Files
Abstract
Recent interest in modeling randomness in crop yields has typically focused on two distinct areas: stochastic trend and non-normality. However, the estimated errors from the stochastic trend model fail a test for normality. Thus, a transformation is used to model the non-normality of errors within the stochastic trend model.