A Multiple Regression Model for Predicting Mean Volume Yield

DOI JASI HANDLE Web Site Open Access

Bibliographic Information

Other Title
  • 平均材積推定のための重回帰モデルについて
  • ヘイキン ザイセキ スイテイ ノ タメ ノ ジュウカイキ モデル ニ ツイテ

Search this article

Abstract

The objectives of this paper were as follows: (1) to construct a multiple regression model which predicted volume yield per unit area for KARAMATSU (Larix leptolepis Gor.) plantations growing in the Kyushu University Experimental Forest in Hokkaido; (2) to investigate the effect of sampling for independent variable values on KARAMATSU plantation yield estimates. Three total sample sizes, three stratification schemes and two methods of calculating mean estimates were exmained by Monte-Carlo simulation procedure. Total sample size had little effect on the amount of improvement in the estimate of mean yield brought on by drawing stratified, rather than simple, random samples. The actual improvement in the precision of the yield estimate found when sample size was increased was similar to those indicated by usual sampling theory. Site index was found to be a useful stratification variable for estimating the yield of KARAMATSU plantations. Furthermore, some multiplicative effects between the two stratifying variables (site index and numbers of trees per unit area) were noted. The two types of mean estimates were evaluated. Method 1 which is theoretically unbiased, estimated the plantation mean yield with the average yield of the sample plots. Method 2 used the single yield computed from the sample average of the independent variables as the estimate of plantation mean yield. Method 1 was superior in terms of average bias, mean square error and average absolute deviation. Method 2 had slightly better precision than Method 1. In general, Method 1 is recommended over Method 2. Stratifying the plantation by index and numbers of trees per unit area before sampling may produce sizable gains in the precision of mean yield estimates over simple random samples if the plantation is heterogeneous.

Journal

  • 九州大学農学部演習林報告

    九州大学農学部演習林報告 56 1-12, 1986-03-25

    Research Institution of University Forests, Faculty of Agriculture, Kyushu University

Details 詳細情報について

Report a problem

Back to top