Abstract
I studied the following four problems in the statistical techniques in preparations of a volume table.
As conclusion;
1. The region of rejection about extreme values in the regression analysis is determined by the value which can be obtained by substracting the variance of the estimated value from the variance of the observation.
2. In order to test the goodness of fit of the current volume table, it seems to be better and easier to use the regression analysis. The hypothesis will be a=0, and b=1 in the formula y=a+bx. (y is a true volume value and x is a value in the volume table.)
3. In the determination of regression constant and coefficient, we, firstly, compute both two in each diameter class, and then pool the, data in diameter classes that we found no significant difference among them. Thus we determine regression constant and coefficient. This method of regression analysis can be applied to the test of significant difference of regression constant and coefficient among various regions or tree-species.
4. In logarithmic transformation, the least-squares estimate is always underestimated. H. A. MEYER computed the correction factor only in normal distribution. In distributionfree, the approximate value of the correction factor has been found to be _??_ (n=number of observations, σx2=variance of logarithmic value of observations, In 10=loge 10=2.302585).
