Using the data on % damaged beans of soybean by the soybean podborer (dependent variable), a multiple regression equation was computed using independent variables based on the five-day mean temperature and cumulative precipitation for about eleven and half months from hatching (assumed as Sep. 1) to just before emergence (assumed as Aug. 15), and other independent variables over an 18-year period. A multiple nonlinear regression equation with the highest coefficient of determination, 0.927, was obtained by using the difference in arc sine transformed % damaged beans between the previous and the said year (
Y) as the dependent variable, the sum of the five-day precipitation with
r>0 (
Z1) showing the positive correlation with % damaged beans, two variables of sum of the five-day temperature with
r>0 and with
r<0 (
Z2,
Z3), showing the positive or negative correlation with % damaged beans, and the arc sine transformed damaged beans of the previous year (
Z4) as independent variables. Standard partial regression coefficients of
Z4,
Z1,
Z2, and
Z3 were-0.73, 0.42, -0.26, and 0.20, respectively.
View full abstract