Lecture 15: Repeated & Mixed ANOVA

Question 1

What is a repeated & mixed ANOVA

Answer 1

An analyses of the variance of the model while reducing error by the within person variance

Question 2

What number of dependent/independent variables does the one-way repeated measures ANOVA have

Answer 2

• 1 DV/outcome variable
• 1 IV/predictor variable
  —> 2 or more levels
• all with the same subjects

Question 3

What number of dependent/independent variables does the factorial repeated ANOVA have

Answer 3

• 1 DV/outcome variable
• 2 IV/predictor variables
  —> 2 or more levels
• all with the same subjects

Question 4

What number of dependent/independent variables does the mixed design ANOVA have

Answer 4

• 1 DV/outcome variable
• 2 or more IV/predictor variables with different subjects
  —> 2 or more levels
• 1 or more IV/predictor variables with the same subjects
  —> 2 or more levels

Question 5

What are the assumptions of a repeated/mixed ANOVA

Answer 5

1. Uni- or multivariate
2. Continuous dependent variable
3. Normally distributed
4. Equality of variance of the within-groups differences (variances of group differences)
   —> Maulchy’s test of Sphericity
   —> this is always met when having only two groups because there’s only 1 variance (between the two groups)

Question 6

How does a repeated or mixed ANOVA differ

Answer 6

It is used for data sets in which the same people are measured twice (paired samples), so the same people in all conditions

Question 7

How many means do we have if we predict with the participant means

Answer 7

As many as how many participants there are

Question 8

What is the rule regarding how many parameters a model can have

Answer 8

Not more than the amount of data points in the data set

Question 9

What is the difference between a wide and a long data format and which one do you use for RM ANOVA

Answer 9

In a wide data format every row is one participant, in the long data format every row is one data point
For RM ANOVA we use the wide data format

Question 10

Why do we use the same people for all conditions and include the IV in the analyses

Answer 10

Because people have different baselines, which could influence the results if for example by luck all people that are really smart are in one condition and the really dumb ones in the other

Question 11

What is a downside of the RM ANOVA

Answer 11

You have many parameters (one for each participant) and you can’t have more than the amount of data points, plus it makes the model less easy to interpret which is the opposite of what we are striving for