Provider: ingentaconnect
Database: ingentaconnect
Content: application/x-research-info-systems
TY - ABST
AU - Seegrist, Donald W.
TI - Statistical Models and Estimators of the Proportion of Volume by Log Position
JO - Forest Science
PY - 1974-12-01T00:00:00///
VL - 20
IS - 4
SP - 357
EP - 360
KW - Conditioned regression
KW - multivariate least squares
KW - mathematical models
KW - least squares
N2 - A conditioned regression model for the proportion of volume by log position was proposed previously. In it, estimates of the conditioned regression coefficients were calculated by a step-by-step application of least squares. This paper presents the statistical theory for the earlier procedures. The estimators are shown to be linear combinations of multivariate least squares estimators of the average proportion of volume in each log position. Expressions for the covariance matrix of the estimator are derived. Simultaneous estimators of the conditioned regression coefficients are derived by multiple regression methods. The two methods of estimating the regression coefficients are compared, using volume measurements on peeler logs from 680 loblolly pines. Forest Sci. 20:357-362.
UR - http://www.ingentaconnect.com/content/saf/fs/1974/00000020/00000004/art00013
ER -