Lecture 10 Flashcards
What is a two-way ANOVA (Factorial design)?
A research design that compares the means of THREE or more groups for TWO independent variables
What is a clinical example of a two-way ANOVA?
Measuring stuttering severity (DV) across three groups: no timeout, 2-sec timeout, and 8-sec timeout (IV1), in children with and without otitis media (IV2).
What are “main effects”?
The effect that IV1 and IV2 have (individually) on the DV. The interaction between IV1 an IV2 can also be observed
What are “marginal means”?
The averages of the columns/rows
What are interaction effects?
The interaction assesses whether the effect of one independent variable is different for the different levels of the other independent variable.
How is the variance of the cell mean split up?
- One associated with the variation among row marginal means
- One associated with variation among column marginal means
- One associated with the interaction between the two IVs.
What does the F ratio tell us in a factorial ANOVA?
An F ratio is calculated for each of the variances (i.e. IV1, IV2, and the interaction). If the F ratio is significant, then we can conclude that there is a real effect for that IV/interaction
What are the assumptions of a two-way ANOVA?
• Assumptions of factorial independent groups ANOVA are the same as for the one-way analysis of variance (independence of samples, independence of cases, DV that is interval or ratio scale, normality of underlying population distributions, homogeneity of variance of the underlying population distributions).
What follow-up tests are used in a two-way ANOVA?
Can use planned comparisons, post hoc tests or simple effects contrasts.
Simple effect contrasts operate the same way as main effects contrasts, BUT the calculation is based on differences between the means of one IV at only ONE LEVEL of the other IV
What is partial eta-squared?
An effect size calculation that estimates the variance explained by the specific IV/interaction you are looking at, as a proportion of the total variance of the sample
What is a repeated measures two-way design? (And an example)?
Both IVs are repeated measures; i.e. ALL participants are tested at every level of every IV.
E.g. all participants undergo treatments A and B (counterbalanced) for stuttering. Their %SS is measured immediately after, 3 months later, and 6 months later.
DV = %SS
IV1 = Treatment type (A or B)
IV2 = Time
What is a two-way mixed design? (And an example)?
One IV is a repeated measure (i.e. “time”) and one IV is a between-groups measure.
For example, you could compare pre-test, post-test and follow-up scores for an experimental and control group.
What is the purpose of a MANOVA?
To compare a set of groups (i.e. 3+) on a SET of DVs (3+), where those measures are correlated and conceptually related.
What is an advantage of using a MANOVA rather than several ANOVAs?
It allows groups to be compared on a combination of DVs, which may not be picked up by individual ANOVAs
What happens if the MANOVA produces a significant result?
Univariate tests (i.e. an ANOVA) is conducted to see which DVs are significant on their own
