Research and Privacy

By Karen Sternheimer

A team of researchers from Harvard and UCLA recently faced criticism for conducting research using Harvard students’ Facebook pages. The researchers studied an entire class of 1,700 students starting in 2006, examining how their connections and interests shifted over time.

According to the project’s Harvard website:

This snapshot of an entire class over its four years in college, including supplementary information about where students lived on campus, makes it possible to pose diverse questions about the relationships between social networks, online and offline.

A rich dataset, no doubt, and in 2008, the researchers released their archive of profiles to the public for others to conduct research. The data offered a complex look at the college experience at an “anonymous, northeastern American university.”

But as it turned out, finding out the identities and profiles of the Facebook users was possible for student research assistants at Harvard—who were attending the same university as all of the study’s participants.

One of the core ethical principles of conducting research is that of privacy. Whenever possible, research participants should remain anonymous (their identities should be unknown to researchers); barring anonymity, at the very least their identities should be confidential (their identities should be known only to researchers).

The American Sociological Association published a code of ethics in 1999, devoting a significant portion to confidentiality:

When using private information about individuals collected by other persons or institutions, sociologists protect the confidentiality of individually identifiable information. Information is private when an individual can reasonably expect that the information will not be made public with personal identifiers (e.g., medical or employment records).

We might question whether someone’s Facebook posts have a reasonable expectation of privacy, given all of the controversy surrounding the site’s privacy settings. People might be less upset to hear that their Facebook postings are available than if they conducted a survey and were told their personal information would be kept anonymous and later found out they were identified.

But sociologists are also supposed to make sure that any research report or publication also keeps participants’ identities private:

Sociologists do not disclose in their writings, lectures, or other public media confidential, personally identifiable information concerning their research participants, students, individual or organizational clients, or other recipients of their service which is obtained during the course of their work, unless consent from individuals or their legal representatives has been obtained.

When confidential information is used in scientific and professional presentations, sociologists disguise the identity of research participants, students, individual or organizational clients, or other recipients of their service.

The code of ethics encourages sociologists to be very upfront with participants about how the information collected will be handled. This is very important especially with sensitive personal information, where some form of personal harm may come if the person was identified, be it embarrassment, a loss of a job, arrest, or even violence.

Researchers are supposed to be sure they obtain informed consent from all research participants, meaning that before people agree to be part of a study they are told what participating in the research entails and any possible downside to agreeing to participate.

But sociologists can observe people in public settings without their consent. Some might even argue that any online communication public; certainly as internet users we need to be careful and presume that anything we post could potentially be made more public than we intend.

Social scientists have faced these ethical dilemmas for decades in order to try and learn more about human behavior. Some limits are clear, but there are many gray areas that can be debated about what constitutes a violation of ethical standards.

As online communication has become so central to social life, we are still navigating the boundaries between public and private. So are researchers. How might sociologists conduct research of people’s online interactions without violating their privacy?

Comments

[Chris Wonch]

I feel as though Facebook can be considered a grey area for research as it is a public forum and you can adjust your privacy settings according to your needs. However when it comes to keeping subjects anonymous the very nature of Facebook as a forum along with the fact people generally don't use the privacy settings to a high degree makes identifying subjects incredibly easy.

[McKenzie West]

I thought this was a very interesting article. It really takes both sides of the privacy issue, and explores how this could be considered an invasion of privacy, or just exploring public life and informatation. I think it's okay to study Facebook and other social networking sites, but it's not okay when the researchers know who they're studying. Some people have their privacy settings restricted, and others have their privacy settings set to public. Those who have their settings set to public, I think it's all right to study their profiles and activity; others, not so much!

[stephen]

Facebook could be seen as semi private, and many people are being more carful about what they put on the site, still some people do not see this. Anything you put on the web is public, so this was fine. I still hope all info was anonymous, even if you’re in public you do not have to fear being filmed without your permission.

[gabe]

I have to point out the grammatical error in the second sentence of the first paragraph. "An" should be used - not "and".

[Danielle Boss]

I think that it's okay to study the facebook's thats are public because there is settings for privacy. If you were so concerned with people being able to see it you'd make sure that was private. But if they're viewing something that isn't is private without consent, that's a problem.

