Problem with probability of hypothesis testing

43 Views Asked by At

Three coins are tossed. If all three show tails, $H_0$ is rejected, otherwise, $H_0$ is not rejected. Using this procedure, what is the probability of committing a Type II error?

I have that $1 - [P(\text{get tail 1st toss}|H_1\text{ true})*P(\text{get tail 2nd toss}|H_1\text{ true})*P(\text{get tail 3rd toss}|H_1\text{ true})]$

But I am not sure of this $1 - ((1/8)(1/8)(1/8))$

1

There are 1 best solutions below

0
On

Something is missing in the text. Assuming that the null hypothesis is

$$H_0: \theta_0\leq\frac{1}{2}$$

Against

$$H_1: \theta_1 >\frac{1}{2}$$

Where $\theta$ is the probability to get tail, type II error is

$$\beta=1-\mathbb{P}[\text{TTT}|\theta_1]=1-\theta_1^3$$

Observe that $\beta$ decreases when $\theta_1$ increases in the Interval $\left(\frac{1}{2};1\right]$ as it must be.