Summary
Contents
Berry and Feldman provide a systematic treatment of many of the major problems encountered in using regression analysis. The authors discuss: the consequences of violating the assumptions of the regression model; procedures for detecting when such violations occur; and strategies for dealing with these problems when they arise. The monograph was written without the use of matrix algebra, and numerous examples are provided from political science, sociology, and economics.
Heteroscedasticity and Autocorrelation
Heteroscedasticity and Autocorrelation
We have seen that the multiple regression model requires the assumptions that (i) the mean of the error term equals zero—E(∊j) = 0 for all j, (ii) the variance of the error term is constant so that VAR(∊j) = σ2 for all j (homoscedasticity), and (iii) no autocorrelation is present—the error terms ...