Skip to main content
On the Finite Sample Behavior of Adaptive Estimators
Journal of Econometrics (1992)
  • Douglas G Steigerwald, University of California, Santa Barbara
With only 50 observations, the adaptive estimator produces confidence intervals that are 20 to 50 percent shorter than those produced by GLS procedures. The key feature is that the underlying error density is symmetric. Under asymmetry the interval length is shortened by a smaller amount.
  • adaptive,
  • efficient,
  • finite sample
Publication Date
Citation Information
Douglas G Steigerwald. "On the Finite Sample Behavior of Adaptive Estimators" Journal of Econometrics Vol. 54 Iss. 3 (1992)
Available at: