NavList:
A Community Devoted to the Preservation and Practice of Celestial Navigation and Other Methods of Traditional Wayfinding
Re: Rejecting outliers
From: George Huxtable
Date: 2011 Jan 8, 16:14 -0000
From: George Huxtable
Date: 2011 Jan 8, 16:14 -0000
With regard to Peter Fogg's sequence of nine altitudes, Peter Hakel's computer analysis deduced, and then employed, a slope value of about 24' (in 5 minutes) whereas the evidence surrounding the taking of the observation implied that the slope couldn't in fact differ much from 32'. So I asked- Is Peter prepared to defend his value [of 24] against the other [of 32]? To which Peter Hakel has replied- You provided the best defense when you wrote: "So let's have a look at the data that Peter Hakel's estimate was based on, in the attachment. I would agree that the better fit, with no other information to go on than those plotted point, would be the continuous line at a slope of 24. Peter's program says so, an Excel fit says so, and my eye says so." But he seems to have missed the point I was trying to make, and perhaps I didn't make it clearly. What I was trying to say was that, in assessing the sequence of altitudes, all three, Peter's program, an Excel straight line best fit, and my own eye would agree that the true rate of change of altitude would have been 24' over 5 minutes. AND WE WOULD ALL THREE BE WRONG! Because, from the other information we had been given, that slope simply had to be 32. And indeed, a slope of 32 would be perfectly compatible with the data, though not the best fit to it. There was so much scatter in the altitudes, over that short observation time, that a wide range of slopes could fit it. Indeed, those observations provided a rotten basis on which to determine the slope. Which is why that analysis gives such a wrong answer for the slope. George. contact George Huxtable, at george{at}hux.me.uk or at +44 1865 820222 (from UK, 01865 820222) or at 1 Sandy Lane, Southmoor, Abingdon, Oxon OX13 5HX, UK.