Dependent relationships are characterised by a disparity in power and status between the researcher and research participant that compromises the ability of the participant to provide free consent to participate in all or part of the project, or withdraw from a project they are already participating in.
This typically occurs when participants feel obliged to participate (or not withdraw), or are likely to have a reasonable belief that they might experience disbenefit, even if relatively minor, by declining to participate or by withdrawing.
Therefore, when planning research and seeking ethical approval, researchers should have the following in place and described in their ethical review application form (Word):
- Proactive reassurances to the participant that they are free to choose to participate and withdraw as they wish without facing any disbenefit, including in the participant information sheet and consent form. For example:
- When inviting students of the University to participate in a research project, it must be made clear that the participation or non-participation in the project, and any subsequent withdrawal, will not affect the student’s academic assessment in any way, and that no information arising from the research will be disclosed to the student’s School, Department or Faculty.
- When inviting users of a health or social care service to participate in a research project, it must be made clear that the participation or non-participation in the project, and any subsequent withdrawal, will not affect the care or treatment they receive. Particular care should be taken if the person recruiting is also the person directly responsible for the participant's care or treatment.
- Any other actions that might be required to mitigate against the effects of dependent relationships in the process of recruiting participants.
Dependent relationships can be particularly impactful when working with vulnerable groups, and researchers should check if they require any additional approvals such as PVG for accessing these populations.