high multicollinearity in a fully factorial experimental design GLM model

I am running proportional binomial GLMs with 2 factors (Test Salinity and Region) but I am running into an issue with high multicollinearity. The reason for this is because this is a lab experiment with a fully factorial design where individuals from each region were spit up across 7 salinities - so each region will have the same 7 test salinities.

Therefore, I understand why I am getting high multicollinearity (since there is complete overlap of all the variables tested) but I am wondering if I can simply ignore this when looking at the performance of the model as this was a part of the set-up and can easily be explained -or- do I have to run some other kind of model to predict my proportional binomial response variable?

Hey there, welcome!

but I am wondering if I can simply ignore this

Your estimates won't change, just your confidence interval will inflate.

This topic was automatically closed 21 days after the last reply. New replies are no longer allowed.

If you have a query related to it or one of the replies, start a new topic and refer back with a link.