The article discusses how values of time for travel demand models are derived using behavioral theories. Different statistical techniques like logit, probit, and discriminant analysis are evaluated to determine the best approach. It is important to ensure the accuracy and adequacy of the models to infer values of time correctly. The study highlights the need to establish statistical validity and confidence in the model coefficients and avoid extrapolating results beyond the data range. Various methodological and conceptual challenges are identified in determining how mathematical techniques relate to choice behavior hypotheses. The article also addresses issues related to user characteristics, mode or route differences, and model specifications for different trip segments.