A few years ago, as I was standing in the bookstore, I heard someone on the radio talk about a new study published in the Proceedings of the National Academy of Sciences (PNAS) showing that a computer algorithm, relying only on the things you “like” on Facebook, makes more accurate judgments of your personality than your friends. If you heard about this study, it might have made you feel a bit squeamish. Maybe it even made you want to delete your Facebook account. In the wake of Cambridge Analytica, it is certainly reasonable to wonder just how much big data companies (like Facebook, Google, Verizon, or Visa) know about you. Having personally reviewed this study before it was published, I was not quite so concerned. Let me explain.
The study itself showed that aggregated Facebook likes (i.e., the things that you like on Facebook) can be used to predict self-reports on a personality test. Further, when the total number of likes is large enough, the aggregated likes show a stronger relationship with self-reported personality than reports from your friends, family, spouse, or colleagues. This was widely reported to indicate that computers make better personality judgments than humans. I have three problems with this conclusion.
1) The data show that self-other agreement with human judges was about r = .49, while self-other agreement with computer-based judgments was about r = .56. In real-world terms, what these numbers mean is that if you judge yourself to be above average (the median) on a trait, your friends are likely to guess that you are above average 74.5% of the time, while the computer algorithm guesses correctly 78% of the time. This is a real difference, so I don’t want to downplay it, but it is important not to oversell it either.
2) One of the most interesting findings from this paper was the fact that both aggregated Facebook likes AND peer judgments of personality predicted self-reports of personality largely independently of each other. Average self-other agreement with human judgments was r = .42 when controlling for computer judgments. Likewise, average self-other agreement with computer judgments was r = .38 when controlling for human judgments. Both the computer algorithm and human judgments were related to different parts of self-reports. That is, peer-judgments of you and like-based judgments of you did not overlap very much.
3) Although the reports made it sound as if computers have some sort of knowledge that we do not, this is of course not true. The computer-based algorithm for making personality judgments is based entirely on the person’s behavior. That is, “Liking” something on Facebook is a behavior. The computer is taking the sum total of those behaviors and using them as a basis for “judgment.” Critically, these behaviors came from the person whose personality is being judged. Thus, one could argue that the computer judgments are merely linking self-reports of behavior or preferences (e.g., I like Starbucks) with self-reports of personality. In other words, the paper showed that how you describe your own personality is related to the things that you like. When you put it like that, it does not sound nearly as disconcerting.
I don’t mean to denigrate the study here. It was an interesting and well-conducted study on personality assessment. Still, what would be more interesting is the knowing the degree to which aggregates of Facebook likes predict (a) one’s reputation and (b) how one will perform in the workplace. Regarding the former, the data from this study indicate the relationship between Facebook likes and reputation is pretty weak, suggesting that Facebook behavior is mostly about identity, not reputation. Regarding the latter, there appear to be no studies speaking to the question.