Re: [PyrNet-L] SAS- Threat to Pyrs?

[Kshoffman@aol.com]

In a message dated 12/21/99 1:50:19 PM Eastern Standard Time, 
JGentzel@aol.com writes:

<<To make assumptions of "significant link" on a very small number of 
 individuals is making a huge jump in logic and is statistically invalid.  It 
 just cannot be done with any degree of reliability.>>

Back up the train a minute here, Joe. <vbg>

I said a *possible* significant link.  I also said: "a controlled scientific 
study and statistical analysis of the data, accounting for all possible 
variables, would be a more reliable indicator of whether or not decreasing 
levels of homozygosity (reduced coefficients of inbreeding) in Newfs is 
significantly statistically correlated to decreasing frequencies of SAS (or 
other defects) in the breed.  It would be nice if some credentialed 
professionals would take on such a study and test this hypothesis" 

This was based on Janice's assertion that statistics on the frequency and 
severity of SAS in Newfs might be available all the way back to 1968.  If so, 
that's hardly one or two cases. Yes, there are a number of variables that 
would have to be considered, but assuming the data is available, and assuming 
a cursory analysis of the data indicates both a reduction in average 
inbreeding coefficients and a reduction in frequency and severity of SAS, 
let's ask some statisticians and population geneticists if there is a valid 
hypothesis to test and analyze. If the experts say "no", then I will agree 
with you.<vbg>  Fair enough?

Kelley Hoffman
kshoffman@aol.com