Websites That Will Make You Fall in Love With Statistics

Mar 7, 2017

From Wall Street to the weather, statistics are a routine part of modern daily life. Statistical analysis is also critical to scientific advancement. And yet, even among scientists, there's a lack of understanding about how statistical tests work and what the results mean.

That's why Brown University senior Daniel Kunin created a new website called Seeing Theory. On one level, it's sort of an animated textbook. Each interactive module includes an explanation of the concept at work. But, as you watch data points rain down the screen or a random sampler jump around a multi-colored honeybomb, it can be easy to forget you're learning anything ... and just be mesmerized.

"[The idea] came from a class I took in the fall of 2015 about interacting with data," said Kunin. "I was asked to make a visualization of data from my own research. But I had no data of my own. So, I decided instead, to make a visualization of a statistical concept."

The project grew from there. Now, there are dozens of interactive modules in which users create their own data, and watch it being analyzed in real time.

Click over to Five Thirty Eight, and you'll get a different take on statistics. There, numbers are the backbone of compelling, human stories about everything from politics, to gun violence, and soon, panda sex.

"We want people to understand what the meaning behind these numbers is," says senior science writer Maggie Koerth-Baker. "I can toss numbers at you all day, but they don't necessarily make sense to you if you don't understand the story that is kind of attached to them and embedded in them."

"You need the narrative behind the statistics, and you need to know where the statistics came from and how those numbers were created. Because the story of how the numbers were created matters as much as what the number is."

Koerth-Baker points to the surprise many felt after the 2016 election as a key example of public misunderstanding of statistics. In addition, she says that understanding what we don't know - the statistics we don't have - is critically important, and often ignored.