Joseph Konstan: Ethical Research and Privacy in the Age of Social Media

In this panel session from the 2013 PRIM&R Advancing Ethical Research Conference, we continue to look at the shifts and trends in social media use that impact the way we culturally perceive privacy and risk. Play the podcast in the player below or download it to your computer, smart phone, or tablet. Or, subscribe to the podcast on iTunes or Stitcher, listen on Google Play Music, or sign up to be notified when a new podcast is available.

Joe Konstan WebinarThe speaker is Dr. Joseph Konstan, a professor in the department of computer science and engineering at University of Minnesota who has conducted research in the field, and who advises the institutional review board (IRB) on data security and IT issues related to internet research.

Dr. Konstan points out that the overriding theme in the field of ethical research is that everything is changing, particularly when it comes to how people balance privacy with the desire to be recognized.

He looks at the issue from two perspectives:

  1. the internet as a venue for research, including online surveys, focus groups, and crowd sourcing research; and
  2. internet activity as a subject of research, including analyzing social media data and online experiments.

In his discussion, the following five issues stood out as areas that still need to be resolved:

  1. The security and confidentiality issues of online surveys - for example, is it okay to use Survey Monkey or an outside source? The answer depends on many factors, including the following questions: How is the data identified? What are the questions asking? What agreements do you have in place with contractors? Is the data encrypted? How compromising are the questions?
  2. What about unexpected coercion? Dr. Konstan points out  the problem with radio buttons on computers, that once you click on one, you can’t un-click it.  Participants need to have an “I don’t want to answer” option to avoid saying something they can’t take back.
  3. There are many risks involved with online focus groups risks, as moderators don’t have much social control for guiding the discussion. Also, there’s no way to debrief somebody who chooses to leave the group suddenly. 
  4. The lines can become blurred when it comes to crowdsourcing, e.g., who is a research subject and who is an employee? This raises a question about which things are really just observations of pubic behavior versus that of paid contractors. Furthermore, there’s no consensus about appropriate compensation.
  5. How to analyze social media data - is it all considered public data? For example, what about Facebook posts that were limited to a certain group? Or what if a post was originally private, but the provider changed their default setting so it is now public?