Classical Linear Regression Model

There are some assumptions relating to CLRM. These assumptions are necessary to show that the estimation technique, Ordinary Least Square Method has a number of desirable properties and also that the hypothesis tests regarding the coefficient estimates could be validly conducted.

The Assumptions are:

1. Linear Regression Model, that is, the model is linear in parameters and not necessarily in variables.
2. X (explanatory variable) is fixed or known to us. Hence, X is non-stochastic.
3. Mean of disturbance term, u is 0.
4. u is homoscedastic, that is, u has equal variance for all the terms.
5. There is not auto correlation between two disturbance terms, u. That is, covariance between them is 0.
6. 0 correlation between u and X. Therefore, X is exogenous.
7. Number of sample observations (n) is greater than number of parameters to be estimated.
8. Variance of X is finite, that is, greater than 0.
9. There is no specification bias or error in model, which means, all the relevant variables must be included in the linear regression model.
10. There exists no perfect linear relationship between explanatory variables.
11. Random error term u, follows Normal Distribution with mean 0.

Click here for government certifications