3 ms·

Not quite on topic, but this headline is a good example of false precision in converted units. Did they precisely measure it to the foot? Of course they didn't.

Not quite on topic, but this headline is a good example of false precision in converted units. Did they precisely measure it to the foot? Of course they didn't. The article makes it clear it's 25 metres, which is probably an estimate within 5-10 metres. 80 feet would have been a more helpful conversion.

cbd1984 10y ago
"Hey, 98.6, it's good to have you back again... " One of the best-known numbers in biology came about due to a similar confusion over significant figures: > The first systematic measurements of human body temperature were performed by the German physician Carl Wunderlich. In 1861 he measured the temperatures of one million healthy individuals (a sample size that seems too large to be believed). The average value was reported as 37 degrees celsius. When converted this value becomes 98.6 degreed fahrenheit. So what's the problem? Wunderlich's value has only two significant figures while the converted value has three. The last digit (the "point six" at the end) should be regarded with great suspicion. Wunderlich's converted value should really be stated as "ninety eight point something" if one is being honest. Actually, if one is being honest and respecting the usual rules, it should be stated as 99°F; that is, rounded to two significant figures. (And, of course, we're not that simple: Core body temperature rises and falls naturally due to a number of factors even when we're completely healthy, so any single number as "the" healthy body temperature is, strictly, incorrect.) http://hypertextbook.com/facts/LenaWong.shtml http://hypertextbook.com/facts/LenaWong.shtml
- pauljohncleary 10y ago
  > Wunderlich's converted value should really be stated as "ninety eight point something" if one is being honest. or we could, you know, agree on a standard and just use that...
  - scotchmi_st 10y ago
    Or you could, you know, use Celsius like the rest of the world... ;)
    - pauljohncleary 10y ago
      I'm from the rest of the world and I do ;-)
    - cbd1984 10y ago
      Or use Kelvin like a scientist.
- cnvogel 10y ago
  Or you just write out the margin of error explicitly... 98.6±0.2F (or whatever). Especially with errors in the order of .1, you might even write out more than the traditional number of digits, e.g. 98.64±.10, as it might make a significant difference if you regard 98.54...98.64 or 98.5...98.6 to be within your confidence interval.
rmc 10y ago
When Mount Everest was first measured, it came in at 29,000 feet. But they released the figure of 29,002 feet, to prevent people thinking that it was just an estimate. https://en.wikipedia.org/wiki/Mount_Everest#Discovery https://en.wikipedia.org/wiki/Mount_Everest#Discovery

cbd1984 10y ago

"Hey, 98.6, it's good to have you back again... " One of the best-known numbers in biology came about due to a similar confusion over significant figures: > The first systematic measurements of human body temperature were performed by the German physician Carl Wunderlich. In 1861 he measured the temperatures of one million healthy individuals (a sample size that seems too large to be believed). The average value was reported as 37 degrees celsius. When converted this value becomes 98.6 degreed fahrenheit. So what's the problem? Wunderlich's value has only two significant figures while the converted value has three. The last digit (the "point six" at the end) should be regarded with great suspicion. Wunderlich's converted value should really be stated as "ninety eight point something" if one is being honest. Actually, if one is being honest and respecting the usual rules, it should be stated as 99°F; that is, rounded to two significant figures. (And, of course, we're not that simple: Core body temperature rises and falls naturally due to a number of factors even when we're completely healthy, so any single number as "the" healthy body temperature is, strictly, incorrect.) http://hypertextbook.com/facts/LenaWong.shtml http://hypertextbook.com/facts/LenaWong.shtml

pauljohncleary 10y ago
> Wunderlich's converted value should really be stated as "ninety eight point something" if one is being honest. or we could, you know, agree on a standard and just use that...
- scotchmi_st 10y ago
  Or you could, you know, use Celsius like the rest of the world... ;)
  - pauljohncleary 10y ago
    I'm from the rest of the world and I do ;-)
  - cbd1984 10y ago
    Or use Kelvin like a scientist.
cnvogel 10y ago
Or you just write out the margin of error explicitly... 98.6±0.2F (or whatever). Especially with errors in the order of .1, you might even write out more than the traditional number of digits, e.g. 98.64±.10, as it might make a significant difference if you regard 98.54...98.64 or 98.5...98.6 to be within your confidence interval.

rmc 10y ago

When Mount Everest was first measured, it came in at 29,000 feet. But they released the figure of 29,002 feet, to prevent people thinking that it was just an estimate. https://en.wikipedia.org/wiki/Mount_Everest#Discovery https://en.wikipedia.org/wiki/Mount_Everest#Discovery