Shulou(Shulou.com)06/02 Report--

This article shows you how to use Eviews to do auxiliary regression to test whether the model has multiple collinearity, the content is concise and easy to understand, it can definitely brighten your eyes. I hope you can get something through the detailed introduction of this article.

Multicollinearity in regression is a condition that occurs when some predictive variables in the model are related to other predictive variables. Severe multicollinearity can be problematic because it can increase the variance of regression coefficients and make them unstable. Here are some of the consequences of the instability factor:

Even if there is a significant relationship between the prediction variable and the response, the coefficient may not seem significant.

The coefficients of highly correlated predictive variables vary greatly from sample to sample.

Removing any highly related items from the model will greatly affect the estimated coefficients of other highly correlated items. The coefficients of highly related items may even contain the wrong symbol.

The multiple collinearity of the model can be tested by correlation coefficient method, VIF method, auxiliary regression method and so on. Auxiliary regression method is to establish auxiliary regression between explanatory variables to judge whether there is a linear correlation between explanatory variables. Take the following data as an example:

First of all, the OLS regression of the data is carried out with Eviews, and the results are obtained.

The regression results show that when the significance level is 5%, the t values of the three variables XI, X2 and X3 are 8.6476, 1.5653 and-4.9971 respectively. Compared with the critical value of their degrees of freedom (15-3) 12, except for the variable X2, the other two variables are statistically significant, indicating that there is a linear correlation between consumption level and per capita disposable income and private car ownership. However, there is no significant correlation between the total number of vehicles, which is not in line with the reality, and their overall test is significant, so the guess is that there is multicollinearity between variables.

Next, use Eviews to do auxiliary regression:

Do the regression of variable X1 to variable X2 and X3 in turn, the regression of variable X2 to X1Magol X3, and the regression of variable X3 to X1Magol X2, get the corresponding R ^ 2 from the regression. By F-test whether R ^ 2 is significantly equal to 0, F = R ^ 2 / (Kmurk 1) is divided by (1-R ^ 2) / (nMuk) to determine whether the explanatory variable is collinear with other explanatory variables.

To make an example of auxiliary regression, the remaining two auxiliary regression can be obtained in the computer:

The results of cubic regression show that the values of R ^ 2 are all high (above 0.9), indicating that the variables X1, X2 and X3 are all collinear with other variables, although they are different in degree of collinearity.

Multicollinearity is a kind of sample phenomenon, and no method can completely eliminate multicollinearity, but can only reduce the degree of collinearity. Common remedies usually include deleting a variable from the model, obtaining new data, or reconsidering the form of the model.

The above content is how to use Eviews to do auxiliary regression to test whether the model has multiple collinearity. Have you learned the knowledge or skills? If you want to learn more skills or enrich your knowledge reserve, you are welcome to follow the industry information channel.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.