Ancova

Question 1

what is an ancova

Answer 1

analysis of variance w/ a covariate

Question 2

what are covariates

Answer 2

variates that are associated with Dv but not part of your experimental manipulation

Question 3

what does this mean covariates are? give an example of this

Answer 3

something that is influences the DV that isn’t the iv

age- should effect DV but not part of exp manipulation and should not interact w/ IV and go up in a scale

Question 4

what should you not do when you have an ancova

Answer 4

throw in as many covariantes as possible
use categorical variates
use lots of covariates until you find a significant result= the variance could be caused by something you don’t want

Question 5

why do you use a covariate?

Answer 5

to reduce within group error variance - therefore more likely to find a significant effect of your manipulation (proportion of between group vars to within group error variance)
i.e. reduce error variance between participants by controlling confounds.

Question 6

how do you eliminate confounds

Answer 6

do a thorough literature review to identify important covariables

Question 7

assumptions of ancova

Answer 7

usual assumptions of anova
idependence of covariate and the IV effect- covariates should not be different across IV groups therefore if you did the anova on covar (DV) and group (IV) should be no signif effect (rarely checked)
homogeneity of regression slops - hard to check
cover - rel. between covar and DV is consistent across groups- slopes are homogenous (rarely checked too)