Online surveys on sensitive topics with non-identifiable (‘anonymous’) populations FAQs

Online surveying is a popular and convenient way of conducting human research on various issues, some of which may explore sensitive personal, psychological and/or social issues.

This is a legitimate and appropriate way to conduct a survey. However, it should also be recognised that the context of online surveying can present particular ethical challenges for researchers and ethics committees reviewing those projects.

These Frequently Asked Questions (FAQs) have been prepared by the RMIT University Human Research Ethics Committee (HREC), to provide guidance and assist RMIT researchers wishing to conduct human research involving surveys on sensitive topics with non-identifiable (‘anonymous’) participants. These FAQs may also apply to surveys conducted through other non-online means.

Human Research Ethics guidelines are provided by the National statement on ethical conduct in human research (National Statement) (PDF 604KB 101p).

These FAQs are intended to complement the National Statement.

What is a sensitive topic?

As has been observed by the National Health and Medical Research Council:

“…it is extremely difficult to establish the categories of information which universally would be considered ‘sensitive’ either because of the nature of the information, the context in which it is handled or the views of the person to whom the information relates.”1

The Privacy Act defines ‘sensitive information’ to mean information or an opinion about an individual’s:

  • racial or ethnic origin;
  • political opinions;
  • membership of a political association;
  • religious beliefs or affiliations;
  • philosophical beliefs;
  • membership of a professional or trade association;
  • membership of a trade union;
  • sexual preferences or practices;
  • criminal record;
  • health (not otherwise defined as ‘health information’); and/or
  • genetic information (not otherwise defined as ‘health information’).

While there is no exhaustive list, ‘sensitive’ topics in human research can include, for example, issues such as sexuality, anxiety, depression, body image, mental health, relationships, self-harm, experience of violence or abuse, homelessness, substance use, etc.

1See submission to the Australian Law Reform Commission.

What risks are associated with human research exploring sensitive issues?

When conducting any human research, it is essential for both researchers and ethics committees to assess risks to participants. In some research, these risks may be negligible (no more than an inconvenience), low (no more than discomfort, for example, physical discomfort in completing a research measure or mild anxiety in completing a questionnaire) or more than low (risk of harm which can be physical, psychological, social, economic or legal).

When the topic of the survey is a sensitive issue, then the inherent risk to participants is increased, as the research is more likely to generate uncomfortable feelings or more severe responses such as distress, anxiety, or relapse in mental illness in participants. The potential impacts of raising sensitive issues with people can be unpredictable as everyone can respond differently and in ways that may not be foreseen or predicted by the researchers, or even by the participants, prior to them undertaking the survey. A survey on sensitive topics may have the potential to cause the participants to discover something about themselves that they weren’t previously aware of or to recall or reflect on a difficult time or event in their lives. Therefore, it is important that the risks of research involving sensitive issues to participants are:

  1. Identified and assessed
  2. Managed and planned for, and,
  3. Mitigated.

What are the risks of using online surveys to conduct this research?

As mentioned above online surveys are convenient for both researchers and participants. When using online surveys for a research project, it is possible to get a greater number of responses and to get those responses sooner than it would be with, for example, face-to-face recruitment and surveying. They are also convenient for respondents as the survey can be completed at any time of the day or night, or on public transport or just as people are preparing for sleep.

However, the context in which online surveys are undertaken mean that it is not possible for researchers to observe or see how participants respond to sensitive survey questions. How will the researcher know if participants have become concerned, anxious or distressed? How are they going to respond to these feelings? How will the researcher know if they are being supported? How can the researcher be sure that they find the right support at the right time? From an ethical point of view, the researcher does have an ethical responsibility towards participants who may find themselves in these situations.

Why do we need to be concerned when participants will be anonymous?

‘Anonymous’ or non-identifiable respondents are those who, as a result of the research design, are not individually identifiable. This can be viewed as a way of protecting participants as, if they cannot be identified by researchers or those interpreting and reading the research, confidentiality is maintained and their privacy is protected. Often in online surveys, anonymity is promoted and may encourage higher rates of participation. An ethical dilemma may arise though if an individual’s responses to an anonymous survey raise concerns for their welfare. If a survey is anonymous, then the potential actions by a researcher are limited, as there is no direct way to contact the participant without breaking confidentiality. For example, what if a participant’s responses to a Depression, Anxiety & Stress Scale (DASS) in a survey indicate concern for that person’s mental health and safety? Certainly, in terms of ethics, the researcher still has responsibilities for the welfare of participants in these circumstances. The researcher is required to implement strategies to assist in mitigating these risks, thus providing a step-by-step guide for the participant group in the specific study. For example, a survey may target young people or older people or males or females only. What sort of strategy is adopted may be influenced by factors such as their demographic profile of the group.

Which participants are regarded as vulnerable?

From an ethical perspective, participants may be assessed as ‘vulnerable’ because of certain personal characteristics and/or due to the particular research design. For example, people with a physical and/or mental disability or illness may be regarded as vulnerable where their ability to make competent decisions may be compromised. Further, where potential participants are known to the researcher due to a pre-existing and/or unequal relationship (i.e. participants are students, friends, colleagues or acquaintances for example), these participants might also be regarded as vulnerable due to the impact this pre-existing relationship may have on the voluntariness of consent. A friend or acquaintance can feel obligated to respond to your invitation to participate when their inclination may be to not participate. Therefore the assessment of vulnerability of participants requires consideration of the individual and/or community circumstances of potential participants, as well as the research context and the potential impacts of the research upon the participants and/or their community. Where research is sensitive or may raise questions that increase the vulnerability of the participants, this raises concerns which need to be managed and mitigated.

