I held the small white envelope with official county seal in my hands with a mixture of dread and curiosity. It was a call for jury duty. Over two days I would be part of a pool of potential jurors who were exposed to a process called “voir dire” which means to speak the truth.
The voir dire process is slow. It involves asking layers of question of each potential juror to see whether they might hold a bias about any aspect of the case. A bias that could interfere with their ability to be fair. So I couldn’t help but wonder, ‘Could this process be improved by the clever use of data?’
Instead of asking each juror if they can serve, would it not be easier to pre-pull jurors who are automatically disqualified? For example, if the court knows they have a trial which will cover 1 week or 3 weeks, could they pull airline records to see if any juror already had pre-booked non-refundable travel? Or could they pull income records to determine if a person earned an hourly income at the poverty level thereby creating an economic hardship. Possibly.
But as I mentioned, the voir dire process is about layers of questioning. While we might find some efficiencies in the basic “can you be here” questions, it’s in the deeper and more personal questions that big data would ultimately fail. These are questions that start with “Do you know any of these people?” and end with “Do you feel people have a right to sue for any amount of damages?” and “Are children more likely to lie than adults?” Questions designed to ferret out an emotional charge which is the foundation of bias. When a potential juror said, “I think I could be fair” despite having a significant opinion, one attorney politely pressed, “Would I want you to fly my plane … if you only think you could?”
As humans, sometimes we do not know how we feel about a topic until we explore it. And when we explore it as a group, new themes and ideas come up which shape our thinking. This is the heart of the legal system’s voir dire process.
But when we work frequently with data, there’s a tendency to think not only can we predict who will buy, click or convert but that we know why they will take action and what messages should be sent. And this is to ignore the very essence of what makes us human – our inalienable right to be fickle, unpredictable, biased and irrational. Just like the stock market. Just like the voir dire process. The one thing big data cannot overcome is human nature.