Stats 2

What is an easy way of remembering what a low P-value means?

"If the P is low, the null must go!" i.e. if the P-value is less than 0.05 it is likely there IS a difference.

Why is it important to consider the difference between clinical and statistical significance?

Sometimes a value might be statistically significant (i.e. one BP drug does lower BP 1% more than another) but might not have a clinical significance (1% more isn't enough to make a difference to the animals health)

How do we evaluate if a difference that we see in results is actually because there IS a difference, or just because of random 'chance' due to the sampling?

Using the 'p-value'

When would you use a null hypothesis?

When you want to compare two sets of data from two similar but different populations.

What is the critical p-value?

0.05 or 5%

What 3 factors do we need to take into consideration and feed into a computer when calculating sample size?

1. An estimate of individual variation 2. The level of difference we would like to detect 3. The desired power of the study

What three things influence the power of a study?

1. Sample size (bigger sample = more precise estimate) 2. Difference between the populations (the bigger the difference the more likely we are to spot it!) 3. Individual variation (if we have similar individuals in our population we are more likely to detect differences between the populations)

What is a null hypothesis?

A hypothesis that we are setting out to disprove, usually a "no-difference" or "no-effect" statement. i.e. the 'opposite' of what you expect to find.

What is a type I error in null hypotheses?

A mistake when scientists decide to reject a true null hypothesis - they decide there is difference between the populations when there is not.

What does 'power' means in terms of a type II error?

A powerful study is when there is a low chance of a type two error, i.e. a low chance we would reject a null hypothesis that is actually true.

What does it mean if the p-value researchers get for their experiment is less than 0.05?

It means that researchers conclude there is evidence of a difference between the two populations.

What is a type II error in a null hypothesis?

Scientists fail to reject a null hypothesis and decide there is no difference between two populations when in fact there is!

What is the P-value?

The probability/chance that the difference we have found in our samples would occur if out null hypothesis were true an d there WAS no difference between the populations.