Spring Intro 2023

16/02/2023

Type 1 error

The probability of a Type 1 error is the same as the α-level set - If α=0.1, the probability of Type 1 error = 10% - If α=0.05, the probability of Type 1 error = 5% - If α=0.01, the probability of Type 1 error = 1%

We usually set α at 0.05, so we routinely accept a 5% chance of making a Type 1 error…

…so 1 in 20 “positive” trials delivers a false positive

The Organisation for Professionals in Regulatory Affairs

53

Type 2 error

We see no difference when there is one

Type 2 error

All these mean the same thing

All these mean the same thing

Null hypothesis is accepted when it is false and should be rejected

False negative

β-error

An effective drug appears not to work

The Organisation for Professionals in Regulatory Affairs

54