Answer to Question #232636 in Economics for KAT

Question

Discuss nature of Multi-collinearity. What are the remedial measures to alleviate problem of Multi-collinearity?

Answer 1

Multicollinearity is a serious problem if the aim of the study is to determine the extent to which each of the explanatory variables affects the dependent variable. The presence of multicollinearity leads to an increase in standard errors, distorting the true relationships between variables.

When predicting future values ​​of the dependent variable with a high coefficient of determination (R2> 0.9), the presence of multicollinearity usually does not affect the predictive qualities of the model.

There is no single method for eliminating multicollinearity. This is due to the fact that the causes and consequences of multicollinearity are ambiguous and largely depend on the results of the sample and the economic content of the explanatory variables. Let's list the most commonly used methods.

Excluding a variable from the model. The simplest method for eliminating multicollinearity is to exclude one or more correlated variables from the model.

However, some caution is required when applying this method. In this situation, specification errors are possible.

In applied econometric models, it is desirable not to exclude the explanatory variable until multipolarity and insanity become a serious problem.

Retrieving additional data or a new sample. Perhaps, to decrease multicollinearity, it is sufficient to increase the sample size. This will reduce the variance of the regression coefficients and thereby increase their statistical significance.

This method is applicable if obtaining additional data is not difficult. For example, due to the large number of observations, the initial sample was smoothed with a simple arithmetic mean. In this case, when considering non-smoothed input data, the multicollinearity problem may decrease.

Modification of the model specification. Model transformation. Sometimes the multicollinearity problem can be solved by changing the model specification. Instead of the original model, a different model is used with the same set of explanatory variables. For example, the multiple linear regression model is replaced with a nonlinear model with the same set of explanatory variables.

The ways to transform multicollinear variables are as follows:

• use non-linear forms;

• use aggregates (linear combinations of variables);

• use the first differences instead of the variables themselves.