Re: Hardy-Weinberg lawt

From: Perplexed in Peoria (jimmenegay_at_sbcglobal.net)
Date: 07/01/04 

Date: Thu, 1 Jul 2004 15:46:20 +0000 (UTC)

"Anon." <bob.ohara@SOD.OFF.Spammers.helsinki.fi> wrote in message
news:cbumko$1r6i$1@darwin.ediacara.org...
> Perplexed in Peoria wrote:
> > "Anon." <bob.ohara@SOD.OFF.Spammers.helsinki.fi> wrote in message
> > news:cbpfed$8kv$1@darwin.ediacara.org...
> >
> >>Too much statistics - it simply isn't a distribution in the sense I
> >>would use the term. The only "frequency distribution" (in the sense
> >>that I would use the term) would be the tri-nomial distribution with
> >>cell probabilities {p^2, 2p(1-p), (1-p)^2}.
> >
> >
> > Bob, you might check your texts. According to what I read, you have
> > just described a binomial distribution (with n=2). The corresponding
> > trinomial distribution would have six cells:
> >
> > ( p^2, q^2, r^2, 2pq, 2pr, 2qr) with the constraint that p+q+r=1.
> >
> No, a binomial distribution has N events, with two possible outcomes per
> event (often called "success" and "failure", or if p=0.5, H and T). For
> H-W, we have three outcomes: two homozygotes and a heterozygote, and N
> is the size of the sample taken from the population.
>
> The number of copies on one of the alleles in an individual is
> binomially distributed, but H-W is a statement of the proportions of the
> three genotypes in the population (this is not to say that the "alleles
> per individual" interpretation is incorrect, but just that it isn't used).

Are you writing this before or after checking your texts regarding
"binomial distribution"?