The analysis of the combined effect of population and precipitation on rice yield is given by Eq. 20. The interaction model explains about 74.75% of the variability of rice, which is 1.33% less than the bilinear regression (≈76.08%). The model is statistically significant with a p-value of 3.292 x 10-6. All the estimates of the variables are not statistically significant because their p-values are greater than 5%.