Common use of Dependent Variable Transformation Clause in Contracts

Dependent Variable Transformation. The field-measured tree heights, are the dependent variables the models are trying to predict. They were examined using the ▇▇▇▇▇▇▇-▇▇▇▇ test for normality to determine if the data needed to be transformed. The original data, and three transformations of the data (log, square root, and squared) were all tested using the ▇▇▇▇▇▇▇-▇▇▇▇ test. Variable y log(y) sqrt(y) (y)^2 ▇▇▇▇▇’▇ Height 0.012 0.001 0.031 0.000 Mean Height 0.075 0.030 0.393 0.000 For the ▇▇▇▇▇▇▇-▇▇▇▇ test, p-values should be 0.05 or higher, for the distribution to be considered normal. If multiple distributions had p-values of 0.05 or higher, the one with the highest p-value was chosen. For these metrics, this test indicates that ▇▇▇▇▇’▇ height could be, but does not need to be transformed, while mean height should be square root transformed. Quantile-Quantile Plots (QQ Plot) were also created to identify unusual residual behavior in the transformed data, shown in Figure 5 and Figure 6.