Abstract---Mathematical models, derived from underlying assumptions of the threshold concept and the
utility maximization principle, continue to grow. Current applied applications of multiple-choice problems.
have generally used simultaneous compensatory models. The models of this class are the multinomial
logit. probit and the generalized extreme value. Within choice modeling, logit is the most commonly used,
generally in a linear functional form. However, these simultaneous compensatory evaluation models do
not appear to reflect the cognitive process by which individuals make their choice.
Work in such areas as psychology, marketing, information theory and artificial intelligence posit that
individuals often use a sequential decision process. This process assumes that individuals process through
the ranking of attributes (ranked according to order of importance), eliminating alternatives in a sequential
manner. The focus of this paper centers around the discussion of sequential processing models. Within
this framework. mathematical models that generate and estimate threshold tolerances and provide choice
probabilities are presented. Furthermore, the use of choice models as a function of individual differences,
and various factors that increase the credibility of the role of sequential processing models are discussed.