Confidentiality of human subjects
If your research involves human subjects, you need to comply to the Concordia Policy for the Ethical Review of Research Involving Humans (VPRGS-3). To learn more about Concordia’s research ethics policies and procedures, please consult this page from the Office of Research.
When managing research data dealing with human subjects, keep in mind that the confidentiality of your respondents is paramount. You should start thinking about preserving confidentiality from the very beginning of your research project. At the same time, consider how best to share your data at the conclusion of your research project.
In particular, pay attention to the following elements:
- Devise your consent form with data confidentiality and data sharing in mind (these are not contradictory);
- Determine if you will need to de-identify or anonymize your dataset before sharing it. These procedures can be time-consuming and may necessitate an appropriate budget;
- Consider where (in what data repository) you will store your data at the end of your research project; this choice should be informed by the type of confidentiality review and access control options offered by that repository
Despite what some researchers believe, even sensitive data can be ethically shared within the research community provided that adequate measures are taken. Make sure not to write your consent form in a way that would impose unnecessary limitations on how your data can be reused. Remember that it is usually very difficult or even impossible to retroactively obtain consent to share data.
When creating your consent form, consider the following elements:
- The consent form should explain to the participants how you will maintain the confidentiality of their records.
- Participants should be informed of the exact type of data that will be collected and the purpose of the data collection
- If you plan to share your data this should be made explicit; you should specify who will have access to the dataset and for what purpose
- Do not commit to destroying your dataset unless this is deemed necessary