Big Data, Blindspots, and Bad Statistics

As some of you know, last spring I wrote a contrarian piece for The Chronicle of Higher Education that posed some cautions to unabashedly embracing big data.  Since then, I’ve found two Ted Talks that add to the list of reasons to be suspicious of an overreliance on statistics and big data.

Tricia Wang outlines the dangers of relying on historical data at the expense of human insight when trying to anticipate the future.

Mona Chalabi describes three ways to spot a suspect statistic.

Both of these presenters reinforce the importance of triangulating information from quantitative data, individual or small-group expertise, and human observation. In addition, all of this information can’t eliminate ambiguity. Any assertion of certainty is almost always one more reason to be increasingly skeptical.

So if you think I’m falling victim to either of these criticisms, feel free to call me out!

Make it a good day,

Mark