10/26/15, 6:11 PMFacebook, private traits and attributes: Predictions from digital records of human behavior – Journalist's Resource Journalist's Resource

Page 1 of 3http://journalistsresource.org/studies/society/internet/private-traits-attributes-predictable-digital-records-behavior

INTERNET, SOCIAL MEDIA

Facebook, private traits and attributes: Predictions from digital records of human behavior Tags: facebook, privacy, technology | Last updated: March 25, 2013

(iStock)

Governments, businesses and policy makers are increasingly able to capture and organize vast quantities of personal information — a trend known as “Big Data” — to help them make better-informed decisions. However, this data can also pose a serious threat to individual privacy. In a widely publicized 2012 episode, the retailer Target was able to guess from a young woman’s recent purchases that she was pregnant, before she’d even told others she was expecting. Even criminal enterprises have gotten in on the trend.

Data gathering extends far beyond simply tracking credit card purchases and includes the information people share online about themselves and others. A 2007 report by the Pew Internet and American Life Project found that 60% those surveyed were not concerned about what was available about them online and 61% do not try to control it; however, by 2012, Pew found that many people were employing privacy settings to reduce the information available about themselves. Other research has indicated that those who use social media sites are typically more relaxed about privacy settings and often feel a false sense of control.

A 2013 study from the University of Cambridge [U.K.] and Microsoft Research in the Proceedings of the National Academy of Sciences (PNAS), “Private Traits and Attributes Are Predictable from Digital Records of Human Behavior,” examines the degree to which information available online can successfully predict an individual’s personal — and private — attributes. The researchers correlated public records of Facebook “Likes” from more than 58,000 users with results from personality and intelligence tests and information from public profiles. The study

10/26/15, 6:11 PMFacebook, private traits and attributes: Predictions from digital records of human behavior – Journalist's Resource Journalist's Resource

Page 2 of 3http://journalistsresource.org/studies/society/internet/private-traits-attributes-predictable-digital-records-behavior

focused on whether a user’s history of Likes could accurately predict sexual orientation, ethnic origin, political views, religion, personality, intelligence, satisfaction with life, substance use, age, gender, relationship status, and “whether an individual’s parents stayed together until the individual was 12 years old.”

Key study findings include:

The researchers were able to accurately predict a user’s sexual orientation 88% of the time for men and 75% for women. While less than 5% of user profiles were explicitly linked to gay policy or advocacy groups, “predictions rely on less informative but more popular Likes, such as ‘Britney Spears’ or ‘Desperate Housewives’ (both moderately indicative of being gay).”

The model was able to predict a user’s ethnic origin (95%) and gender (93%) with a high degree of accuracy. “Patterns of online behavior as expressed by Likes significantly differ between those groups, allowing for nearly perfect classification.”

The model predicted whether a user was Christian or Muslim (82%), a Democrat or Republican (85%), and used alcohol, drugs or cigarettes (between 65% and 75%), and was in a relationship (67%) with a high degree of accuracy.

The model was less accurate when attempting to predict the length of the parents’ marriage (60%). “Individuals with parents who separated have a higher probability of liking statements preoccupied with relationships, such as ‘If I’m with you then I’m with you. I don’t want anyone else.’”

The low prediction accuracy for satisfaction with life (only 17%) may be linked to an inability to separate short-term affect (such as bad mood or mood swings) with longer- term happiness.

“Although liking ‘Barack Obama’ is clearly related to being a Democrat, it is also relatively popular among Christians, African-Americans and Homosexual individuals.”

The researchers caution against the potential negative outcomes that ready access to this type of personal data might have: “Commercial companies, governmental institutions, or even one’s Facebook friends could use software to infer attributes such as intelligence, sexual orientation or political views [that] could pose a threat to an individual’s well-being, freedom or even life.”

In a related study, “Silent Listeners: The Evolution of Privacy and Disclosure on Facebook,” researchers found that although Facebook users exhibited more privacy-seeking behaviors, this was offset by changes to Facebook’s privacy policy. “The amount and scope of personal information that Facebook users revealed privately to other connected profiles actually increased over time and because of that, so did disclosures to ‘silent listeners’ on the network: Facebook itself, third-party apps, and (indirectly) advertisers.”

