The Flaw of Averages: Why We Underestimate Risk in the Face of Uncertainty

CHAPTER 41

Visualization

Two very young kittens are harnessed in a vertical drum with patterns painted on the inside. One kitten is able to walk on the floor of the drum, and the other is suspended just above it. Both kittens are presented with the same visual environment, but only the one on the floor can interact with it; that is, the drum rotates when it walks. By the end of the experiment, the suspended kitten requires a seeing eye dog. It has not learned how to process visual information and is effectively blind.

This experiment is described in Action in Perception by the philosopher Alva Noë, who argues that without action on the part of the observer, there can be no perception.¹ The retina is not just a digital camera that sends multi megapixel snapshots off to the brain for interpretation. According to Jeff Hawkins and Sandra Blakeslee, in On Intelligence, the eye explores a scene by moving its gaze around, much in the way your hand would explore the top of your desk in a darkened room by groping around and bumping into things.²

Visual Statistics

So how should one explore statistical data? That’s easy, because there was an undisputed master of the subject. In 1977 John W. Tukey (1915-2000) published Exploratory Data Analysis, a book on interpreting statistical data.³ Tukey was the author of many enduring Mindles, including the Box Plot, Stem and Leaf diagram, and even the words “bit” and “software.”

Today, powerful visualization software continues in the Tukey tradition. ...