Spring Intro 2023

16/02/2023

Sample size (number of patients)

Based on statistical assumptions (best done by a statistician)

Magnitude of expected treatment effect

Variability of expected treatment effect

Required power to detect a real

Level of significance testing

treatment difference

α

β

The Organisation for Professionals in Regulatory Affairs

51

Type 1 error

We see a difference when there isn’t one

Type 1 error

All these mean the same thing

All these mean the same thing

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

False positive

α-error

An ineffective drug appears to work

The Organisation for Professionals in Regulatory Affairs

52