Tags: technology, Facebook, privacy

10/26/15, 6:11 PMFacebook, private traits and attributes: Predictions from digital records of human behavior – Journalist's Resource Journalist's Resource

Page 3 of 3http://journalistsresource.org/studies/society/internet/private-traits-attributes-predictable-digital-records-behavior

Writer: Margaret Weigel | March 25, 2013

Citation: Kosinskia, Michal; Stillwell, David; Graepelb, Thore. "Private Traits and Attributes Are Predictable from Digital Records of Human Behavior." PNAS, March 2013. doi: 10.1073/pnas.1218772110.

10/26/15, 6:12 PMDeen Freelon of American University: Research chat on digital scholarship – Journalist's Resource Journalist's Resource

Page 1 of 6http://journalistsresource.org/tip-sheets/research/deen-freelon-american-university-research-chat-digital-scholarship

RESEARCH

Deen Freelon of American University: Research chat on digital scholarship Tags: research chat | Last updated: April 29, 2013

(dfreelon.org)

Deen Freelon is an assistant professor of communication studies at American University in Washington, D.C., where he examines a wide variety of topics in the area of politics and the Internet. His papers have ranged from Twitter analysis relating to the Arab Spring to research on youth and media. Among his areas of interest is the emerging field of “sentiment analysis” — analysis of social media data — in the service of social science research. He is an advocate of

having more collaborative projects between traditional social science researchers and researchers who specialize in computational analysis, or Big Data studies.

As part of our ongoing “research chat” series, Journalist’s Resource recently caught up with Freelon to ask him for some insights about a variety of digital topics. The following is an edited transcript.

_______

Journalist’s Resource: How do you see the ongoing debate over how the Internet is changing American politics?

10/26/15, 6:12 PMDeen Freelon of American University: Research chat on digital scholarship – Journalist's Resource Journalist's Resource

Page 2 of 6http://journalistsresource.org/tip-sheets/research/deen-freelon-american-university-research-chat-digital-scholarship

Deen Freelon: I think that in the popular media what you tend to see is a polarization of the debate about the Internet’s impact on politics. People tend to fall on one side or the other, in terms of whether it’s positive or negative, good or bad. But for over 15 years, scholars have been saying, “It’s not all positive or all negative, it’s not utopian or dystopian — it’s a mix.” So what I do in my work is to define those specific areas in which we might be able to talk about and define a better vocabulary to discuss more precisely how digital technologies are used in politics, particularly by what I call “non-elite citizens” — people who are not in Congress and don’t have a lot of money. Social media is really big for me as far as that goes, because obviously regular people use it. There’s often a nice interaction between people who don’t have a high national profile and people who do. It’s a really nice space where very interesting political phenomena can bubble up.

Journalist’s Resource: You read reporting on Internet politics all the time. What are your criticisms of what you see? What should good journalists know?

Deen Freelon: One of the biggest things is that there’s a tendency to totalize. This isn’t just exclusively journalistic — I’ve seen some scholars do this, as well. There’s a tendency to pass judgment on the entire Internet with regard to its impact on politics or health or whatever substantive area you want to talk about. That leads to a lot of unhelpful smoothing over and a loss of granularity in terms of different aspects of the Internet. It’s really good for headlines. You see things like “The Internet Is Really Good for Protests.” But the real story is a lot more complex.

If I were to talk to journalists, the one thing I’d say is to try to reduce the scope and scale of your claims a little bit, and try to do justice to the specific topic you’re talking about. There’s a paper that I just finished up last year, which I’m looking to publish, that looks at coverage in both the mainstream and national press and the tech press of the Arab Spring and the Occupy movement. The premise is that a significant proportion of coverage of both protests and events is going to be focused on the technology. That was actually more pronounced for the Arab Spring than for Occupy. A significant amount of these reports talk about technology in a way that was very celebratory; and then you’d have articles where technology wasn’t so much the focus, but you’d have a passing reference of the type, “Everybody knows that technology is revolutionizing the way that protest happens.” That would be way down in the article. It wouldn’t be the explicit frame for the story, but it’s one of those received wisdom things that might even shape readers’ impressions even more than an explicit frame. That’s one of the main things I’d look out for: Try to be more specific about the scope of the claims you’re making.

