being honest about potential harms, and being willing to relinquish direct control over recruitment and consent when conflicts of interest are present.

The message is clear: dependent relationships require careful attention, not automatic barriers to participation.

AI Disclosure:  Initial drafts of the content for this article 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.

​

​