Statistical Power Statistical Power n Definition The probability
Statistical Power
Statistical Power n Definition: The probability that will reject a false null hypothesis
Decision process Population Decision H 0 is true H 0 is false Keep H 0 Correct (1 -a) Type II error (b ) Type I error (a ) Correct (1 -b) Reject H 0 Power
Statistical inference a (1 -a) m = 72 (1 -b) b m = 69
Statistical inference a (1 -a) m = 72 (1 -b) b m = 70
Statistical inference a (1 -a) m = 72 (1 -b) b m = 64
Effect size (One group) n Definition : It is the difference between the null hypothesis and the alternative hypothesis DI = 0. 2 (small effect) DI = 0. 5 (medium effect) DI = 0. 8 (large effect) n It gives us an idea of the magnitude of the difference that we want to detect. (Treatment effect)
Effect size (Two groups)
Effect size and power n Example: n To know the power we need to use a software like G*power.
Effect size and power
Factors influencing the power n n Signification level The magnitude of the treatment effect The variability within the population Sample size
Factors influencing the power n n n Signification level The magnitude of the treatment effect The variability within the population Sample size The more a increases the more the power increases.
Factors influencing the power n Signification level n The magnitude of the treatment effect n n n The variability within the population Sample size The higher the effect of treatment is the higher the power will be
Factors influencing the power n Signification level The magnitude of the treatment effect n The variability within the population n Sample size n n The less variability in the population the higher the power will be
Factors influencing the power n Signification level The magnitude of the treatment effect The variability within the population n Sample size n n n If n increases the power will increases.
Sample size estimation n n Type I and II errors Effect size
- Slides: 16