Journalist’s Resource: We have been curating a page of new scholarship relating to digital technology and the Arab Spring, and there are some other scholars out there who are doing the same. What has your work in this area revealed, and what do we know that maybe we didn’t know a couple years ago?

10/26/15, 6:12 PMDeen Freelon of American University: Research chat on digital scholarship – Journalist's Resource Journalist's Resource

Page 3 of 6http://journalistsresource.org/tip-sheets/research/deen-freelon-american-university-research-chat-digital-scholarship

Deen Freelon: I’ve been involved in a few research projects — one peer-reviewed paper and a couple of foundation reports. A lot of the data that are coming out of here don’t really lend themselves very obviously or unequivocally, I would say, to any particular interpretation. The first thing that I worked on was a report by Phil Howard from the University of Washington, currently on leave at Princeton. His interpretation of the empirical data, which included analyses of data from Twitter, YouTube and blogs — I handled the Twitter data — was that those services were really helping protest in that area. The next thing I did was work on something with a team primarily at George Washington — Marc Lynch, Sean Aday, Henry Farrell and John Sides. In our research we were looking at Twitter again, but instead of looking at the location of tweets as we did in the Howard report, we were really looking hard at the locations of people who clicked on links that came from Arab Spring-related hashtags. We found that the majority of those individuals were located outside of Arab Spring nations and indeed out of the Arab region entirely.

In the Howard report, one of the things that was really emphasized was the fact that, at times when the mainstream media was not paying a whole lot of attention to the events on the ground, the relative percentage of tweeters who said they were within the region in question increased. In other words, when external media attention spiked, so did [Twitter] attention from outside the region. When it went down, external attention regained parity with attention from closer to the region. But then in the “Blogs and Bullets” report, which came from the U.S. Institute of Peace, we found that when you looked at these huge spikes — the media attention really drove the overall amount of attention paid to certain links — the vast majority of the clicks to these hashtags came from outside the region.

You can focus on the small number of folks who use Twitter and other tools in these countries; that’s something that Zeynep Tufecki and Christopher Wilson did in a Journal of Communication study focused on Egypt that they published in 2012. Their idea is that some of the folks who were the most critical in terms of circulating information had access to social media and allowed those messages on social media to those who didn’t have access to it. They also argue for the importance of tools like Facebook over Twitter.

I think the jury is still out. I think we’re going to see more studies on the interaction between social media and politics. We’ll see a corpus of literature developing about the 2012 U.S. election. As technology continues to develop, we’re going to see communications technologies become more domesticated tools in the practice of politics. What’s interesting to me is how they’re used — where the innovation is coming from and whether these things are “successful,” from the standpoint of how success has traditionally been defined by political science and

10/26/15, 6:12 PMDeen Freelon of American University: Research chat on digital scholarship – Journalist's Resource Journalist's Resource

Page 4 of 6http://journalistsresource.org/tip-sheets/research/deen-freelon-american-university-research-chat-digital-scholarship

communications scholars. This means shifting voting counts, changing people’s minds, reaching groups of individuals and, of course, allowing everyday people to have a greater influence in politics.

Journalist’s Resource: What are some of the social media-related research insights that people should know about in terms of national politics here in the United States?

Deen Freelon: There isn’t a whole lot of research on that yet. But a few years ago, there was a book by Kirsten Foot and Steve Schneider called Web Campaigning. It gives a really nice look at national campaigns, several successions of campaigns, and evaluates primarily the websites of campaigns. This is pre-social media. The book chronicles the move from brochure sites in the beginning to more interactive features as time went on. It incorporates multiple campaign cycles and looks at how things were. In terms of the 2008 election, there’s The MoveOn Effect, by David Karpf at George Washington University. It gives a really great look at how information technology really changes what he calls the “organizational layer” of politics, primarily affecting these interest groups. These groups have not been a big area of focus of political science. They are not official campaign organizations, but they do play a big role in terms of getting out the vote or pushing issues into the public consciousness. In terms of the actual campaign, you have Dan Kreiss’s Taking Our Country Back, which is a really nice ethnographic look at the 2008 campaign, with a technological focus. It looks really strongly at technologies that are not really public-focused: The software that figures out who you are and how to target you with specific pitches and SMS texts. He looks at how this evolved from 2004, when you had the Dean and Kerry campaigns. That’s a really nice look at those dynamics.

