A knowledge-based model that emulates human behavior in a Dynamic Decision Making task is proposed. The model, MAIDEN-DSF, uses a connectionist representation of knowledge and a value function to compute the best alternative. In order to validate MAIDEN-DSF, two data sets have been used: a training set and a test set that contain the behavior of participants that performed the task with different conditions. The results suggest that MAIDEN-DSF is a considerable framework in order to model human behavior. The aim of this paper is to use MAIDEN-DSF to prove that participants do not perceive delay conditions when dealing with Dynamic Decision Making tasks.