10 Assumptions of the OLS Method скачать в хорошем качестве

10 Assumptions of the OLS Method

Link for Derivation of Ordinary Least Square estimators:    • Derivation of Ordinary Least Square (OLS) ...   Assumptions of the OLS Method: 1. Linear Regression Model Linear in the parameters Yi = α + βXi + ui 2. X variable(s) and error term are independent Cov(Xi,ui) = 0 3. Zero mean value of disturbance ui The mean value of the disturbance term = 0 given the value of Xi i.e., E(ui|Xi) = 0 If X is non-stochastic, then: E(ui) = 0 4. No specification error or specification bias Specification error means: Inclusion of unnecessary variables Exclusion of necessary variables Wrong functional form 5. Homoscedasticity or constant variance of ui The variance of the error, or disturbance, term is the same regardless of the value of X. Var(ui) = E[ui – E(ui|Xi)]2 = E(ui2|Xi) = E(ui2) = σ2 Var - variance 6. No autocorrelation between the disturbances Cov(ui,uj|Xi,Xj) = 0 Cov(ui,uj) = 0, if X is non-stochastic 7. The number of observations n must be greater than the no. of parameters to be estimated Minimum observations= 2 8. All values of X must not be the same 9. No multicollinearity Multicollinearity occurs when two or more independent variables are highly correlated with one another in a regression model. 10. Error terms follow normal distribution