home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: sci.math.stat
- Path: sparky!uunet!cis.ohio-state.edu!zaphod.mps.ohio-state.edu!darwin.sura.net!spool.mu.edu!agate!iat.holonet.net!uupsi!psinntp!isc-newsserver!jsvrc
- From: jsvrc@rc.rit.edu (J A Stephen Viggiano)
- Subject: Re: approx. of binomial dist., help needed
- Message-ID: <1992Oct8.202310.10461@ultb.isc.rit.edu>
- Sender: news@ultb.isc.rit.edu (USENET News System)
- Nntp-Posting-Host: bruno.rc.rit.edu
- Organization: RIT Research Corp
- References: <1992Oct7.025857.28031@news2.cis.umn.edu>
- Date: Thu, 8 Oct 1992 20:23:10 GMT
- Lines: 27
-
- In article <1992Oct7.025857.28031@news2.cis.umn.edu> lee@hecto.cs.umn.edu (YoungJun Lee) writes:
- >
- > The stat. book I have just says that
- > "Binomial distribution with n trials and success probability p can be
- > approximated by possion distribution(when n is large, p is very small, and
- > np is of moderate magnitude" and normal distribution(when n is large and
- > p is not too near 0 or 1)"
-
- For the Poisson approximation, it is much more important for p to be small
- than it is for n to be large. The approximation is quite acceptable for n = 1
- with p <= 0.01. This can be seen by looking at the relative error in the
- variance; it involves only p.
-
- With n = 30, a p of 0.05 or less is usually considered satisfactory.
-
- > What's large n ?
- > What's "the very small p" or "not too near 0 or 1"?
- > What's np of moderate magnitude ?
-
- For the gaussian approximation, we were taught that the tails of the
- approximating distribution should be well within the limits of 0 and n. Thus,
- np - 5 (npq)**0.5 should be greater than zero, and np + 5(npq)**0.5 should
- be less than n, where q = 1 - p. In other words, the mean should be at least
- five standard deviations from both zero and n. Again, this is a rule of
- thumb. For p = 0.5 (the best case), n should be at least 25. Smaller or
- larger values of p will require a larger n to reduce the coefficient of
- variation.
-