RMA Week 11: statistical inference

read the slides to look at z-scores, hypotheses critical values and critical z-scores on distribution graphs

what is probability?

a measure of how likely it is that some event will occur

what does p vary from

0 (never) to 1 (always)

what is a type 1 error and how often does it happen?

rejecting H0 when we should not - false positive

• happens 0.05% of the time when p>0.05

what is a type 2 error?

rejecting H0 when we should - false negative

if we increase the threshold from p>0.05 to decrease the chance of type 1 error, what is more likely to occur instead?

a type 2 error

give an example of a directional (one-tailed) alternative hypothesis

the patient’s score will be lower than the scores of healthy controls

give an example of a non-directional (two-tailed) alternative hypothesis

the patient’s score differs from that of healthy controls

what would make you decide whether to choose a directional or non-directional hypothesis

it’s based on whether there’s been prior research! if you have an idea of what’s expected to happen, you’d choose directional. no research = no idea = non-directional!

what will H0 of directional and non-directional always say?

does not differ - don’t fall into the higher/lower opposite trap. H0 does not differ whether the hypothesis is directional or not