Week 13: Two-Way Within-Subjects ANOVA

Question 1

• What is a two-way within-subjects ANOVA? What are they also called?

Answer 1

What is a two-way within-subjects ANOVA?

• Also called two-way repeated measures ANOVA
• Two-way = 2 IVs
• Remember, an IV consists of levels
• Within-subjects - a person participates in ALL levels
• When you have 2 within-subjects IVs, try are examining someone who participates in ALL levels of 1 IV and ALL levels of the other IV
• Levels = sometimes referred to as conditions

Question 2

• 2 IVs and DV: categorical or continuous?

Answer 2

2 IVs = both are categorical

DV = continuous

Question 3

• Again, what happens to your error variance in a within-subjects design?

Answer 3

Your error variance or noise (denominator) tends to be lower because participants are taking part in all levels of each IV reducing individual differences within each level.

Question 4

• What are the assumptions of a two-way within-subjects ANOVA?

Answer 4

What are the assumptions of a two-way within-subjects ANOVA?
-Normal distribution - no skewness or kurtosis on DV
-Sphericity - homogeneity of variance of the conditions
(only important when you have 3 or more levels in an IV;
*note: if 2 levels only, sphericity is assumed) **

Question 5

What happens when your IVs have two levels – do we violate sphericity?

Answer 5

Violation of sphericity increases Type I Error
(seeing a differences when in actuality there is none)

You correct for violation of sphericity by using:
- Greenhouse-Geisser - if correction estimates are BELOW .75

-Huynh-Feldt - if correction estimates are ABOVE .75