Yesterday, I attended the 2017 New York R Conference. A talk that piqued my interest was titled “The Unreasonable Effectiveness of Empathy” by JD Long.
In the talk, JD espoused the values of modeling other people’s perspectives, and the ways in which it is crucial to a technical career:
It is important when getting requirements for a project
He told a story about how he was asked to write code that would fit a curve to a dataset. He spent a week building a stand-along Windows application. When he demo’d it to the user, he found out that the user wanted an MS Excel plug-in.
If JD hadn’t assumed the end user’s needs, he would have had a conversation with the user. It would have served everyone’s purpose better, and the whole process would’ve been smoother.
It is important when working with data
JD showed a screenshot of the website of the D.C. Homicide Watch website. It contains homicide data from Washington D.C., including geo-data.
It would have been very easy for the creators of this website to put a flashy map of the homicides front-and-center on the landing page. Instead, they opted to put up photos of some victims. You have to click on a link in the navigation bar to see the map.
This humanizes every data point and concretizes the motivation of this project. It’s a reminder that at the end, this is about real people.
It is important when presenting data
Taking care to recognize the human factor in the data when visualizing it it is an important exercise. Our mind thinks in narratives. It functions much better when presented with stories.
A good illustration of this point is Periscopic’s Gun Deaths Visualization. Every data point on this visualization represents a person’s life, and is a link to the gun death’s news story.
The data is presented in a way that facilitates empathy for each individual. It takes a lot of care not to lose the individual in the aggregate statistics. It tells an impactful story.
Why bother?
The point of all this is to remind ourselves that the data is a representation of something in the real world. Regardless of how academic or impersonal the subject matter, at the end, it affects someone’s life or paycheck. It is our responsibility as data scientists to respect this fact.