Frankly speaking, it is the first time I know Facebook has a Data Science Team. But that comes as no surprise – a company sitting on petabytes (1 petabyte = 1 million gigabyte) of personal data definitely would not let the valuable information go to waste. Looking from this perspective, I begin to think Facebook not just as a company, but a social experiment as well, an experiment in which we are all test subjects.

When we share a video/picture from some websites, it shows our personal preferences to certain forms of intellectual creation; when we click on an advertisement, it indirectly tells people (whoever knows we clicked on that advertisement) our current needs and wants; when we engage in discussions/tagging with our friends, we are trying to build or strengthen our social ties. It is hard to observe all these actions if they happen in real world, especially when they involve a billion users literally everywhere on earth. But with Facebook, that observation becomes possible as all these social interactions are converted to “0”s and “1”s and logged in the data centers. And many experiment proposals that were “beyond imagination” to social scientists in the past can now be implemented with ease. In fact, that is what scientists are doing right now, both scientists from the Data Science Team and researchers from many other institutions.

While gaining economic return for Facebook is certainly one purpose for the Data Science Team to conduct numerous user-based researches every year, I believe the major objective is for us to better understand ourselves and the society we live in. Our understandings of how the interactions of individual social elements influence each other and the functioning of society as a whole remain shallow and sometimes naive. With the development of platforms like Facebook, we are building a virtual community which is essentially a projection of our conventional society onto the Internet. And the basic working mechanism of this online community is very similar to that of our real society – the mutual connections and interactions of “cells” give rise to an “organism” which is constantly growing and self-correcting. So by studying data collected by Facebook, scientists can definitely deduce many interesting/intriguing/shocking theories about human beings. And the unique part of this huge social experiment is that the test subjects do not even know they are being studied (people are merely socializing on Facebook and seldom consider the possibility that their actions will be closely observed). As a result, their responses to certain incidents are totally natural and genuine, and that is exactly what researchers want. Actually, even the purpose of gaining higher economic return can, in some sense, be incorporated into the experiment of understanding social behavior – firms are simply “researching” what kind of persuasions users are more prone to listen to.

However, some parts of the experiment process are not as smooth as we want them to be. Facebook currently restricts its data access to employees from the Data Science Team only. Although it recently agreed to let outside researchers examine its data in order to varify its research results, the process for researchers to get the permission itself is already time-consuming and involves too many restrictions. On the one hand, the scientific community needs access to Facebook’s data in order to give consent to the validity of its researches and to run more social science researches; on the other hand, Facebook has the obligation to “protect its users’ privacy and anonymity” (or so as it asserts) and could not let outsiders use its data freely. This dilemma is what keeps more revolutionary findings from coming out. The Data Science Team has done its job well, but some contributions from other competent scientists in the world would definitely make the experiment a more fruitful one.

Moreover, we users are not notified before being treated as test subjects. Facebook anonymizes data before conducting researches so that researchers could not identify certain users from the data they have (and many researchers do not pay much attention to individual data as compared to mass trend). And by not informing users they could be the target of the research, researchers could get the most genuine online behavior they want. But do we deserve the right to know that our actions are being monitored? Even though the feeling of being observed by some total stranger might make some people unnerved, at least we should have the option to be notified about what is going on.

When he founded Facebook, Mark Zuckerberg never imagined his website to become a platform for social research. And now Facebook has evolved into this giant virtual community, it is running both as a billion-dollar business and as the biggest social experiment ever. With greater data comes greater responsibility – at least it is true in this digitized world.