Discovering Statistics; Chapter 12 Flashcards
1
Q
What is a covariate?
A
Not a main part of experimental manipulation, but can influence the dependent variable.
2
Q
How does ANCOVA allow for a more accurate SSm?
A
If we can explain SSr in terms of a covariate, we reduce the error variance.
3
Q
What is the assumption of independence in terms of ANCOVA?
A
To fully explain the unexplained variance, there should not be any overlap in variance explained by IV and covariate.
4
Q
What is the assumption of homogeneity of regression slopes?
A
Regression lines for entire dataset; lines for IV, covariate and IV*covariate); they must already follow the same pattern (ie. all -ve or +ve relationships)
5
Q
How do you report ANCOVA?
A
F(df, df) = F value, p = sig value
