Statistical Issues Of Design Flashcards
Type 2 Error
We do not reject the null
Null is false in the population
Probability = beta
Type 1 Error
We reject the null
Null is true in the population
Probability = alpha (.05)
We reject the null
The null is true in the population
Type 1 Error
We reject the null
The null is false in the population
Correct decision!
We fail to reject the null
The null is true
Correct Decision!
We fail to reject the null
The null is false in the population
Type 2 Error
The theory about decision errors
Fisher’s Classical Statistical Decision Theory
Factors the Affect Beta
Sample size
Alpha
Difference between means
variability of DV
Sample size
Sample size increases, beta decreases
Experimenter
Cost is a constraint
Solution is to test as many subjects as possible
Alpha
Alpha goes up, beta goes down
Experimenter’s fault
Tradition is a constraint
Solution is to make alpha .05
Difference between means
Difference increases, beta decreases
Nature’s fault
Relation of IV and DV are at fault
Choose your levels wisely
Variability of DV
Variability goes up, beta decreases
Nature’s fault
Due to participant homogeneity
Reduce variability of other factors