1st Dec

Question 1

What is the connection between estimation and hypothesis?

Answer 1

Estimation allows you to estimate an effect, hypothesis testing will tell you how unlikely you are to see an effect this big by chance if there is no genuine effect

Question 2

What is the relevance of confidence interval?

Answer 2

If the 95% confidence interval is small we are more certain of our estimate
If the 95% confidence interval does not include the null then the association is significant
As we increase sample size, our estimates become more accurate, so our confidence intervals shrink

Question 3

What will make our CI smaller?

Answer 3

Less variation in data

Larger sample size

Question 4

What is the power of a study?

Answer 4

Probability of rejecting the null hypothesis when the null hypothesis is false ie the probability of finding an association if it is there to be found

Question 5

What makes a greater power?

Answer 5

Mean effect is bigger
Variation in the effect is smaller
Your sample is bigger

Question 6

When do you calculate power?

Answer 6

Before conducting a study: to make sure it is well-designed and to justify your study
After and when no association found: to demonstrate you were sufficiently well powered to detect an association, to detect how small the association is

Question 7

What is the odds of an event?

Answer 7

Probability an event occurs divided by probability an event doesn’t occur

Question 8

What is an odds ratio?

Answer 8

Odds of an event for an exposed group divided by odds of an event for unexposed group
If no effect, odds ratio = 1 (the null)

Question 9

What is a 95% CI?

Answer 9

95% of the time the effect is true