[LEAPSECS] Imprecise inaccuracies

[Rob Seaman]

On Apr 13, 2011, at 1:47 PM, Ask Bjørn Hansen wrote:


> On Apr 13, 2011, at 12:39, Rob Seaman <seaman at noao.edu> wrote:

> 

>> ...bearing in mind that "everyone" includes the nearly seven billion people who are completely unaware that a discussion is occurring and who would be perplexed to understand that the idea here is that the clocks on their wrists, walls, computers, and bedside stands will have nothing whatsoever to do with the obvious diurnal cadence of their lives...

> 

> For "human perception" this minute difference will be completely lost in the noise compared to the seasonal changes.  The "diurnal cadence" is also shifted significantly by moving, traveling, etc.  Most places the sun isn't on the top of the sky at noon, anyway.

> 

> I think you'll find more people who accurately set their watches be surprised that every few years time jumps (seemingly randomly) by a second.


Asked and answered a hundred times over since 1999.  Apparent solar timekeeping is a red herring.

Focus instead on the word "accurately" here.  What does this word mean?  Rather than look online, let's consult an actual book.  On p. 9 of "Data Reduction and Error Analysis for the Physical Sciences" (P. R. Bevington, McGraw-Hill, 1969) we have:

	'Accuracy:  Measure of how close the result of the experiment comes to the "true" value.
	Precision:  Measure of how exactly the result is determined without reference to any "true" value.'

(Double quotes in the original.)

The ITU is asserting that the true value doesn't matter, they are not asserting that there is no true value.  If they were doing the latter, they could simplify things even further by declaring there are 100,000 SI-seconds per day.

Rather they are saying that accuracy and precision can be smooshed together into the same concept.  Good luck with that.

(I see that Steve has already answered with a more parsimonious response - as usual :-)

Rob