Week 9

Question 1

If Levene’s p>.05

Answer 1

one way anova
if p<.05
post hoc bonferroni or Tukey HSD

Question 2

If Levene’s p<.05

Answer 2

Robust Welch ANOVA
if p<.05
Post hoc Games-Howell

Question 3

Factorial ANOVA

Answer 3

Examining impact of 2 or more factors
factors can be independent, repeated or mixed models

Question 4

interaction effect can be significant

Answer 4

even when the main effect is not

Question 5

main effects

Answer 5

mean differences across levels of one factor, collapsing/averaging/marginalising the other factors

Question 6

when choosing contrasts (planned comparisons(

Answer 6

use CG as reference point
each contrast should compare only 2 pieces
contrasts must not interfere with each other (unique hypotheses, independence)

Question 7

orthogonal

Answer 7

all contrasts are independent of one another
planned and have a reason

Question 8

no. orthogonal contrasts=

Answer 8

k-1

Question 9

Guidelines for contrasts

Answer 9

1. choose sensible comparisons (only 2 chunks, if a group is singled out in one comparison, that group should be excluded from subsequent contrasts so the contrasts are orthogonal/independent)
2. groups coded with positive weights will be compared against those of negative
   sum of weight for a comparison should be
3. Sum of weight for a comparison should be 0
4. if a group is not involved in a comparison, automatically assign weight of 0
5. for a given contrast, the weights assigned to the group in one chunk of variation should be equal to the number of groups in the opposite chunk of variation

Question 10

for k groups, there are ? possible orthogonal contrasts

Answer 10

k-1
k-1 contrasts are required to filly separate the variance explained by the model (SSM)