ANCOVA when slopes assumption is not met

Since ANCOVA is essentially regression with some additional assumptions, we can still construct an ANCOVA-like solution using regression even when the homogeneity of slopes assumption is violated. This assumption is violated when there is a significant interaction between the covariate and the independent variable.

Topics

Reference

Seltman, H. J. (2018) Experimental design and analysis
https://repository.iit.edu/islandora/object/islandora%3A1012018

2 thoughts on “ANCOVA when slopes assumption is not met”

  1. Dear Charles,

    first of all, thank you for your great work! Which alternatives are there, if the assumptions of an ANCOVA are not met? You mention that it is possible, but i can’t perveive a suggestion of an alternative. I hope i don’t oversee something on this page…

    Thank you very much,
    Severin

    Reply

Leave a Comment