Striking a balance

Mon 01 September 2014 by Rick Gilmore

YouTube is a fantastic service. I use it all the time in my teaching. But, those of us working on Databrary find ourselves saying "We're not YouTube" on a regular basis. Here's why.

Our core community consists of researchers who study infant and child development. We use video to collect the richest possible sample of behavior in natural or laboratory settings. Video often contains faces, voices, and names spoken aloud. This means that the videos could be used to determine the identity of research participants. When researchers collect information from participants, we promise to protect their confidentiality and preserve their privacy. Those promises stem from longstanding principles of research ethics with human participants such as the Belmont Report. The challenge comes in determining how to keep these promises while enabling researchers who haven't been part of the original data collection to see, and possibly use in their research, the recordings collected by others.

Many others have spoken about the virtues of making data sharing common practice in science. Our colleagues at the Center for Open Science have been particularly visible on these topics, especially in behavioral science. The Databrary Team has weighed in on the issue, as well. Much more can and should be said on the matter, but I won't do so here. Instead, I'd like to explain how we solved the practical problem of sharing identifiable research data.

It's simple, really. We ask permission. Whether as part of the formal procedure when researchers ask permission or consent from a participant to study them or as a separate release related to photos and videos, we ask the people we've recorded to give permission to share the materials with other researchers. This is one way we aren't YouTube. We restrict who will have access to Databrary. We also require researchers who want access to sign a formal agreement to become authorized. The researcher's institution has to sign, as well. Researchers granted access promise to protect confidentiality of research participants and follow other ethical practices. YouTube has a terms of service agreement that every user has agreed to, but how many of us have actually read it?

We want Databrary to be as easy to use and as powerful as YouTube, but with a much more limited target audience, with users who've read the terms of service and understood their responsibilities, and with more constraints on how the materials will be used. To do so, we've tried to strike a balance between enabling greater video data sharing while holding true to ethical principles governing research conduct. Some balancing acts are easier than others, but we're optimistic this approach will work.


Fork me on GitHub