“The statistical method shows the facts in the light of the ideal average, but that does not give us a picture of their empirical reality.” – Carl Jung
Pithy, isn’t it? Okay, it’s actually a rather dense quote. What it means is “stop putting people in buckets”. Thanks for coming to our TED talk and we hope you enjoy the day. Just kidding, let’s dig into this a bit.
First, isn’t it interesting how people can often spot problems early, long before the rest of us catch up. Typically, we ignore them and their concerns until it is years, sometimes decades later and someone else remembers the lost insight. That is the case here. That quote from the great psychologist is from 1957, decades before the digital revolution was underway, yet it is incredibly relevant to the present day. It is an indictment of our over reliance on statistics in our decision making processes.
Even the fact we tend to ignore insights like this, insights that are ahead of their time, proves the point of the quote. We ignore things like this based on an unconscious analysis that is grounded in statistics. Fifteen years ago, most people would have said, “I’ll never really ignore people in favor of my phone or an attractive spreadsheet.” Because a thing has never happened or has only happened rarely, that doesn’t mean it can’t or won’t happen. We hear this kind of thing in politics all the time. “No one has ever been elected with this….” Insert whatever statistical fact you want. And then it happens.
The truth is, statistics are great predictors, until they aren’t. Just because a thing usually happens in a certain way, there is no particular reason to think they will always go that way. What’s worse is that we think knowing some statistics is the same thing as really understanding something. We tend to treat them as explanatory, when they are only descriptive at best. There are many times when statistics aren’t even properly descriptive. Instead, they are illustrative of the analyst’s biases.
This is particularly true when applied to people. Imagine someone who gets a ton of ads for Christmas music. Why might that be? Because they often buy Christmas albums? Not necessarily. Remember, the algorithms that drive the ads operate my cross referencing certain behaviors. In this case, let’s imagine that this person with all the Christmas music ads tends to order a new ugly sweater on Amazon every year. The algorithm assumes that the person likes everything having to do with Christmas. Maybe this individual does like most things associated with the holiday. Everything but Christmas music. In fact, our sweater wearing friend hates Christmas music but endures it for the sake of the annual ugly sweater party with his friends. I can guarantee those ads are not going to convert him into a sale for the latest Mariah Carey Christmas album.
Why do we do this? Why do we make all of these guesses? Why rely so much on assumptions and allow our decisions to be guided by statistics and algorithms? Because it is easy. Find a few statistical correlations and develop an algorithm from them and then run all your data through that. Broadly speaking, the picture it forms may even be accurate. But you don’t really know for sure. You certainly don’t know where it falls short or why. The only way you really can be sure is by going to the individuals behind the statistics, the people actually generating the data that all these programs are trying to classify. Then ask them, “what were you thinking when you did ‘x’?” That’s how you get real knowledge, and real understanding, by treating data with the respect you give to the people who generate it. Because that data represents them and their thoughts. That is powerful and understanding is the first step on the path to real, truthful knowledge.
What’s your data worth? Sign up for the TARTLE Marketplace through this link here.
