Answers from my updated Capen Quiz:
Don’t read any further until you’ve taken the quiz! Be distracted by the little bird. Scroll no further.
- 79.75 meters (261 feet 8 inches) http://www.airbus.com/aircraftfamilies/passengeraircraft/a380family/a380-800/specifications/
- 384,400 kilometers (238,855 miles) http://en.wikipedia.org/wiki/Lunar_distance_%28astronomy%29
- 1957 http://history.nasa.gov/sputnik/
- 410 AD http://en.wikipedia.org/wiki/Sack_of_Rome_%28410%29
- 714 http://en.wikipedia.org/wiki/500_home_run_club
- 1,389,000 http://arstechnica.com/apple/2007/10/apple-q407-financials-triumph-of-the-steve/
- 3,100,000 http://en.wikipedia.org/wiki/Transistor_count
- 40.1 million http://population.govt.nz/myth-busters/myth-8.aspx
- 43 mm (1.68 inches) http://en.wikipedia.org/wiki/Golf_ball#Regulations
- 105 meters (344 feet) http://www.worldwaterfalldatabase.com/waterfall/Victoria-Falls-111/
How did you do? A perfect score is 8 out of 10 right. Recall that I told you up front that people are typically overconfident, so hopefully you got at least half right. If you got less than 2 right, Barclays needs a new CEO.
I guess I will check out Barclays jobs available. Richard
At least you’ll be in good company …
https://metasd.com/2012/07/the-capen-quiz-at-the-system-dynamics-conference/
When conducting the test, how do encourage people not to be too broad in their answers – e.g. saying 1 inch to 3 trillion miles for the moon/earth distance?
For the record I got 6 right. Not quite CEO material.
May I post this survey on my web site. I will of cite you as the source.
6 is pretty good.
Feel free to post.
In practice, I haven’t seen people choose ridiculously wide bounds. I’m not sure how to prevent it though. That was one of the reasons that I chose to lower the target confidence interval from 90% (Capen’s original, though he tried others) to 80%. When I did this at the SD conference, I did point out rather dramatically that getting all 10 right was a sign of failure (though really, it’s plausible to have well-calibrated bounds and get 10 right by good luck).
One possibility would be to target a 50% interval, which would make 10 right a more obviously bad outcome. But then the overconfidence wouldn’t be as visible.
It would be an interesting experiment to ask people to do this in small groups. Might take longer though (10 minutes is about right for individual work).