Bernoulli trials

• November 12th 2010, 02:38 PM
wik_chick88
Bernoulli trials
a series of n Bernoulli trials is to be observed as data for testing $H_{0} : p = \frac{1}{2}$ versus $H_{1} : p > \frac{1}{2}$.
suppose that the null hypothese will be rejected only if k, the observed number of successes equals n.
for what value of p will the probability of making a Type II error equal 0.20?
• November 12th 2010, 10:59 PM
CaptainBlack
Quote:

Originally Posted by wik_chick88
a series of n Bernoulli trials is to be observed as data for testing $H_{0} : p = \frac{1}{2}$ versus $H_{1} : p > \frac{1}{2}$.
suppose that the null hypothese will be rejected only if k, the observed number of successes equals n.
for what value of p will the probability of making a Type II error equal 0.20?

You are being asked to find $p$ such that the probability $pr(k=n)=0.2$, where $pr(k=N)=b(N,n,p)$, so $pr(k=n)=b(n,n,p)=p^n$

CB
• November 28th 2010, 12:30 AM
wik_chick88
well u dont know k or n, so does that mean $p = \sqrt[n]{0.2}$, depending on what n is?
• November 28th 2010, 05:40 AM
CaptainBlack
Yes.

CB
• December 15th 2010, 05:56 AM
wik_chick88
• December 15th 2010, 06:15 AM
CaptainBlack
Quote:

Originally Posted by wik_chick88