Good afternoon Frank,
Although I failed statistics as an undergraduate, I later became fairly good
at it when I was involved in research. My memory seems to recall that the
minimum number of samples for a reasonable approximation of a normal
distribution was in the low twenties. But as I haven't had to worry about
minimum sample numbers for a while (my current set is > 5,000), I may well
be incorrect on this. Doubtless someone else on this list will gently
correct me!
But would Fitzroy really have known much about statistical distributions?
Perhaps more likely he was taking multiple backups, plus a few freebies?
During my field work I have learnt the hard way about backups. A notebook
computer completely died on me in the middle of outback Queensland, and I
had to spend hours reloading Windows and essential software. Now I travel
with two notebooks, two GPSs, etc. And this is for trips of only a few
months. If I'd been Fitzroy on a research voyage of several years, I'd have
made very sure I wasn't going to have chronometer problems
Cheers, John
John Pickard
john.pick...@bigpond.com
----- Original Message -----
From: "Frank Evans" <frankev...@zooplankton.co.uk>
To: "Sundial" <sund...@rrz.uni-koeln.de>
Sent: Thursday, June 23, 2011 6:51 PM
Subject: time, off topic
During Darwin's famous voyage aboard the "Beagle", Captain Fitzroy had 22
chronometers aboard, no doubt to obtain accurate longitudes. This seems
pretty excessive and I'm wondering how many (or few) chronometers would
have reduced his time errors to an acceptable level. Any thoughts? Poisson
distribution, perhaps?
Frank 55N 1W
---------------------------------------------------
https://lists.uni-koeln.de/mailman/listinfo/sundial
---------------------------------------------------
https://lists.uni-koeln.de/mailman/listinfo/sundial