Econometric methods
- 2nd ed.
- Auckland McGraw - Hill Kogakusha 1972
- 437 p.
A new introductory chapter has been added on The Nature of Econo metrics. Chapters 2 and 3 treat the two-variable linear model and its extensions as before. Chapter 4 on Matrix Algebra has been considerably expanded with greater attention to partitioned matrices, rank and the solution of homogenous equations, quadratic forms and definite positive matrices. Chapter 5 on the general linear model contains substantial amounts of additional material on such topics as the correlation matrix, prediction. linear restrictions, multicollinearity and specification error. Chapter 6 is essentially new; it contains all the material on dummy variables and pays especial attention to covariance analysis. The treatment of generalized least squares has been simplified in Chapter 7 and new material added on pure and mixed estimation, grouping of observations and grouping of equations. Autocorrelation still receives a chapter to itself and recent important results have been included. Chapter 9 introduces some important asymptotic results and also treats instrumental variables and errors in variables. Two entirely new chapters follow, namely Chapter 10, which is concerned solely with the problems raised by lagged variables and Chapter 11, which covers other important multivariate techniques. The final two chapters. on simultaneous equation problems have also been expanded and include. among other things an assessment of the latest simulation studies of the small sample properties of various estimators. I have not included Bayesian techniques or spectral analysis for, at the present stage of their development in econometrics, they are best treated in specialized monographs and, at the level of prior knowledge assumed for this book, their treatment would have taken up an inordinate amount of space in relation to the rest of the material.