How do you manage and mitigate risks of online surveying of vulnerable populations?

All research has a certain level of risk, even if the risk is no more than the risk of inconvenience (for example, time spent completing a survey). The National Statement identifies three categories of risk for human research:

  • Negligible risk: Research where there is no foreseeable risk of harm or discomfort and any foreseeable risk is no more than an inconvenience;
  • Low risk: Research where the only foreseeable risk is one of discomfort; and
  • More than low risk: Research where the risk to participants is more serious than discomfort and there is a risk of harm.

Research that is conducted with vulnerable participants and/or that involves a sensitive topic has a higher level of risk and would arguably be regarded under the National Statement guidelines as being ‘more than low risk’. Careful consideration should be given to the research participants, design, context and foreseeable risks, to inform the classification of the risk level.

No matter what the level of risk, the researcher is responsible for managing and mitigating those risks. The first stage is ensuring that foreseeable risks are identified and a plan or protocol put in place to manage these. It is also critical to ensure that the participants are fully informed and understand the nature of the research and any risks (National Statement, 2.2.2).

What are the Committee’s expectations?

All applications reviewed by an ethics committee are considered on its merits, using the National Statement guidelines as a basis for its decision. It is a key responsibility of the Committee in granting ethics approval to establish that the potential benefits of a project outweigh the likely risks. Further, the greater the risks to participants, the more certain ethics reviewers must be that the risks identified will be well managed and that the participants clearly understand these risks (National Statement, 2.1.8). It is important that all participants are fully aware of any risks in a project. While risks and the management of them are reviewed by the Committee, their expectation is that the researcher will adequately address these matters. Risks also need to be managed and addressed. If your project involves a sensitive topic area, obesity for example, then part of your risk mitigation strategy might be to provide information about relevant support groups or help lines.

What are the researcher’s responsibilities after formal approval of an application?

Once an application is approved, the researcher still has a responsibility to monitor the research and the individual responses of participants. Should participants contact researchers, then a timely and appropriate response is important. But it is also important to monitor the activity related to the research and ensure that ethical standards are maintained. As much research in this category involves the use of social media, it is recommended that related social media traffic is monitored. In the past participants have ‘outed’ themselves as a participant on social media, perhaps not realising the implications of doing so. In this case what is the appropriate response by a researcher? Maybe a discreet private message to the participant is appropriate. In other cases people have used social media to cajole friends, relatives and colleagues to participate, or have utilised websites without appropriate permissions. Researchers need to be careful that such efforts do not contradict the conditions of approval of an ethics application and ensure that all research protocols are followed. Remember that researchers must report immediately any serious or unanticipated adverse effects of their research on participants, and/or any unforeseen events that might affect continued ethical acceptability of the project, to the HREC secretary. If in doubt about any of the ethical aspects, then please contact the Committee that approved your application.

Suggestions for your application

  • In projects such as those that fall into this category of sensitive research with vulnerable populations, it is common to utilise a variety of research instruments like the DASS 21 or 42. Make sure that the use of such instruments is explained and contextualised. For example, the DASS is a self-report questionnaire designed to measure the severity of a range of symptoms common to both depression and anxiety and can be confronting for participants as it provides information about a person’s mental health symptoms. The severity rating of this survey range from “Normal” to “Extremely Severe”. Be sure to explain the purposes and limitations of such tools and include ways to encourage the respondents to seek further advice where they are concerned about their responses or self-scoring.
  • The Participant Information Statement & Consent Form (PICF) template (DOCX 56KB) includes specific reference to online surveys. You should also provide advice to participants about how to print a hard copy or download or screenshot a digital copy of the PICF on their own device. They may wish to take this to their preferred health practitioner should they feel they need to seek assessment or advice.
  • Management of risks is an important responsibility of the researcher. It is important that researchers carefully consider how they provide participants with information about support services or other ways to self-manage potential distress or other outcomes of participation.
    • If there are support groups or services available for specific topics or participant groups, communities, issues or other groups, reference these in the step-by-step information for participants.
    • It is important to provide this information at the beginning of a survey and near the end in a way that is easy for participants to access (e.g. alerting them to download or screen shot the PICF and a support service information sheet/numbers/websites, providing clear hyperlinks).
    • Including services that specifically address the issues you are researching may be more effective than general services e.g. LGBTIQ mental health services as well as Lifeline; a variety of substance use services including those in regional areas if your inclusion criteria is Victoria-wide, etc.
    • Depending upon the recruitment strategy of the project, it is possible that overseas-based participants may be recruited, which may mean that their needs will have to be addressed in the application as well.
  • Often social media will be used for recruitment in this type of research. This is usually appropriate, if managed correctly, but must also be fully explained in the human research ethics application form. Do not use your own social media accounts to recruit but create new dedicated accounts for the purpose of the research.
  • Remember also that if participants are assured that their participation will be anonymous, then the identity of anyone who has or has not participated must not be divulged before, during or after the research. Absolute discretion is required here if anonymity has been promised to participants. It is inappropriate to name or identify participants in any context, especially on those social media sites that you may have used for the purpose of recruitment. If as part of the research, raffle prizes or incentives are provided, then do so in a way that ensures the anonymity of participants. Do not publicly announce ‘winners’.
  • When submitting copies of any documentation to the HREC, ensure that it is presented in the way that it is presented to the respondent. This may require you to develop a trial survey on a survey platform. Remove any irrelevant or redundant content.

For more information