Information Technology Reference
In-Depth Information
Is it reasonable for anyone to expect that the information they post online will be kept as
private as information shared verbally among a few confidants?
This is an important issue. It is easy for a researcher to simply say “if it is publicly available, then I can
take it”; but that simple statement doesn't necessarily fit within the broader tenets of research ethics.
Our concern should be with the subject: what was the intention of that post? Whom did they think
would see it? Did they understand it is visible to everyone? Did the default settings of the platform
change since it was originally posted (consider how Facebook has suddenly made people's “likes”
publicly viewable, when previously they could be hidden)?
I don't mean to suggest that it is never acceptable to mine these Web sites for research data, but simply
we must take great care to consider the context and expectations. It is not simply a matter of “already
public.”
What is your fundamental objection to the research methodology used in the T3 study?
Fundamentally, my concern is centered on the fact that even well-intended researchers—and their
Institutional Research Board (IRB)—failed to fully understand the implications of their methodology.
Like many, they seemed to be holding onto the traditional dichotomy of “public versus private”
information, assuming that because someone posted something on a (possibly) public social media
profile page, it is free for the taking without consent or concern over the poster's original intentions
or expectations. I'm concerned that as more powerful tools to automate this kind of scraping of social
media platforms are developed, and more research—both from highly experienced scholars and novice
undergrads—takes place, this kind of potential breach of privacy and anonymity will become more
common.
If the researchers had been more careful and had succeeded in their goal of making the
dataset truly anonymous, would you still have criticized their study?
Better protecting the source of the data would have helped, and it appears that the researchers have
rewritten the original codebook to remove the unique names of the majors and also make the geo-
graphic origin of the subjects more generic. Despite these improvements, the methodological concerns
persist, and I likely would have still expressed concern over the need for informed consent before scrap-
ing the students' Facebook data.
Are you saying that social scientists engaged in research projects should be required to
get written permission from subjects before gathering information those subjects have
posted on social networks?
This is a complicated issue, and it certainly isn't possible to get written consent from all subjects in
every case. Each research project should be considered separately and reviewed by an IRB and related
experts. I do feel that the intents of the subjects should be strongly weighed in the decision-making
process. I suspect few people with public Twitter feeds ever expected their 140-character utterances—
typically lost in a sea of thousands of tweets every moment—would be archived by the Library of
Congress for research purposes. These are the kinds of scenarios that should force us as a research
community to think about what is the most ethical approach to social media-based research projects.