I’m very interested in what will be said about the 2012 campaign. In the waning days of the election, I published a blog post that took a very inductive look at what was driving the engagement behavior around Obama’s and Romney’s Facebook comments. It wasn’t super- theoretical or anything, but I did it because I thought it would be fun. I looked at the top-five posts on Obama’s and Romney’s Facebook pages. I found that all of the top-five on Romney’s page involved entreaties to get him over some “Likes” threshold — “Let’s get 8 million likes, or 9 million likes” etc. All of Obama’s top posts were pictures of his family. It was one of those pieces of research where there’s no real theory that’s needed; it’s so unequivocal that you know people will get it without much explanation behind it.

The year after an election there’s always a lot of interesting stuff. I’m super interested to see what’s going to come out this year at both the meetings of the American Political Science Association and the International Communications Association, which is going to be in London.

Journalist’s Resource: For people interested in all of these areas — technology and politics generally — what are some places on the Net that are really geeky and great?

10/26/15, 6:12 PMDeen Freelon of American University: Research chat on digital scholarship – Journalist's Resource Journalist's Resource

Page 5 of 6http://journalistsresource.org/tip-sheets/research/deen-freelon-american-university-research-chat-digital-scholarship

Deen Freelon: I’m really glued to my Twitter feed [@dfreelon]. What you should do is go to all the people I follow and follow those people! [laughs] Twitter is really a great place where I find out about a lot of papers that I otherwise would not have found out about. My colleague Katy Pearce [@katypearce] is good and does a lot of tweeting about online communications, as well as the former Soviet Union. There is an account called @InternetPapers that automatically tweets out the table of contents list from several journals that cover the Internet in general and related social science topics. I also

like @VincentR, who tweets out a lot of research. Of course, there’s the Pew Internet and American Life Project [@PewInternet], too.

Another thing that I use is Google Scholar, which recommends books and papers based on your own scholarly output. It’s not something I’d rely on solely as a scholar, but it’s really nice in that it has exposed me to journals that I wouldn’t ordinarily look at. It’s helpful as a silo-breaker.

Journalist’s Resource: Finally, tell us a little about the collaboration you advocate between traditional social science and computational analysis, or Big Data studies.

Deen Freelon: I’ll give you one example. In information science, there are a lot of these studies that are really interested in the concept of “influence.” How do you measure influence on the Web? There are special issues of journals and conferences and panels that are dedicated to this. But very little of the work on the information science side spends a lot of time defining what influence is, or when it does define influence, it defines it operationally in terms of link propagation or retweets. That doesn’t engage with the long tradition of thinking about this concept from a social-psychological perspective or from a communications perspective, or some of the other perspectives that are out there.

This also doesn’t situate the concept of “influence” in terms of everything that it could possibly be. For example, if you’re looking at the concept of influence on Twitter, you could look at whether somebody retweeted it or copied some of it or retweeted a link. That’s a kind of influence. But is that the kind of influence that a politician or marketer would care about? Does it really change personal or voting behavior? Does it change opinions? I don’t think that we can really say that, at least in terms of the studies I’ve seen in this area. We need to be really specific when talking about influence.

I think this is where the communications discipline’s perspective comes in, because we have these qualitative methods that involve asking people if they think of themselves as influencers and, when they are the target of messages, if they think of themselves as having been influenced. This is very different than a researcher coming in from outside of a community that’s already engaged in communication and saying, “Well, this and that is influence, and this other stuff is not really influence.” We have to try to figure out how people understand how influence

10/26/15, 6:12 PMDeen Freelon of American University: Research chat on digital scholarship – Journalist's Resource Journalist's Resource

Page 6 of 6http://journalistsresource.org/tip-sheets/research/deen-freelon-american-university-research-chat-digital-scholarship

is occurring and the ways in which some forms of invisible influence might be happening in these spaces. I think that helps get us to a broader conception of

Related Tags