Lecture 29- Multiple Testing

Question 1

What is multiple testing?

Answer 1

• Occurs in many cases where you are not just measuring one outcome but a number
• This requires multiple hypothesis tests

Question 2

What is the chance of getting a type 1 error for any individual hypothesis test and what does this mean?

Answer 2

• Alpha level (significance level)

- This means that for a 95% confidence interval there is a 5% chance of rejecting the null even though t is true.

Question 3

In multiple testing does the chances of getting a type 1 error change depending on sample size?

Answer 3

No

Question 4

Answer the question on slide 554. What R function did you use and why could you use this?

Answer 4

Answer on slide
dbinom, could use because there is two options either p is less than 0.5 or it is greater than 0.5 it is therefore a binary variable that can be represented by a binary distribution

Question 5

What happens to the chances of getting a false positive result when you do multiple tests in one day?

Answer 5

Chances are lot higher than alpha (the significance level) because tests may not be independent