Many countries conduct national surveys to collect information on the income of a household and their basic expenditures such as automobile expenses. Studies in the past have shown that income data from such surveys is largely under reported. Additionally, for those households that do not own vehicles, their expenditure data is censored at zero. To study the relationship between automobile expenditure and income, Tobit regression introduced by Tobin (1958) is used to model the non-negative response variable (automobile expenditure) while the errors-in-variables model is adapted to model the independent variable (income) with the use of validation data. We propose a nonparametric test procedure utilizing kernel regression estimators and U statistic to verify the underlying regression function. We further derive the asymptotic normality under the null and alternative hypotheses. Simulation studies indicate robustness and higher power of the test when compared to other leading methods. A real data application of the proposed method provides consistent results and validates the current understanding of the dataset.