I have a dataset where my dependent variable measures 'How much Trump’s locker room video should have mattered in the election'.
The categories are coded between 1-5, where 1 represents 'should not have mattered at all' and 5 represents 'should have mattered a lot'. This is presumably an ordinal scale.
My independent variables include:
A) 'Disgust towards Hilary Clinton'- which is coded similar to the above variable where 1 represents 'no disgust' and 5 represents 'a lot of disgust'
B) Party ID - which has 7 levels and goes from Strong Democrat to Strong Republican
C) Ideology - which is a 7 point scale where 1 is 'extremely conservative' and 7 is 'extremely liberal'
C) Gender - Male or Female
D) Education - which has 5 levels which goes from 'Less than High School' to 'Post Graduate'
I am worried there might be an issue of multicollinearity between the first 3 variables but I am not sure how to test for this. Should I use a pairwise Chi-squared test for this? And, on what basis should I drop variables which may show high collinearity? Further, is there a way to visualize this similar to scatter plots which are used for continuous variables?
I am trying to implement this in R.
