From: Steve Hilliard (hilliard@zookeeper.zoo.uga.edu)
Date: Tue Aug 31 1993 - 08:05:22 EST
As the subject line indicates, I need some stats advice. I've got a major user that has spent a lot of time working on a project exploring the utility of increased CV as an indicator of genetic damage (on the population level in animals). They've done dozens of experiments to make sure they've accounted for staining variability, etc. The problem is, now that this student's committee is reviewing the work, one member say you can't do ANOVA's on the CV's to test for differences between the treatment groups. He says that since CV is a ratio, it can't be analyzed by ANOVA, unless the numerator and denominator are unrelated. Obviously SD and MEAN are related--that's why we calculate CV. I've suggested she talk to members of our stats dept. but does anyone have any feelings/prior experience on this subject. Is there some other approach that would be valid? Thanks in advance for any replies, sgh Steve G. Hilliard | Cell Analysis Facility, Univ. of GA | I brew the beer I drink! (706) 542-9474 hilliard@zookeeper.zoo.uga.edu| Let's have a tasting These opinions are mine--get your own! | at Lake Placid!
This archive was generated by hypermail 2.1.6 : Thu Jan 01 2004 - 17:25:58 EST
