2

How do I test for homogeneity of regression slopes in ANCOVA? I really need some conceptual and practical help with this assumption. I've scoured the web but a lot of them are not too helpful. Also ideally I would like to be able to test this assumption with R.

Many thanks!

Sarah
  • 59

1 Answers1

1

Technically speaking, the slopes are the same if there is no interaction between the continuous covariate and the categorical factor. The simplest thing you can do is to add that term to the model and see how small the interaction p-value is.

Nik Tuzov
  • 531