ANCOVA

Question 1

Covariates

Answer 1

Metric-independent variables

Question 2

Use of ANCOVA

Answer 2

to test differences between means where there is an extraneous variable affecting outcome

Question 3

Benefit of ANCOVA

Answer 3

reduces error variance, step-down analysis, eliminates some systematic bias, greater experimental control and baseline adjustment

Question 4

Step-down analysis

Answer 4

Makes data cleaner and clearer

Question 5

Formula for ANCOVA

Answer 5

Yi = b0 +b1Xi +b2Covariate

Question 6

What is ANCOVA similar to

Answer 6

hierarchical regression

Question 7

Omnibus test definition

Answer 7

Tells you there is a difference but not where the difference lies

Question 8

Look at bootstrap or nah?

Answer 8

Bootstrap

Question 9

Assumptions of ANCOVA

Answer 9

Normal distribution, homogeneity, independence, independence of covariate (multicollineraity), homogeneity of regression slopes

Question 10

Checking for independence of coavariate

Answer 10

t-test or ANOVA, must be non-significant

Question 11

Achieving independence of covariate

Answer 11

Randomisation or matching groups of covariate measure

Question 12

Checking homogeneity of slopes

Answer 12

Scatter plot or non-significant interaction

Question 13

Effect size (eta-squared)

Answer 13

SSeffect/SStotal

Question 14

Partial eta-squared

Answer 14

SSeffect/ (SSeffect-SSresidual)

Question 15

How many covariates is best

Answer 15

Small

Question 16

When should covariates be gathered

Answer 16

Before experiment