July 20, 2016 - David M Fishman | Storytelling with Data

When data talks, who’s listening? Hadooponomics Podcast, Episode 10

Two young fish swimming along happen to meet an older fish swimming the other way, who nods at them and says “Morning, boys. How’s the water?” And the two young fish swim on for a bit, and then eventually one of them looks over at the other and goes “What the hell is water?” — David Foster Wallace, “This is Water”

Every day, we all swim in data, much like the fish in the story above swim in water. And if you’re a data professional reading this blog post–whether you’re in IT, data science, or hands-on data analyst–you’re the older fish. So let’s face it: we’re outnumbered by those who may use the word ‘data’ a lot, but are secretly asking “What the hell is this data?”

Hadooponomics Brought to you by Arcadia Data and Blue Hill ResearchHadooponomics Episode 10 is “Using Data for Influence: From Spurious Correlations to Presidential Polling” features Dr. John Johnson, CEO of at Edgeworth Economics and author of the recently published book, Everydata. He discusses how data professionals can help everyone from executives to everyday data consumers better understand and make use of data insights. Dr. Johnson, who was an expert witness in the infamous Sony breach, often testifies in high-stakes litigation in front of extremely smart non-statisticians:

“There are some people I deal with … that are really, really smart but just don’t have the training or the background in these things. … Spending the time to think about what the essence of your work is and what it means is so, so critical. … as an expert witness, the most important thing I have is I need to have credibility. Which means I have to give objective opinions. … I think when you can actually show your audience strengths and weaknesses, what your analysis shows, what they need to be aware of, that’s really important.”

As the world floods with data, it becomes increasingly important to understand how and when that data is misunderstood. Dr. Johnson premises that as a pattern-seeking species, humans are driven to connect dots in direct violation of the maxim: “the plural of anecdote is not data.” At the same time, data professionals need to maintain humility and avoid preaching from an analytic pedestal:

“I try to be very pragmatic. I’m not trying to turn the world into a bunch of statisticians that can run derivatives in their head. But for more of an everyday audience, there’s an awful lot of basic skills and approaches which involve awareness, thinking, developing some intuition, and not being afraid of some basic details that could actually be really informative.”

Naturally, translating data into visualizations is fundamentally important to delivering a translation of your analysis into other hands. Of course, we at Arcadia Data are more than happy to show you how you can go beyond static visualizations to let the data tell as complete a story as all Big Data allows. How well this works for your audience must be balanced with the oceans of data they swim through every day.

Here’s a simple example from Dr. Johnson’s book. A 2014 study by the US Bureau of Labor Statistics compares time the average American in 2014 spent watching TV vs. socializing. Visualize data with the area of each circle showing that time, and it looks like this:

Slide1

Visualize this same data using the diameter showing the amounts, and you see something pretty different:

Slide2

No real surprise, but how often are these differences made obvious? Do your visualizations support the drill-down needed to expose those differences?

In a prior post we discussed the necessary skills that an ideal data scientist should have, including subject matter expertise, technology savviness and deep knowledge of statistics. Many of you are already there. But not everyone has an ‘inner data scientist.’  Dr. Johnson offers a salutary corrective as he suggests ways we can use our data skills to teach others, and to make our work as data professionals more useful to everyone.

To hear more of John Johnson’s commentary, check out Hadooponomics Podcast, Episode 10 – “Using Data for Influence: From Spurious Correlations to Presidential Polling.  The Hadooponomics Podcast series is produced by Blue Hill Research in partnership with Arcadia Data. You can listen to prior episodes here.


Related Posts