Anova Flashcards
(13 cards)
ANOVA RM sphericity
One more thing for repeated measures in ANOVA , testing the same person multiple times is like the t test but during t test you are only testing somebody twice
Repeated measures Anova sphericity or circularity assumptions
The two terms are synonymous
- the assumption of sphericity states that the variance of the differences between treatment A and B equals the variance of the differences between A and C, which equals the variance of A and D
Which equals the variance of differences B AND D
- we test for the sphericity assumption using MAUCHLYs test of sphericity
ANOVA example : alcohol attitudes data set towards water, wine, and beer
We are testing the same people multiple times , we can use a repeated measures ANOVA to test whether participants have different attitudes towards different kinds of drink
Anova stands for ?
Analysis of variance
Why not use multiple t test for the previous example
This can create a type 1 error a false positive
The probability of type one error increases as the number of statistical comparisons increases
When to use F distribution
Working with more than two samples
ANOVA used when ?
Used with two or more nominal independent variables and an interval/ratio dependent variable
What is ANOVA ?
One or more independent variables
With different levels/ conditions
Analyzing differences in means
On a continuous dependent variable
Why ANOVA ?
Therefore, you use an ANOVA to cover the chance of type 1 error
However ANOVA have the disadvantage of involving extra steps or post hoc tests
- usually t- tests correcting for multiple comparisons
Parts of ANOVA - IV
Independent variable - this variable has to be dichotomous variable
You can put people into groups
Or your experimental manipulation , ex: instructions vs no instructions
Parts of ANOVA - DV
The dependent variable needs to be continuous variable
Your DV should be the measurement you took in your study, or what information you are expecting to see changed over groups
Types of ANOVA
Between group independent - for independence measures where participants are in only one level condition
repeated measures/dependent/within group : another for dependent measures where participants are in all levels/conditions
Mixed designs ANOVA - for when you have two variables, one between subjects and one repeated measures
Types of ANOVA for two or more way # x #
Designates the number of levels for the multiple IVs in a two way design
