Exam 3 Flashcards
Factorial ANOVA:
One way: Factor A (effects?)
1 effect:
A
Factorial ANOVA:
Two Way: Factors A & B (effects?)
3 effects:
A, B, AxB
Factorial ANOVA:
Three Way: Factors A ,B, & C (effects?)
7 effects:
A, B, C, AxB, AxC, BxC, AxBxC
2x2 ANOVA: Main effect of A?
Is there a difference between means of A1, and A2?
2x2 Anova: Main effect of B?
Is there a difference between means of B1, and B2?
2x2 ANOVA: Interaction effect?
Is there a difference in the difference between A1 and A2 depending on the level of B?
2x2 ANOVA: Main effect of A H0 and H1?
H0: uA1=uA2
H1: uA1/=uA2
2x2 ANOVA: Main effect of B:
H0 and H1?
H0: uB1=uB2
H1: uB1/=uB2
2x2 ANOVA: Interaction effect H0 and H1?
H0:uA1B1-uA2B1=uA1B2-uA2B2
H1:uA1B1-uA2B1/=uA1B2-uA2B2
SS A tells me?
How much variation is there between levels of factor A
SS within tells me?
How much does each person differ in their own condition
Unsystematic Variance
Individual differences + error
RMANOVA
Researcher measures more than 2 responses per participant
Benefit of RMANOVA
Helps weed out individual differences which lead to smaller within level variance
Goal of RMANOVA
analyze the repeated measurements as changes or differences
Why do we use an ANCOVA?
To remove a variables influence that we are not interested in
When is an ANCOVA good to use?
When we are unable to match participants on a variable, but we have the data available. often we realize the confound after the fact
ANCOVA Assumptions (#1)
homogeneity of regression (covariate is equal across groups)
ANCOVA Assumptions (2-4)
Normal distribution of variables, variances homogeneous, relationship between DV and covariate is linear
First thing to do in ANCOVA
Test effect of covariate, treatment, and subtract them to get its interaction effect, if not significant proceed, if sig you must treat them as separate groups
MANOVA
2+ DVs are analyzed at the same time, in combination
Example of a MANOVA
Age x Education, measuring DV’s of vocab, speed, math, spatial, memory, attention
Loglinear Analysis
used to measure dichotomously-scored variables
Ex: scented/unscented
Write up for a 2x2/RMANOVA has what components?
all?
1: What was the study, your measurement?
2: Stats? Factors? Levels of those?
Hypothesis, supported?
3: Significant? F’s df’s pwr, Not sig? list f’s df’s pwr
4: Explain results
5: Means, SD’s and R2