Anova Lectures Flashcards
What analysis technique is used when there is one IV with 2 levels?
T test
F ratio formula
Variance of the model/ variance of the error
=MSm/ MSr
Levenes test
Independent groups
Must be non significant to assume homogeneity of variance
Mauchlys test
Repeated measures
Must be non significant to assume sphericity
Bonferroni
Pair wise comparison
Most conservative measure of risk reduction
Tukey
Pair wise comparison
Used when there are lots of comparisons
Why do we do pairwise comparisons rather than lots of Ttests as post hoc exploration
To reduce type 1 errors
When to use planned contrast
When you have prediction about levels
Which proportion of the total variance is ignored in repeated measure ?
Between subject variance
When you have a 3 main effects, a 3 way interaction and several 2 way interactions.. Which should you explore and report first?
3way interaction
2 way interactions
Main effects
What is a covariate
A factor which also affects the DV
How does it improve analysis
Makes it more pure
Protects internal validity
More realistic than controlling for everything in the design
The covariate should affect what…?
It should affect the DV but have no influence on the IV
When to use an ANOVA
Continuous DV
Categorical IV with more than 2 levels
OR
More than 1 Categorical IV and 2 or more levels
