ANOVA

Question 1

Type I error

Answer 1

false positive, you say there’s a difference when there’s not

Question 2

Type II error

Answer 2

false negative, you say there’s no difference but there really is

Question 3

How are Type I and Type II errors related to ANOVA?

Answer 3

ANOVA tests the difference among all of the groups simultaneously, avoiding inflation of alpha level

Question 4

What is one-way ANOVA used for?

Answer 4

• one IV with 2 or more conditions (e.g., emotion)

Question 5

What is the general logic of ANOVA?

Answer 5

• decomposing variance to examine mean differences
• testing the mean difference through analyzing variance (between treatment variance; measures differences due to systematic treatment effects and random unsystematic factors; within treatment variance measures differences due to random, unsystematic factors)

Question 6

factor

Answer 6

independent variable

Question 7

level

Answer 7

conditions of the independent variable

Question 8

K

Answer 8

of groups

Question 9

n

Answer 9

the # of scores in each treatment or condition

Question 10

N

Answer 10

the total number of scores in the entire study (sum of all n)

Question 11

SS

Answer 11

sum of SDs

Question 12

MS between

Answer 12

the average squared SD between the group means and the grand mean

Question 13

MS within

Answer 13

the average squared SD between each group mean and the individual scores within each group

Question 14

Degrees of freedom for one-way ANOVA

Answer 14

Df between: K-1
Df within: N-k
Total df: n-1

Question 15

Degrees of freedom for factorial ANOVA

Answer 15

(2 or more IV, splitting up variance to between and within)
Df between (overall): # of cells - 1
DF between for A: Ka-1
DF between for B: Kb-1 
DF within: n - # of cells
Total df: n-1
DF interaction: DFa*DFb

Question 16

F value (what is the numerator/denominator)

Answer 16

F is similar to t-test, uses variance as a measure of difference between groups

numerator: differences including any treatment effects
denominator: difference with no treatment effects

Question 17

factorial design

Answer 17

when a study design involves more than one factor/IV that are completely crossed. every level of each IV appears in combination with every level of the other IVs.

Question 18

main effect

Answer 18

independent effects of A and B separately

when examining main effect of A, average across B conditions and vice versa

Question 19

interaction effect

Answer 19

effect of IV1 on the DV varies across several levels of IV2

Question 20

degrees of freedom for each effect

Answer 20

dfbtwnB: Kb-1 (columns-1)
dfbtwnA: Ka-1 (rows-1)

Question 21

difference between descriptive statistics reported in factorial ANOVA and estimated marginal means?

Answer 21

mean response for each factor, adjusted for any other variables in the model