Revisit problems encountered in linear regression models

This paper revisits the main problems faced by linear regression models. The exogenous condition for linear regression is verified by using the triangular law of vectors and dot product of vectors. Errors in variables and missing variables are analyzed analytically. This paper shows that these two problems will cause large sample bias of LS estimator for the parameters. Practical suggestions are made on how these problems can be minimized on a case to case basis