Week 6: Factorial Repeated Measures Design (within and mixed) Flashcards
What does factorial mean?
More than one
What are the design possibilities when you have a factorial ANOVA?
- Between subjects
- Within subjects
- Mixed (SPANOVA - split plot)
Explain fixed or random factors?
Variables can be either fixed or random
Fixed: IV levels are specifically chosen and do not change randomly from one experiment to the next
Random: levels are randomly selected from a larger population of potential levels. They would differ from one replication to the next
Does sphericity still apply?
Yes - if there are more than 2 levels it is likely that sphericity has been violated and corrections should be applied
For a within-subjects factorial RM design, how would you set the data out in jamovi?
1x row for each participant
1x column for each level of each IV and the DV
How do you do a completely within-subjects factorial ANOVA?
ANOVA - RM ANOVA
IV - RM factors
Levels - RM cells
DV - DV
Ask for descriptives, sphericity
Plots (EMM)
What is the output of a completely within RM ANOVA in jamovi?
2 main effects, and an interaction effect
Generating a plot for a within RM ANOVA assists what?
It assists in the interpretation of the interaction
If you have a significant interaction, what do you do next (within RM factorial)?
Investigate further using paired t-tests
- Bonferonni adjustments
Instead of using paired t-tests to investigate a significant interaction, what can you do? (within RM factorial)
Do 2 separate one-way repeated measures ANOVAs - do this using the filtering procedure
Do Bonferroni adjustments to see if it is significant
If so, need post hoc tests (Bonferroni or holm) to see where exactly the effect is occurring
What is a mixed factorial design?
Have a combination of between and within IVs
How do you set out data in jamovi for a mixed factorial ANOVA?
1x row for each participant
1x column for every between IV, every within level and the DV
How would you begin the analysis for a mixed factorial ANOVA in jamovi?
Whenever there is a within-subjects IV, the analysis must be performed using the repeated measures ANOVA option
Within variables are then put into RM factors box and their levels into RM cell box
Between variables are put into between subjects box
DV - DV
Do we need to look at sphericity or homogeneity in a mixed factorial design?
Both.
For within we need to look at mauchly’s test well as considering how many levels the variables have. Greenhouse corrections may be needed and have to be done on both the main effect AND the interaction
For between, levenes test still applies
If a mixed factorial ANOVA has a significant interaction, what should you do next?
You have 2 options:
1: You can do independent samples t tests. Make the grouping variable the between IV and fill the DV box with the within variables
Bonferonni adjustments need to be made.
2: You can do separate one-way repeated anovas for each of the levels of between IV
- Bonferonni adjustments and sphericity considerations
- can do post hocs to see individual levels of within IV on that level of the between IV