Sampling estimators of total mill receipts for use in timber product output studies

Abstract

Data from the 2001 timber product output study for Georgia was explored to determine new methods for stratifying mills and finding suitable sampling estimators. Estimators for roundwood receipts totals comprised several types: simple random sample, ratio, stratified sample, and combined ratio. Two stratification methods were examined: the Dalenius–Hodges (DH) square root of the frequency method and a cluster analysis method. Three candidate sizes for the number of groups were selected from the cluster analysis and subsequently used in the DH stratification as well. Relative efficiency improved when the number of groups increased and when using a ratio estimator, particularly a combined ratio. The two stratification methods performed similarly. Neither the DH method nor the cluster analysis method performed better than the other. Six bound sizes (1%, 5%, 10%, 15%, 20%, and 25%) were considered for deriving samples sizes for the total volume of roundwood receipts. The minimum achievable bound size was found to be 10% of the total receipts volume for the DH method using a two-group stratification. This was true for both the stratified and combined ratio estimators. In addition, for the stratified and combined ratio estimators, only the DH method stratifications were able to reach a 15% bound on the total (six of the 12 stratified estimators). These results demonstrate that the utilized classification methods are compatible with stratified totals estimators and can provide users with the opportunity to develop viable sampling procedures as opposed to complete mill censuses.