What do we know about big data and understanding human behavior?
New Data, New Insights
The premise in Everybody Lies seems sound to me. Here we have an entirely new dataset that reveals things about people that they are not willing to reveal. Advice: don’t listen to the audiobook in the car with kids. It goes into a lot of challenging areas, e.g., racism, sexual proclivities (some quite dark).
The author is especially proud (and one can see why) of his ability to offer a view of racism in the US. He argues that academics have largely focused on implicit racism (non-conscious/automatic). They have recently somewhat glossed over old-fashioned racism. People still make a surprisingly large number of searches with the n-word in. This may be true but I still think non-conscious racism is worth studying. Given it doesn’t fit with his methods he seemed a little dismissive which was a bit disappointing.
The core idea of the book is hard to argue with.
In other words people’s search for information is itself information.
Stephens-Davidowitz, 2017
What people ask Google tells us a lot about what they are thinking about. Their secrets, worries, and hatreds. Much of it was fascinating. I think we can learn a lot. As Stephens-Davidowitz emphasizes people will tell Google things they won’t tell humans.
We Won’t Vote Conservative, Everyone Lies
Stephens-Davidowitz will probably annoy a whole bunch of academics with his writing. He does seem to dismiss survey research. The title “Everyone Lies” seems to imply that it isn’t worth asking people what they think. After all, they lie. He comes across as a bit of a big data evangelist telling us that all past methods are dead. I am somewhat (a little) with him. I don’t want to dismiss past methods but we certainly do need to think of whether the majority of marketing scholarship should be lab-based student subjects studies in a world where a host of other (sometimes better, sometimes worse) data is available.
The Dangers Of Big Data
I enjoyed his discussion of the dangers of big data. It made me laugh a bit because it felt forced. Almost as though someone told him he needed to warn of the dangers of big data but his heart wasn’t it in. There are intriguing questions about invasion of privacy but I personally don’t see privacy extending to your right to plan to murder your girlfriend in secret. I don’t think he does either.
I was a bit confused at the end about what he thought were the dangers. This was given his most cogent argument was that you’d get a load of false positives if you visit everyone who is looking into killing their girlfriend. Let us get beyond the scary idea about how many people are considering murder. Still, I’m not sure that his is a proper moral argument not to monitor potential murders (i.e., the argument that we shouldn’t). Instead, I saw it as a practical argument (e.g., we can’t even if we might want to).
On the moral front maybe we should risk annoying a few innocent men with intrusive questions if we help prevent a few girlfriend murders? Murders see many times worse than questioning however annoying. If you search, “how to murder my girlfriend”, even if you turn out to be a mystery writer doing research, I don’t really have a problem with a police visit to check on you. The moral arguments are fascinating but probably the weakest part of the book.
Big Data And Understanding Human Behaviour
Overall though this is a great book, filled with fascinating and somewhat disturbing detail. What we do online really can tell us a lot about who we are.
For more on big data see here, here, and here.
Read: Seth Stephens-Davidowitz (2017) Everybody Lies: Big Data, New Data, and What the Internet Can Tell Us About Who We Really Are, Harper Collins