Social media has emerged as a promising space for data acquisition and study recruitment, due to the vast number of diverse users and abundant, accessible data. Taking advantage of this incredible research resource also raises important ethical questions. As part of a larger project examining ethical challenges posed by digital technologies, the IU Ethics, Values, and Technology project recently hosted three webinars covering the promise and perils of research using social media, focusing on studies on sexual health, attitudes towards disease, and adverse drug-drug interactions. With support of Lilly Endowment Inc., Indiana University is studying ethical issues raised by digital technologies that span human activities “from science to business to culture to politics,” aiming to develop “integrated initiatives to better equip students and faculty to confront and address these ethical challenges presented by our digital world.” All the webinars are available on the project’s website (https://digitalethics.iu.edu) and through the links provided at the end of this post. The researchers shed light on numerous topics involving research ethics but perhaps the most persistent and The challenging topics discussed in the webinars center on study recruitment and use of data from social media.
Drs. Lisa Carter-Harris and Amanda Gesselman each focus on study recruitment as a part of their respective webinars. Dr. Carter-Harris shares her experience using social media to recruit for lung cancer screening studies, while Dr. Gesselman explains how social media has helped recruit participants in her sexology research. Taken together, these two webinars highlight the strengths but also raise important questions about consent and the relationship of the researcher to her participants when they are recruited over social media.
There are several advantages to using social media for recruitment. First, this sort of recruitment can save researchers money, which may be reallocated elsewhere in their projects. Another benefit is keeping participants better informed and involved in the project. Social media offers an open and user-friendly avenue for communication between researchers and their participants. This is a promising area for creating stronger relationships and trust between researchers and the public. One final advantage is the potential to recruit a larger and more diverse participant pool. Not only does this increase the validity and reliability of the study’s results, but it also diversifies the participant groups and ethically represents historically excluded and exploited groups in the scientific literature. This can ultimately bring our scientific work towards more equitable and inclusive ends.
One of the most ethically challenging parts of using social media to recruit participants is consent and communicating risk. While similar issues arise for other ways of recruiting, what is relatively unique is the increased risk to a participant’s social media contacts. Social media is a place where people message, post and repost, and tag other people regularly. These interactions increase the likelihood of an individual’s information finding its way into a study’s data set without their consent. Researchers must therefore be aware of the increased risk of unethically involving a participant’s contacts when recruiting on social media.
For example, consider Dr. Gesselman’s current research on how members of the LGBTQ+ community use social media to mediate their relationships. Participants’ posts will often feature sensitive information about other users. Consequently, the researcher must must communicate not only the risks of identification to the participants themselves but also the risk they introduce to their contacts. The high social connectivity characteristic of social media makes this “bystander risk” a much more prominent consideration for social media researchers as they recruit in this space.
Recruitment represents just one use for social media in furthering scientific research. It also has great potential as a place where people discuss their experiences with health and disease, which researchers can tap into to establish clinically important relationships. Drs. Natalie Lambert, Wendy Miller, and Rion Correia each share their experiences using social media to discover information regarding health, disease, and treatments in their webinars. Dr. Lambert’s most recent work performs a qualitative analysis of posts in the Facebook group “Survivor Corps,” a group for people dealing with long-term COVID symptoms. Dr. Miller’s work largely focuses on correlations between post content and sudden unexpected death due in epilepsy (SUDEP). Lastly, Dr. Correia’s research bridges social media activity with electronic health records (EHRs) to better understand certain drug-drug interactions (DDIs). Using social media in these ways pose important questions regarding de-identification and obtaining consent to study social media behaviors.
There are a few options available to researchers when turning to social media for data. One option is focusing on publicly-available activity (e.g., public posts and reposts), which often does not require the researcher to obtain permission to use the information in her data. Proper de-identification, however, is still necessary to protect the owner of the post. This also comes with the important caveat that study type (i.e., longitudinal studies) may necessitate stricter standards of consent regardless of whether the post is public or private. Another option is private posts. These often do require prior participant approval for research use. Ultimately, it is up to the type of study and IRB review process to determine the level of consent researchers must obtain for their social media studies.
Other studies, like Dr. Correia’s on DDIs, not only perform qualitative analyses of posts involving certain drugs but also link the posts to information stored in health records. Linking these two is a promising step, but one which requires a much more involved consent process.
Perhaps the most involved is reaching out and obtaining individual consent. Dr. Correia and Dr. Miller used such a strategy while studying SUDEP and social media post content by having the participants’ families posthumously donate their social media timelines. This type of research requires a lot of work, which is evidenced in their only having 12 participants in the study. Other ways to bridge the two is to use anonymized records. Though less time consuming, it is harder to draw concrete connections between an individual’s social media behaviors and their clinical health. We can expect researchers to find ways to make these connections easier, but participant confidentiality and privacy remain the key ethical consideration in these types of studies. Solving these problems of privacy and confidentiality will decide how rapidly researchers draw clinical connections to peoples’ social media presence.
Many of the considerations common to our discourse on research ethics receive added emphasis when considering social media as a recruitment tool and place for data acquisition. Among these are consent, risk communication, de-identification, and confidentiality. These webinars provide only a glimpse into the budding ethical conversation surrounding researchers’ social media usage, and we will be tasked with rethinking our ethical understandings to accommodate social media’s growing role in advancing our health sciences. To view the full webinars, follow the links below: