Optimized Endogenous Post-Stratification in Forest Inventories

• Activity Number: 478
• Type: Contributed
• Date/Time: Wednesday, August 3, 2011 : 8:30 AM to 10:20 AM
• Sponsor: Section on Statistics and the Environment
• Author(s): Paul L. Patterson*+
• Companies: Forest Service
• Keywords: endogenous post-stratification ; environmental monitoring ; forest inventory ; remote sensing ; post-stratification ; logistic regression

Abstract

An example of endogenous post -stratification is the use of remote sensing data and a sample of ground data to build a logistic regression model to predict the probability that a plot is forested and then use the predicted probabilities to form categories used for post-stratification. Breidt and Opsomer showed design consistency of the endogenous post-stratified estimator and, under a super-population model, the consistency and asymptotic normality of the endogenous post-stratified estimator by showing it has the same asymptotic variance as the traditional post-stratified estimator with prespecified strata. An optimized endogenous post-stratified estimator where the optimization occurs over all cut points has been recently proposed in the literature. There are no known literature results describing the operating characteristics of this new estimator. This study reports the results of a detailed Monte Carlo investigation of the small and large sample performance of the optimized endogenous post-stratified estimator under a variety of realistic scenarios and compares its performance with earlier approaches. The results provide guidance as to the appropriateness of these estimators.