This study is proposed for two-year period of research implementation. The first year of the project focused on developing computerized cognitive diagnostic assessments. The second-year project designed an experiment to explore the performance of remedy instruction based on DINA model in educational practical settings. The methodology of this study is the non-equivalent quasi-experimentation design. The samples are 107 4th grade students chosen from two elementary schools. The results showed: 1) the test has good properties in reliability and validity; 2) the personalized digital remedial teaching method can enhance the remedial result, especially for medium-achieved and low-achieved students.
