Testing and Correcting for Distributional Misspecifications in the Tobit Model: An Application of the Information Matrix Test.

Recognizing that misspecification in the form of heteroskedasticity and non-normality yields inconsistent parameter estimates of the Tobit model, the Information Matrix Misspecification test was used to detect violations of the distributional assumptions of the Tobit model applied to fresh winter vegetable consumption. The test suggested that the Tobit model was misspecified. However, following a correction for heteroskedasticity and the introduction of the inverse hyperbolic sine transformation to limit outliers, the resulting specification could not be rejected at conventional levels of significance.