In the rehabilitation field, there is a need for quantitative evaluation independent of the therapist's subjective evaluation. In our previous work, we developed a binary-valued quantitative evaluation algorithm for the motor function of the shoulder joint based on sensor data from a test performed by therapists. In this paper, we developed a quantitative evaluation support system by implementing the binary-valued quantitative evaluation algorithm of the motor function and verified the effectiveness of the proposed system. The quantitative evaluation support system uses a three-dimensional force display robot that can imitate the test motion and the evaluation by the therapist. For this purpose, we analyzed the test motion and the evaluation by the therapist for ten subjects to determine two threshold parameters used in the binary-valued quantitative evaluation algorithm. In order to verify the evaluation accuracy in the test using the developed quantitative evaluation support system, we conducted the test using the developed system for three subjects. The sensitivity was 75% and the specificity was 75% in the test using the developed system. We plan to improve on the evaluation accuracy in our future work.