Addressing privacy issues when sharing videos of classroom research

In academic research, data sharing has become an aspiration. After all, why waste resources to reproduce data that has already been collected or why not take advantage of data that cannot be re-collected, especially as the ability to quickly share this data has become so much easier? Some funding agencies have even begun to require researchers to make data management and sharing plans.

But what if the data is qualitative? What if the subjects are school children? Do attitudes about data sharing change when the research data includes classroom video?

A new study from researchers at the University of Michigan School of Information answers this question with a resounding “yes.” Researchers and educators are expressing significant concerns about protecting the privacy of their subjects, and have come up with a number of strategies to mitigate these concerns.

Video records of practice are a unique and important type of qualitative data used in the field of education. These videos can give a window into how students learn from different teaching methods, or highlight skills and techniques for novices. As the ability to share these data quickly has become much easier, the worry about protecting the teachers and students in those videos has grown, too.

The paper is part of a larger project on qualitative data reuse in education. For this part of the study, the team focused on 44 interviews with researchers and teacher-educators who have reused video as qualitative data for research or teaching.

Rebecca Frank

Dr. Rebecca Frank, first author of the study and a former doctoral student in the U-M School of Information, says people talk about the subjects of the videos in very protective ways, and have developed a variety of means to try to address their privacy concerns.

Perceptions of Harm

“One of the interesting findings of this research was how many avenues of harm people were thinking of,” says Frank. “Violating the privacy of a subject in a video isn’t just seen as having a negative impact for the subject in the video, but also for everyone involved with the creation, use, and sharing of the video.”

Areas of concern included:

  • Personal harm for the data producer: Frank says that data producers worry that if the data they collect is misused at any point, they could be personally culpable.
  • Reputational or financial harm to the subjects: These videos usually include many days of filming, Frank says, and participants expressed concern that a video showing a teacher or student on “anything less than their best day” could be pulled out of context and used against them.
  • Harm of identity exposure of people in the video: Because the education field is relatively small, there is a risk that subjects in the videos are members of the same community – or sometimes in the same audience – as those viewing the videos.

In other types of research, anonymization could solve these concerns, but the type of information captured by classroom videos does not lend itself to this because reuse depends on high fidelity to the video.

“When you blur out faces, you miss much of what your research project is about, because a lot of it is about facial expressions, where attention is focused, and how students respond to engagement,” says Frank.

Instead, researchers and teacher educators have embraced a few different tactics to mitigate and offset the potential harm to privacy.

Mitigating Harm

For some researchers, reducing the potential for harm starts with educating and training future teachers and researchers to be responsible with these data that includes personally identifiable information.

Others limit what they capture when producing videos, or edit videos to exclude the faces of those subjects who did not consent to participate. The latter can be expensive and time-consuming.

Finally, interviewees revealed that many data producers were carefully crafting narrowly defined consent forms, restricting future access to their data.

“So instead of doing what people are being encouraged to do now, which is write a consent form that lets you deposit your data for maximum availability for reuse and sharing, they were using the ethics process to narrow the shareability of the data instead of broaden it,” says Frank.

“It really drove home for me the idea of how protective the researchers and data producers in this space are of the subjects.

The bottom line, says Frank, is that any efforts to facilitate sharing of video records in education will need to take these concerns into account.

“Privacy Concerns in Qualitative Video Data Reuse” was published in volume 13/1 (2018), of The International Journal of Digital Curation.

Other U-M authors were Allison R.B. Tyler, Kara Suzuka, Anna Gault, and Elizabeth Yakel.

- Jessica Webster, PR Specialist

Posted February 5, 2019