One Way And Repeated Measures ANOVAS L8 Flashcards
What is the probability of a type 1 error occurring?
Whatever the alpha level is set as, typically a 5% chance.
What is the Bonferroni correction?
This adjusts the p value criteria based on the number of comparisons.
Equation = p value/no. of comparisons
The null hypothesis is then only rejected when p value is less than this. Considered a harsher way of doing it.
What is a one-way ANOVA?
Compares the means of different people/groups
One DV and one IV that has 2 or more levels of different participants
What is a repeated measures ANOVA?
Compare means of the same people
One DV and one IV that has 2 or more levels of the same participant
What are the assumptions of a one-way ANOVA?
Similar to independent samples T tests,
Treatments are randomly formed and are independent of each other
The DV must be measured on an interval or ratio scale
The data must be normally distributed
The samples should have approximately equal variances
What is the equation of the F value?
F = average between groups variance / average within groups variance
If variance between > variance within then F>1, if F>1 it might be significant
How is homogeneity of variance assessed?
Levene’s test, which looks at whether the variances in different groups are the same.
A significant result = variances are not equal <0.05
Non significant result = variances are equal >0.05
Welch and brown forsythe corrects for lack of homogeneity
What are typical effect sizes for partial eta squared (np2)?
Small = 0.01 Medium = 0.06 Large = 0.14
What are the assumptions of a repeated measures ANOVA?
- measurements are not independent of one another
- the data should be normally distributed
- the samples should have approximately equal variances (sphericity assumption)
What can be taken from looking at Mauchly’s test of sphericity?
If p>0.05 one must read from sphericity assumed
If p<0.05 one must read from greenhouse geisser
