As revelations continue to unfold in the latest Facebook scandal, we now know that millions of its users have unwittingly participated in research revealing details about their friends, their shopping habits, and their psychological profile, otherwise known as “psychographics,” or profiling of personality. According to one New York Times report, Facebook users opted in to complete a personality test using Qualtrics, software widely adopted by social science researchers to gather legitimate data in a convenient online manner. University researchers in the U.S. who use Qualtrics must undergo review by their institutions regarding the protection of study participants. Those review panels require that each Qualtrics survey begins with a statement of the rights of the participants, including confidentiality, anonymity (in most cases), the right to withdraw, risks and benefits, and a clear statement of what the participant can expect to have happen with the data. After completing the questionnaire, the researcher provides participants with a “debriefing” form that reveals the real purpose of the study along with contact information of the investigator.
It appears that these protections weren’t taken when Alaksandr Kogan, a University of Cambridge (England) psychologist who worked for Cambridge Analytica, partnered with Facebook to provide a means of profiling users (who had to opt-in) with a personality test assessing the Five Factor Model. This is a test that is widely used in legitimate research on everything from narcissism to psychopathology and every other personality constellation in between. There are short and long versions; this article will lead you to some of these free versions.
It is true that those who gave their most intimate personality data to Kogan had to agree to participate, and then click on the link that took them to Qualtrics. However, what they didn’t realize was that the answers they provided would then provide Cambridge Analytica with profiles that could influence their Facebook feed. Other data about users also got drawn into the profile, which in turn gave even more personal information.
As part of the expose now coming to light, one study, in particular, has not received a great deal of attention, but in some ways is even more ominous than the Cambridge Analytica story alone. In 2015, Kogan published a scientific article with collaborators from well-respected academic institutions as well as his company, and Facebook researchers, in which the claim was made that people of higher social status have fewer international friends. The underlying theory was that people with greater wealth and power don’t need to affiliate with people who aren’t like them; i.e., people from other nations. The authors didn’t seem to think that using data from millions of Facebook data, without their awareness, would constitute an ethical violation. See what you think after reading the details of this paper.
You can begin by considering the source of the paper. Published in the journal “Personality and Individual Differences,” which sounds reasonably legitimate (not exactly a grocery store tabloid), the article’s authors are listed as including an “Aleksandr Spectre.” This was Kogan in disguise, using his married surname. It would be impossible to make the connection between him and the Cambridge Analytica psychologist unless you knew to read the study’s footnote to this effect. Second, the journal itself is “Open Access.” This means that you can read the article yourself without requiring an institutional subscription, such as the very expensive university online databases. Sounds great, until you realize that the business model for Open Access journals involves having authors pay substantial sums to see their work reach the scientific community and popular press. In the case of this particular journal, the publisher (Elsevier) lists the fees as approximately $2350 USD per article. In return for this fee, you’re guaranteed review by academic readers, so you can’t just publish anything.
The articles that the Open Access journal publishes may very well be as high quality as those published in non-open journals, but there is this important distinction to consider. Those considered to have met the highest standards for Open Access journals are listed in the Directory of Open Access Journals. These are the journals that must exercise rigorous peer review and editorial control. Personality and Individual Differences is not listed in this directory.
Now let’s turn to the article itself and how the authors stepped over the line in their use of Facebook data. They note that the population of Facebook users is highly appropriate for research on this topic: “Several aspects of the Facebook platform allow us to overcome certain classic challenges in social sciences. First, Facebook’s user base is massive, spanning over 1.3 billion users; thus, in our second study, our findings provide insights based on data from every corner of the earth and most walks of life” (p. 225). Kogan and his collaborators clearly believe that overcoming the “challenges” of having to recruit participants who deliberately agree to be in a research study justifies their use of data obtained without permission.
There were two studies published in the Yearwood et al. article. In the first, participants agreed through the normal route of providing consent to complete an online survey. As the authors point out, “no deception was used” (p. 225). All participants had at least one Facebook friend. When they agreed to be in the study, they agreed to authorize Facebook to gather information from their profiles automatically which may or may not have been in the “fine print” of the consent form. This agreement, though, meant they would now provide information that could be used to find out their total number of friends and where their friends were currently located. The total number of friends whose locations and contact information was obtained, totaling 287,739 Facebook users. In other words, over a quarter of a million people had their Facebook data accessed without their knowledge, and all through the pushing of a “yes” button by the actual study participants.
The results of this first, survey-based, study showed a small but significant relationship between people’s social status and the number of international friends. The findings, the authors claimed to support the “restricting social class hypothesis” that wealth narrows your friendships to those in your own country. With this as their starting point, the Yearwood et al. team moved on to the next study using all the Facebook data in the world with, of course, the help of Facebook. As the authors stated, “Facebook provided data on every friendship formed in 2011 in every country in the world at the national aggregate level. These datasets included a total of 57,457,192,520 friendships. From these data, we knew how many friendships were made within each country (domestic friendships) and also how many friendships were made between every country pair (international friendships)” (p. 226). Although these were aggregated data (i.e. no data from individuals), profile and contact data of individuals clearly had to be fed into the analyses in some form.
To be sure, the Facebook data used in this second study was nation-, not individual-based. At this national level of analysis, the authors concluded that people from “high status” countries had fewer international friends than people from “low status” countries, a determination based on Gross Domestic Product of the user’s home country. The effect, though significant, was relatively small, with people from low-status countries having 35% international friendships and those from high-status having 28% of their friends located in other countries.
With these numbers in mind, the authors conclude that people of higher status (or at least those living in high GDP countries), are more likely to have outgroup biases, greater anxiety about people from groups other than their own, and higher levels of prejudice. People with greater wealth, in other words, “tend to think and act in ways that reinforce their social class” (p. 228), despite their greater opportunities to travel and conduct work at an international level.
This paper was only one of the studies, published or not, that Cambridge Analytica performed on Facebook users without their explicit consent. It fails to conform to the ethical standards that psychologists must adopt, as well as the standards that academic journals require before they will publish a study. Additionally, funding for this study was provided by a U.K. research grant as well as by a grant from St. Petersburg State University, in addition to the personnel and resources made available from Facebook. In the U.S., funding by the National Institutes of Health or National Science Foundation will not be provided to a researcher without clear identification of the methods used to recruit participants.
Psychology research can provide you with information that can enhance your life and make it that much more fulfilling. The Facebook studies were an anomaly. The next time you read a study, or consent to be in one yourself, this Facebook teaches you that it may be worth reading the fine print before you push that “agree” button.