Lecture 6 - factorial ANOVA Flashcards
ANOVA can simultaneously compare two means. What else can it do?
It can also allows us to test more than one independent variable at a time and see how they interact with each other
When would you use a factorial ANOVA?
When you have more than one independent variable
Want to know if the means are really different or if the differences can be explained by error variability
Want to know if your independent variables interact with each other
Factorial ANOVAs increase what in our study?
Generality of results - more complex experiments reflect the real world
What are the main effects in a factorial ANOVA?
The separate effects of each of the independent variables
The effects of each IV collapsed across (ignoring the levels of any other variables)
Effect of what you’re testing on its own, effect of task on its own
Shows if the effects of one variable change at different levels of another variable
When do two IVs interact?
When the effects of one variable are different at the different levels of the other variable - levels of the IV affect performance differently at different levels of the task IV
What is interpretation of main events affected by?
The presence of interaction
If we have an interaction between our IVs and we plot the data on a graph, what will this show?
The lines on the graph will not be parallel
What is disordinal interaction?
Disordinal (crossover) interaction is when each variable is having the opposite effect at different levels of the other variable. It is the strongest form of interaction and has very misleading main effects
What is ordinal interaction?
If the task Type only affected one variable and had no affect on the other - can interpret main events with caution
What is meant by no interaction?
Main effects but no interaction, parallel lines on the graph, main effects accurately reflect the results
In a repeated measures design, what are all the factors?
Within subjects
In a between subjects design, what are all the factors?
Between subjects
What are the factors in a mixed design?
Some factors are within subjects and some factors are between subjects
What are the advantages of a factorial design?
Economic- more info less cost
Eliminating confounds - instead of controlling extraneous variables they can be included as extra IVs - more realistic
Factorials allow you to study interactations
What does a factorial ANOVA give separate tests for?
Main effects
Interactions
In which ANOVA test you find within treatment variance?
Between subjects ANOVA
In which ANOVA would you find residual error variance?
Repeated measures ANOVA
What would we need to add for a factorial design on ANOVA?
Between treatments variance is further sub divided into components for the main effects of each factor and for their interaction - the estimation for each component follows same logic as before
How do you work out the interaction variability?
Work out goal between group variability (deviation of each treatment mean from overall mean)
Work out main effect variability (look at variability due to each IV separately) - deviation of the means for each condition of each IV from the overall mean
Total between group variability - main effect variability = interaction variability
What does ANOVA mean?
Analysis of Variance
How can total variability be divided?
F = (main variance + interaction variance) divide by error (individual differences + residual error)
What does k mean in an SPSS table?
Number of levels of variance
How would you find the F value for variable A?
MSbtw/MsError
How would you get MSbtw for variable A?
SS/df