3 Comments

"Take, for example, the Space Shuttle Challenger disaster of 1986. Seventy-three seconds after liftoff, the shuttle exploded, tragically ending the lives of seven astronauts and costing taxpayers billions of dollars. The prior day, engineers working on the rocket had sent charts to NASA with evidence that it was not safe to launch due to the forecast of cold weather and the way that o-rings (circular seals used in the shuttle) responded to cold temperatures. Unfortunately, and despite the noble efforts of these engineers, the story that higher-ups took away from these charts was not one of serious danger."

There's another way in which this is similar to the Kahneman / Deaton data, what's not shown in the data is that the amount of hot gas that passed the O-rings was much higher on 51C (the coldest) than was seen on other flights, and that on 61A much less (as indicated by the amount of soot in the space between the primary and secondary O-ring. This indicated that the O-rings sealed more quickly and completely on the latter, despite the fact that three distinct field-joint O-rings were bypassed by hot gas.

That information was discussed verbally on the call, so the people at NASA who made the launch decision were aware of it, and certainly Thiokol management who changed their minds from recommending against a launch to recommending it went ahead were aware of it because the person who brought it up was in the room with them. However it was not shown on the chart.

Expand full comment

Very interesting and illuminating article on how data visualisation and statistical analysis can be interpreted and presented in 'stories'. And the pitfalls of 'cherry picking' the data to suit the stories!!

Expand full comment

I think

"In short, some people are wired such that even favorable changes in their circumstances cannot make them measurably more happy over a certain point"

should read,

"In short, some people are wired such that even favorable changes in their circumstances cannot make them measurably less unhappy beyond a certain point."

More awkward, but it makes the distinction you're trying to make, whereas the first version re-instates the researchers' original confusion.

Expand full comment