Probability; type 1 and type 2 error and sample size

Question 1

What assumption is the p value calculated under?

Answer 1

The assumption that the null hypothesis is correct

Question 2

What is the p value?

Answer 2

It gives the probability that result tested occurred by chance

Question 3

What is type 1 error?

Answer 3

Rejecting the null hypothesis when it is correct (False positive)

(Overall positive results)

Question 4

What is type 2 error?

Answer 4

Failing to reject the null hypothesis when it is correct (False negative)

(Overall negative results)

Question 5

What is type 1 error designated by?

What is it set to?

Answer 5

Represented by alpha level (α)

0.05

Question 6

What data is type 1 error

Answer 6

Higher values

Multiple tests

Question 7

How can type 1 errors be reduced?

Answer 7

• By setting a lower value 0.01 (1% probability)

- Reporting p values to 3 decimal places to give more accurate probability estimates

Question 8

What data is type 2 errors more likely with

Answer 8

Small samples

Small effect size

Question 9

What is type 2 error designated by?

Answer 9

Beta

1-beta (power of a test)

Question 10

How can type 2 errors be prevented?

Answer 10

Using a large sample size

Larger effect size (this helps to consider what is ‘clinically meaningful)

Question 11

How does relaxing the alpha level (e.g. p=0.01) have an effect on type 1 and 2 errors?

Answer 11

Increases type 1 errors (false positives)

Decreases Type 2 errors (false negatives)

Question 12

how does tightening the alpha level (e.g. p=0.01) have an effect on type 1 and 2 errors?

Answer 12

Decreases type 1 errors (false positives)

Increases type 2 errors (false negatives)