Creating an Informed Consent Script That Anticipates Sharing Data

In this lesson you will learn

  • Basic outlines of informed consent language
  • The potential interactions between informed consent and sharing data generated through interacting with human participants
  • What to do if you are already using informed consent language that precludes sharing

Initial questions

  • You likely already created an informed consent script. Did you think about how it might affect later sharing of data?
  • If you wrote both a DMP and an IRB application, did you make sure they matched?

Informed consent is at the heart of ethical research with human participants. Whenever your research involves human participants, you need to make sure they consent to being part of your research, and they fully understand, are informed about, what this participation entails. This is an ethical requirement for social science researchers. In the US, it is also a legal requirement, mandated by guidelines from the federal government. This course does not replace a course about ethical research practices. In the US, you will typically be required to take such a course prior to conducting research with human participants. Our focus in this lesson is on developing and using an informed consent script that does not unnecessarily prevent you from sharing data. In the process, we hope to also provide you with some suggestions on effectively gaining consent.

A typical informed consent script includes four main elements

  1. Basic information about the study you are conducting
  2. Any risks and benefits associated with participating in the study
  3. Information about you, the researcher, as well as your institution, with contact information for both
  4. What participating in the study entails

Commonly, 1-3 are included on a separate study information sheet, that participants can take home.

Did you know?

You will often see participants referred to as “human subjects,” which is also how the US federal law refers to them. At QDR, we follow the American Psychological Association’s lead and prefer the term “research participants,” highlighting their willing participation as well as their active contribution to the research.

Qualitative research sometimes has an uneasy relationship with informed consent procedures derived from laboratory research, mainly in the medical sciences. One of the advantages of qualitative research is that it often takes place in a setting close to participants’ everyday life. Introducing a legal document to sign in such a context can significantly disrupt your research. But precisely because qualitative research may not seem like research to your participants, informed consent is particularly important.

One way to make the consent process more natural is oral consent: instead of having participants sign a sheet, you verbally run through a script and ask participants to agree with it. One disadvantage of oral consent can be that it leaves you without proof of consent. Where you record interviews, it is common to restate consent in a highly abbreviated form at the beginning of the recording.

Tip from the field

While oral consent is often more appropriate for qualitative research, don’t assume it is. We’ve heard from a scholar who found that her interviewees (ex-combatants) were reassured by the legitimacy indicated by the written consent form and information sheet. Ask others who have engaged in a similar type of research in the context you are studying about their experiences.

Qualitative research also has an advantage over lab-based research in obtaining truly informed consent. As you build lasting relationships with participants, you are more likely to be able to communicate to them what your research is about and what they are participating in.

Traditional informed consent forms will often preclude data sharing. This happens in the form of two main elements.

  1. Where the informed consent specifically states that only the researcher (or the researcher and his team of x collaborators) will have access to the data, the data cannot be ethically shared.
  2. Where the informed consent specifies that data will be destroyed after a given time, the data cannot be ethically shared.

Such restrictions can be adequate for sensitive, identifiable data, such as audio-recordings or sensitive interviews. They are almost always in order for the keys which allow you to re-identify subjects from numbers or pseudonyms. However, you should, wherever ethically possible, allow for sharing of de-identified data. You can do so very explicitly. Consider, for example, this language recently developed by Cornell University’s Institutional Review Board:

“De-identified data from this study may be shared with the research community at large to advance science and health. We will remove or code any personal information that could identify you before files are shared with other researchers.”

Alternatively, you can mention the possibility of sharing of data in the context of de-identification, as in this version suggested by ICPSR:

“Any personal information that could identify you will be removed or changed before files are shared with other researchers or results are made public.”

Do not make any promises that you cannot keep. You should specify that you will remove identifying information, but you should not promise anonymity to participants, as there are no ways to guarantee that data are truly anonymous.


Evaluating Informed Consent Language

Consider these two examples of language from informed consent scripts, identify problems, and suggest improvements.

  1. Three people will have direct access to these observations: me, and the two other members of the research team. The written observations will be kept during the analysis of the data, that is, about 3 years. After that, they will be destroyed.
  2. We will anonymize all data before sharing it in a trusted digital repository. No one will be able to identify you from that data.
  • show solution
    1. The two biggest issues here are the specification of a restricted group of people with access to the data and the eventual destruction of the data. We would suggest something like this instead: “Three people will have direct access to the full observations: me, and the two other members of the research team. We will remove all information that can identify you from these observations before sharing them with other researchers.”
      1. As explained above, do not promise that no one will be able to identify participants. Moreover, avoid complex technical terms such as “anonymize” and “trusted digital repository. Instead, something like this may work: “We will remove information that can identify you, such as your name and exact date of birth, from the data. Then we will make the data available to other researchers, so that your answers can help scientists better understand why …”

      Re-consenting Participants

      What can you do if you have already interviewed participants, want to share interview data, but used consent language that prevents you from sharing? Often, the answer to this question is, sadly, nothing. But where you have good rapport with your participants and you are able to get in touch with them with relative ease, it is possible to retroactively ask them to consent to data sharing. Note that your consent form has been vetted by your IRB, so you must file an amendment to your IRB application with the updated consent form. Only once it is accepted, should you re-contact your participants. Keep close notes on who did and who didn’t agree to the updated script and remember that you can only share the data of those who did.


      Craft Your Own Consent Language

      1. Now spend some time with writing or editing your own consent script. Reassure your participants that their data is safe, but do not preclude the sharing of data. If working in a group, exchange scripts with a partner.
      • show solution
        1. Read through your consent script (or that of your partner if working in a group) and consider some of the following: - Is the language easy to understand for the targe audience? - Does the script make clear that the data will be shared? - If there is going to be de-identification involved, do you explain what that entails? Do you avoid overpromising?