*** the assumptions are the same as for the 2 variable linear regression model, but it also assumes that the independent variables are not collinear.
- a linear relationship exists between the dependent and independent variables
- no correlation with the error term for the independent variables
- Conditional expectation of the error term is 0
- Homoskedasticity (the variance of the error terms is constant for all observations)
- The error term for one observation is not correlated with that of another observation
- The model only includes the appropriate independent variables and does not omit variables.
- The independent variables are not collinear (the independent variables are not random, and there is no exact linear relation between any two or more independent variables). |