Week 3 revision

Question 1

Explain what an omnibus test is, and list the effects in a 2-way ANOVA that can be
categorized as omnibus tests

Answer 1

Any test resulting from primary partitioning of variance - They can be Main effect F1, Main effect F2, Interaction effect.

Question 2

Identify the circumstances in which follow-up tests are required in 2-way factorial ANOVA

Answer 2

If a variable has more then 2 levels.

Question 3

Explain what a significant main effect comparison tells you

Answer 3

Which marginal means are different

Question 4

When testing simple effects in a 2-way ANOVA what are the research q’s/hypothesis being tested?

Answer 4

What is the effect of F1/F2 at each level of of F2/F1

Question 5

When testing simple effects in a 2-way ANOVA what are the type of statistical test (e.g. F, t, z) that is used

Answer 5

F test

Question 6

When testing simple effects in a 2-way ANOVA what is the error term used?

Answer 6

The same MSerror used in the original ANOVA.

Question 7

When testing simple effects in a 2-way ANOVA what degrees of freedom are calculated for SSeffect / treatment and SSerror

Answer 7

DF for simple effects are just the df for the associated main effect (observations -1)

Question 8

When testing simple effects in a 2-way ANOVA how would you determine the maximum number of simple effects that could be tested for a factor?

Answer 8

(a, N-a) where a is the amount of levels in F1 and N is the number of all levels int he experiment.

Question 9

When testing simple effects in a 2-way ANOVA
what does a significant simple effect tells you?

Answer 9

A sig simple effect tells you which groups among a veriable have significant effect on the DV

Question 10

In a 2-way ANOVA, explain how the variance is re-partitioned when testing the simple effects of Factor A

Answer 10

The sum of the simple effects of Factor A
for each level of Factor B are equal to the sum of the main effect of Factor A and the interaction.

Question 11

Identify the circumstances in which simple effects would need to be followed up

Answer 11

When they are significant with more then two levels

Question 12

When testing simple comparisons what are the research question (and hypothesis) being tested?

Answer 12

Which level of A at B is significantly different.

Question 13

When testing simple comparisons what is the type of statistical test (e.g. F, t, z) that is used

Answer 13

F test

Question 14

When testing simple comparisons what does a significant simple comparison tells you?

Answer 14

The observed difference between the compared groups is unlikely to be due to chance.

Question 15

Identify the problems (and solutions) associated with simple comparisons (2)

Answer 15

They increase familywise/Type 1 error from happening
Could redundantly be explaining the same mean difference more then once

Question 16

Explain why significance tests are not that helpful when we want to determine the importance of findings (3)

Answer 16

Arbitrary acceptance leads to binary results
No practical information
Large p-value will eventually be insiginficant under certain population parameters (EFFECT SIZE)

Question 17

Define effect sizes and explain why they are useful

Answer 17

Assess the reliability of the results in terms of variance, and can compare effects within a factorial design.

Question 18

What does eta-squared (η2) describe

Answer 18

describes the proportion of variance in the sample’s DV scores that is accounted for by
the effect

Question 19

What does omega-squared (ϖ2) describe?

Answer 19

describes the proportion of variance in the population’s DV scores that is accounted for by the effect

Question 20

What factors influence the difference between eta-squared and omega-squared? (2)

Answer 20

sample size and error variance

Question 20

Is eta-squared or omega-squared more biased, and why?

Answer 20

eta-squared is considered a biased estimate of
the true magnitude of the effect in the population due to sample size and error variance making it proportionally bigger.

Question 21

What is the difference between eta-squared (η2) and partial eta-squared (η2p) in terms of
what each estimate describes?

Answer 21

Eta-squared is a proportion of total variance accounted for by the effect. Partial-eta-squared is a proportion of residual varriance account for when other variable are partitioned out

Question 22

What are the two key limitations of partial eta-squared (η2p)

Answer 22

It is more inflated because it adds error+effect, not just total
Can add up to more then 100%