People who don’t recognize the name Edward Tufte nevertheless are likely to know his most famous critiques—reports on how poor data presentation contributed to the space shuttle disasters.
Before the 1986 Challenger launch, NASA engineers had collected data indicating that rubber O-rings failed to seal correctly in cold weather. But Tufte says the charts that made their way up NASA’s chain of command failed to convey fully the dangers of a frigid-morning liftoff. The typography was sloppy. Cutesy icons of rockets obscured key numbers. (See below.)
 |
| Courtesy Edward Tufte |
Worst of all, the O-ring performance data was arranged by launch date instead of by the critical factor, temperature, making it all but impossible for decision makers to see that a launch in weather below 66 degrees probably would involve O-ring failure. (See bottom chart.) The launch proceeded when the temperature was 36 degrees, and the shuttle exploded 73 seconds after liftoff.
The 2003 Columbia accident represents the flip side of data debacle, Tufte says. When a chunk of insulation foam broke off during liftoff and hit the wing, engineers lacked the data to predict whether the damaged shuttle could safely re-enter Earth’s atmosphere. Tufte says a PowerPoint presentation obscured the technical ignorance and reassured NASA officials, tragically, that no rescue was needed.
 |
| Courtesy Edward Tufte |
Tufte loathes PowerPoint, whose format he says gives short shrift to evidence and encourages cheerleading, pitching and a preoccupation with form over content in ways that ultimately corrupt thinking. “The worst thing Microsoft has ever done is to replace the fundamental unit of analysis, the sentence, with the bullet, the grunt,’’ he says.
The day NASA posted its Columbia slides online, in response to a Freedom of Information Act request, Tufte stayed up most of the night furiously writing his analysis. Two years later, the Columbia Accident Investigation Board included much of his material in its final report. Not everyone agrees with his analysis, much less his full-bore antipathy to PowerPoint. Nobody, however, disputes Tufte’s basic assertions about the use and misuse of information.
