本研究提案以Rasch Model 的參數模式，運用灰關聯分析方法 GRA (Grey Relational Analysis)，與S-P 表的方式建立Rasch Model GSP (Grey Student Problem) 表，並且以結合灰色結構模型法GSM (Grey Structural Model) 建立GSM 結構圖。經由研究結果，實質獲得了學生與試題的鑑別度「α」，以及學生測驗平均值與試題難度「β」，而從「γ」的分佈圖，可以辨識學生的得分率和試題答對率情形。透過GSM 結構圖進行研究解析，可清楚了解結構成分的分群狀態，以及界定彼此的關聯性，這說明GSM 結構圖是一種學習地圖的概念，是具有方向性和定位的結構圖。本研究提出GSP 表與GSM 結構圖的研究組合，明白顯示全部受測者的答題結構，讓研究結果清楚定位與呈現，GSP 表對於人數及試題數沒有限制，而且可以比較程度近似的班級，GSM 結構圖針對試題的層級結構組織之關係，能夠有系統的呈現出來，提供給教師進行命題的參考，同時也是一種可以取代S-P 表的研究方法。

The paper aims to apply Rasch Model, Grey Relational Analysis (GRA) and S-P chart to establish the Rasch Model GSP (Grey Student-Problem Chart). With the combination of Grey structural modeling, the GSM (Grey Structural Model Figure) could be established. Based on the findings, the GSP is able to be set up with discrimination from students and items, and be used as discriminationα. The average of item difficulty and the average of examinees’ scores could be used as difficultyβ. The lowest score and the lowest rate of correct test items could be recognized as guessingγ. Through the GSM figure, it is clear to understand the structural components of the clustering state, and define the relevance of each other, which means that GSM figure is a learning map concept is a direction and positioning in the process. This study proposes the research portfolio of GSP chart and GSM figure. GSP is not limit on the number of people and questions, and comparison of similar classes. Moreover, the GSM figure clearly distinguishes the clustering and structure of data between the orientations, and providing teachers a reference to the topic, it’s a research method can be replaced by S-P chart.