Unit 5

Question 1

What is the goal of hypothesis testing?

Answer 1

To quantify how much it would take to prove that our previous idea about what was going on was false

Question 2

What is the standard value for the cutoff point?

Answer 2

5%

Question 3

What is a null hypothesis?

Answer 3

A statement that a population parameter takes on a specific value that we would like to test. It is always stated with an equality, that there is no difference. It is the starting point of our investigation. (H0: parameter = number)

Question 4

What is an alternate hypothesis?

Answer 4

A statement that makes an opposing claim that we are trying to support (HA: parameter > number, parameter < number, parameter != number)

Question 5

What is the decision rule?

Answer 5

Give a precise statement of what must happen to reject H0:
- reject H0 if Zc > Za
- reject H0 if Zc < -Za
- reject H0 of Zc < -Za/2 or Zc > Za/2

(replace Z for t if you are testing mu with an unknown sigma)

Question 6

What is a critical value?

Answer 6

Your cutoff points that denote the most extreme a% of your sampling distribution relative to Ha

Question 7

What should be the form of your conclusion?

Answer 7

• As (rejection rule met), reject H0. Conclude there is sufficient evidence for Ha
• As (rejection rule not met), fail to reject H0. Conclude there is sufficient evidence for Ha

Question 8

What are the steps for hypothesis testing?

Answer 8

1. Level of Significance (set your alpha value)
2. Hypotheses (null hypothesis and alternative hypothesis)
3. Decision Rule (give a precise statement of what must happen to reject null hypothesis)
4. Test statistic (determine the z-score assuming null hypothesis is true)
5. Critical value (find cutoff points)
6. Conclusion (state your decision)

Question 9

What is the p-value interpretation?

Answer 9

the p-value in a hypothesis test is the probability of observing a statistic as or more extreme than we did (extreme being relative to HA) if H0 were true

Question 10

When can we use a confidence interval to perform a test of hypothesis?

Answer 10

if:
1. we are performing a two-tailed test
2. The level of significance and confidence level sum to 1

Question 11

How can you use a confidence interval to make a hypothesis test?

Answer 11

Instead of making a test statistic we will make our confidence interval and use the following decision rule:
Reject H0 if mu0 falls outside of the confidence interval