Approximate Sampling Variance of Adjusted 3P Estimates
Lewis R. Grosenbaugh · Forest Science · 1976
Abstract Two new approximations for sampling variance of adjusted 3P estimates are presented. One employs population parameters and is closely related to an expression tested by other authors with ambiguous results. The other new approximation is sample-based and appears to be a considerable improvement over several previously published expressions. Both new expressions perform well in a small population where exact computation of theoretically expected variance is feasible, and the parametric expression not only checks closely with well-replicated Monte Carlo estimates for larger populations, but permits prediction of the relatively low sampling intensity at which adjusted 3P sampling becomes more efficient than PPS sampling with fixed number of samples and replacement. Forest Sci. 22:173-176.