On Apr 30, 2017, at 6:27 PM, Anderson M. Winkler <[log in to unmask]> wrote:

Hi Kamy,

I understand these two variables are only partially correlated, so to take both into account, both should be included in the model.

Hope this clarifies!

All the best,

Anderson

On 30 April 2017 at 19:20, Kamy Wakim <[log in to unmask]ter.edu> wrote:

Hi Anderson,

Thanks so much for your reply!  Do you know if it's necessary to control for both, or whether there is a statistical test that can tell me if controlling for one or the other is sufficient?  An issue in my data is that I'm missing education information from two subjects, and I would prefer not to remove their (perfectly fine) data from my GLM on account of that.

Best,
Kamy Wakim