NavList:
A Community Devoted to the Preservation and Practice of Celestial Navigation and Other Methods of Traditional Wayfinding
From: Andrew Nikitin
Date: 2013 Jun 24, 18:19 -0700
Antoine,
> i=-0.17
> i=+0.22
> i=+0.21
> i=+0.36
> i=+0.15
Looking at intercepts only they look pretty one sided. This sample has 0.154 as sample mean and s=0.2 as dispersion. Assuming "true" i is 0 and using Student-t with 4 degrees of freedom: t(-0.154/(S/SQRT(4)))~0.1: there is only 10% that such a big deviation is due to random chance only. Some people might imply that there is 90% probability that the sample indicates presence of positive systematic error in altitude measurements or calculation, but we are not those people.
In my laymans opinion, 10% is borderline case where you cans either shrug it off or get 10 more samples to see where it will take you or look at the measurement setup more carefully and try to imagine where systematic error may come from.
----------------------------------------------------------------
NavList message boards and member settings: www.fer3.com/NavList
Members may optionally receive posts by email.
To cancel email delivery, send a message to NoMail[at]fer3.com
----------------------------------------------------------------