NavList:
A Community Devoted to the Preservation and Practice of Celestial Navigation and Other Methods of Traditional Wayfinding
Re: Lunars: Jupiter's BIG.
From: Frank Reed CT
Date: 2003 Dec 24, 03:01 EST
From: Frank Reed CT
Date: 2003 Dec 24, 03:01 EST
FredACF said:
"With a regression line, it is important that a point near the middle of the line be chosen, and, in general, averaging is much better than these graphical methods for determining the time."
I may have posted on this already, but I don't think so. The real reason that linear regression is not useful with a set of a half-dozen lunars is because the individual data points are not really independent trials. When you shoot a lunar, you react to the last lunar you shot. For example, if I feel my hand was shaky on the last trial, I will change my bracing technique on the next. If I think, after the fact, that I made contact too close on the last lunar, I will bias a little wide on the next one. After a set of five or six individual sights over ten to fifteen minutes usually, you have data that you can draw a line through, but you can and should bias that line based on your best guess of the quality of your sights. Statistically, you should never ignore an outlier, but if you remember your eyes watering and a sudden icy breeze on a sight that ends up way off the line when plotted, it is reasonable to drop it. These sights are not independent trials.
Frank E. Reed
[X] Mystic, Connecticut
[ ] Chicago, Illinois
"With a regression line, it is important that a point near the middle of the line be chosen, and, in general, averaging is much better than these graphical methods for determining the time."
I may have posted on this already, but I don't think so. The real reason that linear regression is not useful with a set of a half-dozen lunars is because the individual data points are not really independent trials. When you shoot a lunar, you react to the last lunar you shot. For example, if I feel my hand was shaky on the last trial, I will change my bracing technique on the next. If I think, after the fact, that I made contact too close on the last lunar, I will bias a little wide on the next one. After a set of five or six individual sights over ten to fifteen minutes usually, you have data that you can draw a line through, but you can and should bias that line based on your best guess of the quality of your sights. Statistically, you should never ignore an outlier, but if you remember your eyes watering and a sudden icy breeze on a sight that ends up way off the line when plotted, it is reasonable to drop it. These sights are not independent trials.
Frank E. Reed
[X] Mystic, Connecticut
[ ] Chicago, Illinois