On 25 Sep 2000 16:26:30 -0700, [EMAIL PROTECTED] (Ronny
Richardson) wrote:
> What follows is a question from the test bank for Complete Business
> Statistics by Amir Aczel. I am embarrassed to say that I cannot figure out
> how to work the problem. I do know from the test bank program that the
> answer is 798. Any hints or solutions would be greatly appreciated.
>
> "A real estate salesperson wants to prove that 65% of all home-owners
> change their house in less than 6 years. This salesperson wants to have a
> 90% probability of success if the true percentage is 60%. If the
> hypothesis test is carried out at a 5% level of significance, what should
> be the minimum sample size in a survey conducted to prove this claim?"
>
This somewhat resembles a question in statistical power.
I don't know what standards of "proof" your textbook endorses,
but I would not subscribe to them. That is, I believe that there is
a serious semantic problem.
I won't try to guess what question was intended, or fiddle to see what
problem happens to require "798."
Rephrased:
Draw a sample of size N, which will 90% probability to
prove that a fraction X=65%
when, in fact, the true X=60%
and do that (somehow) using a "5% level of significance."
We don't try to "prove" the palpably or presumably untrue.
--
Rich Ulrich, [EMAIL PROTECTED]
http://www.pitt.edu/~wpilib/index.html
=================================================================
Instructions for joining and leaving this list and remarks about
the problem of INAPPROPRIATE MESSAGES are available at
http://jse.stat.ncsu.edu/
=================================================================
