Please reply to the enquirer, [log in to unmask], and not to the messenger
(me).
---------------------------------------------------------------------------
Dear Allstat,
We have what is probably a very simple circular statistics problem that is
impeding the progress of a (visual psychophysics) project and I wondered
if you could give us any pointers. Here's the problem: the variance on an
estimate of the mean decreases in inverse proportion to the number of
samples but for wrapped data this isn't so; at high variability wrapping
means one need more samples. Is there an analytical form for this
dependence?
Thanks in advance for any assistance you can give
Sincerely
Steven Dakin
Institute of Ophthalmology
11-43 Bath Street
London EC1V 9EL
Lab: +44 20 7608 6988
Office: +44 20 7608 6924
Fax: +44 20 7608 6850
Web: <http://www.homepages.ucl.ac.uk/~smgxscd>
http://www.homepages.ucl.ac.uk/~smgxscd
|