Contents
How to test for multicollinearity using Vif and CI?
I am testing my dataset for multicollinearity using VIF and condition indices (CI).My dataset is cross-sectional macroeconomics data. I have 6 independent variables ( x 1, x 2, x 3, x 4, x 5, x 6) plus 2 dummies ( d 1, d 2) plus 2 interactions terms ( d 1 ∗ x 1, d 2 ∗ x 1 ).
Which is an indicator variable with a high Vif?
“The variables with high VIFs are indicator (dummy) variables that represent a categorical variable with three or more categories. If the proportion of cases in the reference category is small, the indicator variables will necessarily have high VIFs, even if the categorical variable is not associated with other variables in the regression model.”
Which is the correct condition number for multicollinearity?
A condition number that “equals infinity” implies that, for any of the M observations, any one of the N variables can be described as a weighted sum of the other ( N − 1) variables. That defines exact multicollinearity. ⇒ q T Z Z T q = λ m i n q T q = λ m i n 1 = 0.
When can you safely ignore multicollinearity?
If the proportion of cases in the reference category is small, the indicator variables will necessarily have high VIFs, even if the categorical variable is not associated with other variables in the regression model.” taken from: When Can You Safely Ignore Multicollinearity? http://www.statisticalhorizons.com/multicollinearity
What should the VIF be for multiple regression?
After removing two problem variables, VIF is ≤ 3.96 for each variable. Do the variables need more treatment or does this VIF seem fine? Condition Index: I have heard that a Condition Index (CI) of 30 or more is a problem. My highest CI is 16.66. Is this a problem? Are there any other dos/donts that need to be considered?
Which is a problem Vif or condition index?
VIF: I have heard that VIF ≥ 10 is a problem. After removing two problem variables, VIF is ≤ 3.96 for each variable. Do the variables need more treatment or does this VIF seem fine? Condition Index: I have heard that a Condition Index (CI) of 30 or more is a problem. My highest CI is 16.66. Is this a problem?