Download (PDF 699 KB): https://research.fs.usda.gov/download/treesearch/4521.pdf
Abstract
As Forest Inventory and Analysis changes from a periodic survey to a multipanel annual survey, a transition will occur where only some of the panels have been resurveyed. Several estimation techniques use data from the periodic survey in addition to the data from the partially completed multipanel data. These estimation techniques were compared using data from two periodic surveys from Georgia. The comparison criteria were based on (1) an estimated mean within the confidence interval derived from using the complete multipanel data set and (2) a small, estimated standard error that does not underestimate the complete data standard error. Multiple imputation matching performed best; the double sampling ratio estimator also performed well. Two methodssingle imputation using group means and single imputation using matched standsboth underestimated the standard error. Replacing the missing observations with growth model predictions using SETWIGS caused an overestimation of the mean.
Citation
Gartner, Dave; Reams, Gregory A. 2001. A Comparison of Several Techniques For Estimating The Average Volume Per Acre For Multipanel Data With Missing Panels. In: Reams, Gregory A.; McRoberts, Ronald E.; Van Deusen, Paul C., eds. 2001. Proceedings of the second annual Forest Inventory and Analysis symposium; 2000 October 17-18; Salt Lake City, UT. Gen. Tech. Rep. SRS-47. Asheville, NC: U.S. Department of Agriculture, Forest Service, Southern Research Station. pp. 76-81
