MANOVA
- Created by: Chloe
- Created on: 02-05-15 14:30
View mindmap
- MANOVA
- A generalisation of ANOVA in the case where a number of DV's are measured at the same time
- Maintains the Type 1 error rate as only carrying out one test
- Following a significant MANOVA test, separate univariate tests may be then conducted to determine which of the DV's are significantly influenced by the IV
- When DV's are correlated, this should make use of ANCOVA: to control for all other DV's while each one is being evaluated
- When DV's are not highly correlated ANOVA'S may be used
- When DV's are correlated, this should make use of ANCOVA: to control for all other DV's while each one is being evaluated
- When carrying out several ANOVAS you need to use the Bonferroni procedure to modify the significance level
- E.g if alpha level is set at 0.5 & you have three DV's you need to set the critical P value of 0.05/3 = 0.017 for each ANOVA
Comments
No comments have yet been made