Testing a Hypothesis

Null hypothesis and level of signification

P-Value

Confidence Interval

Significant Treatment Difference

Power and Sample Size

Test Statistic

Null Hypothesis

p1 = probability of positive outcome for operative group

p0 = probability of positive outcome for non-operative gruop

If H0 is true, the sampling distribution will be normal

Level of signification

Alpha represents the rejection region willing to be accepted

Probability of a type I error

Critical value

Start of the rejection region

Scaled version of the sampling distribution

Critical Value

Helps to accept of reject H0

Proportion of repeated samples under H0

If p < a there is enough evidence to reject H0

If p > a there is not enough evidence to reject H0

Constructed around the test statistic

Represents the proportion that contains the true population difference

Prove with a certain number of samples to test the true difference

Contains the whole range of values and also the proportion of the confidence interval

H0 is true

True negative: do not reject H0

False positive: reject H0 (type I error)

H0 is false

False negative: do not reject H0 (type II error)

True positive: reject H0 (power)

Power: probability of rejecting a false H0