Heteroscedascity
The variance of the error term is not the same for all observations. For example, the variance of the error term might increase with the mean of the dependent variable as is often the case with health expenditures. Or the variance might be larger for observations in one group, for example, one health plan, than in another group. The presence of heteroscedascity does not bias the estimators, but it does bias estimates of their standard errors. Common tests to detect the presence of heteroscedascity include the White test, the Breusch-Pagan test, and the Goldfeld-Quandt test.
Kennedy, P. 1998. A Guide to Econometrics, Cambridge, MA: The MIT Press.
Greene, L.. 2003. Econometric Analysis, 5th ed., Upper Saddle River, NJ: Prentice Hall.
Wooldridge, J.M. 2003. Introductory Econometrics, Mason, OH: Thomas South-Western.
|