This paper proposes a method to measure the effect of learning by case method. The method has the following unique features: 1) a set of alternative characters is prepared, 2) the learner’s change of the role understanding is detected by evaluating each alternative character, 3) the analysis method is designed with an orthogonal array technique in the experimental design. We have conducted experiments to examine whether a clear change of the
viewpoints is detected quantitatively by the proposed method. The results have shown the effectiveness of the proposed method.