Statistical hypothesis testing 1: Introducing hypothesis testing

Question 1

Describe hypothesis testing

Answer 1

uses data from a sample to test whether a statement about a whole population is believable or not

Question 2

H0 and H1

Answer 2

• H0 represents the null hypothesis
• H1 represents the alternative hypothesis

Question 3

Method for conducting a hypothesis test

Answer 3

1. establish H0 and H1
2. decide on the significance level
3. use the data to find the test statistic
4. calculate the p-value and compare with significance level
5. if p-value < significance level reject H0, else don’t reject H0
6. state a conclusion with reference to the context of the question

Question 4

If the test statistic is small…

Answer 4

find the P(X<= test statistic)

Question 5

If the test statistic is big…

Answer 5

find the P(X>= test statistic)

Question 6

One tail test

Answer 6

• test either the bottom end or top end of the tail
• P > … or P < …

Question 7

Two tail test

Answer 7

• the significance level is halved and tested on both ends of the tail
• P ≠ …

Question 8

Hypothesis testing conclusion

Answer 8

• There is / isn’t sufficient to reject H0
• There is / isn’t sufficient evidence to suggest that … (refer back to context)