[Jackie]

I would have to agree with what Danielle said. It is their own fault if they set their facebook to public. On my facebook, I only show my profile picture, name and my gender. We never know who could be out there "surveying" our facebooks, and for that very reason, I have mine on private.

[Maya]

I feel that using Facebook as an online source for research shouldn't be a problem. If the person really wanted to have all of their information hidden then they would have blocked all of their information on their Facebook page. So I feel like the privacy matter shouldn't be taken into consideration if using Facebook as a source.

[Katie]

Facebook, though public domain, should still be protected and not available for people to simply access one's information without his or her consent. All Facebook users should have the right to know who is viewing their information if it breaches their privacy settings.

[Alexa]

I disagree with the position of this article. I think it is the responsibility of all Facebook users to be aware that the things they are posting could possibly be seen by anyone, no matter what their security settings are. Facebook users should not get upset by finding out their profiles are being used in a study because posting something makes it public information, no matter what it is.

[Ralyn Allen]

I think that if sociologist make it clear to the people participating in the experiment the consequences and what could happen (the information becoming public or people knowing their identity) and they still agree to do it then it isn't the sociologist problem because the volunteers knew from the beginning what could happen. So I disagree with the position of this article.

[Annalie Konyndyk]

I am completely in agreement with this article. The experiment being performed contradticted the code of ethics in many ways. Number one, everything is supposed to remain private and that did not happen. And number two, the people were told that their personal information would be kept anonymous and later they found out that they were identified. I personally, would be very upset with that. However, I can see the other side of this argument. when dealing with facebook it is a very open network, and in reality, any person can view anything you say. I feel that if someone is conducting research through facebook they should include the fact that it may not be kept private.

[Elizabeth S.]

I think that people do not take what they put on the internet as seriously as they should. What is put on the internet stays there forever. People should not put their profiles on public if they do not want other people from around the world looking at their own stuff. To answer your question I would research the profiles that are set on public because obviously they do not care if they are looked at.

[Sarah]

It is interesting sociologically that a lot of other commenters seem to be focused on the fact that it is "your responsibility" to control your Facebook privacy settings. It seems like we're assuming individual responsibility for technological sophistication in controlling privacy settings when it doesn't seem reasonable nor achievable. While I always do this, I am often surprised by how uninformed my students are about how to do this; every time Facebook in particular changes anything, I am tasked with explaining the changes and what they mean to family members in terms of how to set their privacy settings. What about people who don't have a "tech support buddy" like me? How would their information be used potentially in a research study, or by less sociologically ethically concerned marketing researchers? What are the structural impediments to ensuring people DO know how to do this stuff and take it seriously? I would imagine necessary components of using participation in an online social network as an ethical source of data includes only using respondents who have: familiarity with and appropriate use of computer technology (the often-cited digital divide); social network contacts with familiarity and frequent use; and emotional maturity. None of those things is equally distributed or present in the population, so assuming that everyone is equally responsible to protect themselves seems unrealistic and possibly unfair.
Regarding the original study, I think the ethicality of the research is dependent on how much identifying information the researchers included with their public release of the data. If they created pseudonyms, made attempts to get consent from the subjects, or did not provide all identifying information necessary to track an individual, then it is probably fairly ethical.

[Cygnis Media]

Really great post, This is definitely an issue that is important for marketers and social media monitoring companies because although "information is gold", there are ethical boundaries that shouldn't be crossed. Many people want to monitor Facebook, for example, and while there are companies that will do that, Synthesio (I'm their CM ps ;) ) only monitors public spaces, including forums. Transparency is an issue that should be pushed, I think, because it's better for the entire market if people are aware what is public and therefore being monitored by multiple companies at any given time, and what is locked and therefore should be unavailable.

[Cygnis Media]

Really great post, This is definitely an issue that is important for marketers and social media monitoring companies because although "information is gold", there are ethical boundaries that shouldn't be crossed. Many people want to monitor Facebook, for example, and while there are companies that will do that, Synthesio (I'm their CM ps ;) ) only monitors public spaces, including forums. Transparency is an issue that should be pushed, I think, because it's better for the entire market if people are aware what is public and therefore being monitored by multiple companies at any given time, and what is locked and therefore should be unavailable.

[Joseph Mooryes]

Wow great article !