The authors administered a Japanese version of the Test of Understanding in College Economics, the fourth edition (TUCE-4) to assess the economic literacy of Japanese undergraduate students in 2006 and 2009. These two test results were combined to investigate students' cognitive process or reasoning with specific economic concepts and principles as measured by the TUCE-4 micro and macro exams. At first, the authors conducted preliminary analyses of the test results by means of an one-way analysis of variance and a multiple regression analysis, assuming five factors would affect the test scores: student's gender, year in school, department he/she belongs to, economics learning and the deviation score of school. Then those factors were inputted into the path analysis, according to causal models devised by the authors, to explore student's reasoning mechanism of economic concepts and principles. The results from this analysis showed whether the causal models were fit for their actual cognitive process or not.