This paper suggests how to develop more effective moral educational programs by utilizing interdisciplinary research methods including neuroimaging, social psychological intervention, evolutionary modeling and deep learning methods. Our interdisciplinary research program consists of three steps: identification of core psychological processes involved in moral functioning at the neural level, development of small-scale intervention programs tweaking the identified psychological processes, prediction of long-term, large-scale outcomes of designed interventions using computational methods. We discuss how this research program will inform moral educators and educational policy makers who wish to develop effective moral educational programs in larger educational settings in the long term with actual neuroimaging and intervention experimental data.