Multicollinearity occurs when there exists perfect or exact linear dependence or relationships between two explanatory variables or among explanatory variables in a given model. It has the potential of adversely affecting regression coefficients. It affects the regression beta weights, standard errors and the corresponding statistical significance levels associated with them. You can identify the presence of multicollinearity if you observe any of the following: R-squared will be high, beta coefficients will not be statistically significant, beta coefficients will be contrary to expected a priori or bizarre, large standard errors, small t-statistics, more likely not to reject the null hypothesis, removing or adding a regressor causes substantial changes to the model, and more often than not, the model breaks down. This video explains multicollinearity and how to correct it.

Follow up with soft-notes and updates from CrunchEconometrix:

Website: [ Ссылка ]

Blog: [ Ссылка ]

Forum: [ Ссылка ]

Facebook: [ Ссылка ]

YouTube Custom URL: [ Ссылка ]

Stata Videos Playlist: [ Ссылка ]

EViews Videos Playlist: [ Ссылка ]-

свернуть