[EMAIL PROTECTED] writes:
>As a statistically ignorant engineer I have a problem in
>establishing an Upper limit based on a restricted set of
>data samples. The problem is described below :-
>
>An infinite, normally distributed population is sampled
>and N samples are collected with mean xbar and variance
>s^2.
>
>What Upper Limit UL can be placed on the population such
>that one can be confident (with greater than say 99%
>confidence) that any future samples will be less than UL
>when
>1. N < 10
>2. 10 < N < 30 and
>3. N > 30
It sounds like you want a prediction interval or possibly a tolerance issue.
these things are not discussed very well in most stat books. Check out Hahn
and Meeker for an excellent overview of what these intervals are and how and
when to use them.
Hahn, Gerald J. and Meeker, William Q. (1991) Statistical Intervals. A Guide
for Practitioners. New York NY: John Wiley and Sons, Inc. (ISBN:
0-471-88769-2). For the intermediate reader. The authors discuss two
important alternatives to the confidence interval: the tolerance interval
and the prediction interval. These intervals are useful when you wish to
characterize a specified fraction of your population (tolerance interval) or
if you want to predict one or more future individual observations
(prediction interval). The authors do an excellent job outlining when to use
which type of interval and they provide formulas for a variety of cases.
Steve Simon, [EMAIL PROTECTED], Standard Disclaimer.
STATS: STeve's Attempt to Teach Statistics. http://www.cmh.edu/stats
=================================================================
Instructions for joining and leaving this list and remarks about
the problem of INAPPROPRIATE MESSAGES are available at
http://jse.stat.ncsu.edu/
=================================================================
