In article <3582AC1E.370AB128 at perry.net>, Chris Perry <chris at perry.net> wrote:
>I have recently been running some experiments with
>drosophilia with the apterus phenotype. I have been getting
>some rather unusual results. I have been using chi square
>tests to determine the chances of my results being
>coincidental or being of something more like human error or
>variables in the environmental conditions. [...]
You'd probably be better off repeating the experiments and
obtaining more data. Statistical methods only go so far with
small sample sets.
Mark D. Garfinkel (e-mail: mg16 at midway.uchicago.edu. Ask for PGP public key)
(c) 1998; all rights reserved. Permission granted for Usenet quotation.
Fed up with unsolicited commercial e-mail? Join CAUCE at http://www.cauce.org