Barring congressional oversight that may change how we understand social media companies and their role in content moderation, the University of Massachusetts Amherst’s Human Research Protection Office (HRPO) considers research utilizing social media (Twitter, Facebook, TikTok, public websites, etc.) involving no interaction or intervention with the individual about whom data are being collected (e.g., public Twitter feeds; public Facebook profiles or wall postings; information from public forums, open chat rooms, etc.) in most cases will not be considered Human Subjects Research as the federal regulations currently define it. While we consider most passive social media research (involving no interaction or intervention with the individual about whom data are being collected) not Human Subjects Research, we strongly recommend that researchers follow these best practices:
- Research personnel should successfully complete human subjects ethics training. We recommend either Group 1: Biomedical research Investigators and Key Personnel - Basic Course; or, Group 2: Social Behavioral and Education Research Investigators and Key Personnel - Basic Course.
- If a post is not from a public figure, we strongly discourage quoting the social media posts verbatim since a simple query on a search engine will lead to the individual. When reporting findings, researchers should ensure that all the private information about an individual is removed. As a courtesy, we recommend that individuals not be individually identified or that the information on the individuals be combined in such a manner that the identity of the group or individuals cannot be readily ascertained.
- If needed, the researcher should submit a Determination Form in Kuali to seek a formal confirmation of not human participant research status for the study. However, depending on the nature of the study, the HRPO office may require a protocol submission out of an abundance of caution.
There is a strong argument to be made that social media posts are by and large meant to be public and engaged with (re-posted, shared, go “viral’, etc.). If an individual or social media platform has not placed any restrictions on access to information, that is, the information is available on a public website, blog, Twitter feed, open chat room, etc., the data should be considered as “publicly available” as understood by a general audience.[1]
We should acknowledge that there has been a healthy public debate regarding the role of social media in research with some pointing out that some data included in posts can indeed be private, not just about the post creator but about others (“doxxing,” intimate pictures, etc.). Others voice concerns about whether a post’s use can actively harm the poster if highlighted in an academic setting. We must concede that currently, these platforms exist in an ambiguous space and that sometimes a “public” post or “public” settings may not indicate an active choice (“Participant” Perceptions of Twitter Research Ethics”). There is currently a lack of consensus in the field as to whether oversight is needed, as well as the proper way to collect, analyze, and report on social media data. It is possible that reputational harm could come not just to those directly quoted in academic research but third parties whose data may be disclosed in posts. Concerns over consent and privacy do not disappear simply because individuals participate in online social media networks and because the current human subjects research federal regulations do not seem to apply. In the past, out of an abundance of caution, our office has erred on the side of some oversight but at the lowest possible risk category (Exempt category).
Lastly, we should also point out that the IRB looks not only to the Code of Federal Regulations but also to the spirit of the Belmont Report and IRB members’ own sense of ethical conduct. Since the regulations do not cover everything, and emerging technologies raise novel issues and concerns, IRBs are given quite a bit of leeway as a body that represents not only researchers but also includes non-scientists, and community members.
If you have any questions or concerns regarding any aspect of this policy guidance please contact us by phone at (413) 545-3428 or via email at @email.
References
- UC Santa Barbara’s In line with online research guidance.
- Cornell University’s Use of Social Networking Sites or Mobile Devices for Human Participant Research
- NIH National Library of Medicine Blog “On the Ethics of Using Social Media Data for Health Research
[1] If an individual has restricted access to the data in any way or the social media platform or site has restrictive provisions in its terms of service, an expectation of privacy has been established and the investigator must seek IRB approval before conducting the research (e.g., the researcher has to request or seek access from the individual or from the group that the individual belongs to, If the researcher has to belong to, be invited to, or invite others to a particular “interest” or “friend” group, etc.).