Final session 13b

Question 1

Analysis of covariance (ANCOVA) does what

Answer 1

controls for (or removes) the influence of extraneous variable on the DV

Question 2

is ANCOVA like ANOVA

Answer 2

Like ANOVA, but with a covariate variable included as a predictor

Question 3

How does ANCOVA work?

Answer 3

1 Influence of covariate is removed from the DV using linear regression
2 Remaining effect of the categorical IV is applied to residuals
i.e., portion of the DV left unexplained by the covariate
The tests of difference across the means of residuals is similar to ANOVA Means of residuals are called adjusted means

Question 4

what are the assumptions for ANCOVA

Answer 4

Same assumptions as in ANOVA and Multiple Linear Regression These can be checked in the same way

Additional assumption: Homogeneity of regression slopes across different groups

Question 5

The assumption of homogeneity of regression lines can be checked how

Answer 5

by adding an interaction between the categorical and continuous IVs

If the interaction is significantly different from zero, the assumption is not tenable

Question 6

what are the ANCOVA post-hoc tests

Answer 6

Field recommends pair-wise comparisons with Bonferroni correction Similar to what we used with repeated-measures ANOVA

Tukey HSD is not directly available in SPSS, though variants of it may be available in other software

Question 7

do you do effect size for ANCOVA

Answer 7

yes

A partial version of ω2 is possible

Question 8

explain A partial version of ω2

Answer 8

The partial version of ω2 is not the same as ω2

ω2 is estimated as a ratio of variance due to the effect and total variance: