Come October 31, hordes of miniature ghosts, ghouls and goblins will come knocking down your door for some sugar-filled treats. Of course it’s not what it seems. You know that behind the colorful masks and spooky costumes stand harmless children celebrating Halloween.
The question is, can you say the same about your data when it appears in front of you? Data, too, often is not what it appears at first sight.
Unlike enthusiastic trick-or-treaters roaming the streets in search of candy, data does not intentionally wear a disguise to trick you, yet so many of us are fooled by it on a daily basis. It’s no surprise that data can often blind us to the truth – there’s just so much of it out there. With the amount of data growing exponentially each and every day, it’s often hard to find time to make sense of it all. Instead, we take what it says on the surface without considering whether or not we are asking the right questions of the data.
Take unemployment for example. After one of the worst recessions in modern history, we’re now hearing much celebration in the media about how unemployment is down to almost record lows, 5.1% as of me writing this article. That certainly seems worthy of praise; after all, the data doesn’t lie. While the numbers are technically right, it does not factor in a ton of variables that may actually change the overtly upbeat conversation about the current economic climate.
Please don’t think of me as a ‘Debbie Downer’ trying to rain down on the recovery parade, but this is a classic example of the data not telling the complete story. The Department of Labor measures unemployment by the number of people receiving unemployment benefits. But that, of course, can be misleading since unemployment benefits expire, leaving the jobless without a way to be measured. It also does not take into consideration those who may be working part-time jobs but are actively seeking full-time work to support their families and pay back student loans – the same loans they took out in hopes of landing full-time jobs.
These are the types of factors we need to look at in order to pose the right questions before we take the data for what it’s worth. Of course it’s easy for politicians to look at the numbers without digging deeper because it makes them look good.
“It’s human nature,” says Anthony Scriffignano, Dun & Bradstreet’s Chief Data Scientist. “We have a tendency to try and organize the world around us. It’s how we probably have survived. The world around us is so chaotic so we try to find patterns in it; we try to find things that we think are there and want to believe.”
Unfortunately, Scriffignano believes, we tend to use data to make decisions based on ego or make rush judgments in response to answers we believe others want to hear instead of digging deeper to discover what really lies underneath the data. “There’s a name for it,” he says. “It’s called confirmation bias.”
Getting back to the unemployment example, it’s really not ethical to make a broad-based assumption on the data without looking at other evidence, not when that assertion clearly contradicts what others themselves are experiencing. Unfortunately there are countless examples of this happening all the time, and in much more precarious instances across business and government.
As Gartner’s senior vice president and global head of research said at this year’s Gartner ITxpo, “Data is inherently dumb. It does not actually do anything unless you know how to use it.” That doesn’t stop most of us from seeking out even more data when we don’t see the answers we want. But reaching an honest conclusion goes beyond data volume. It takes an inquisitive approach and relentless curiosity to turn data into dependable insights.
Don’t be tricked by seeing data only on face value. Look deeper.