This paper examines the possibilities and limitations of introducing simulated experiments in the psychology domain by practicing a course with graduate students in psychology, in which simulated experiments were conducted in place of real human experiments. The class-learning object was the dual-storage model of human memory. The simulation results showed anomalous results that differed from intuitive predictions. The results were also inconsistent with the results of prior human experiments. Analysis of reports submitted by 15 participants revealed that they focused on the anomalous results emerging in the simulation results and examined them based on the dual-storage model theory. On the basis of these results, future directions for this practice are discussed.