透過您的圖書館登入
IP:18.222.111.24
  • 學位論文

人類心智模式衡量中不同相似係數之評估

Evaluation of Different Similarity Coefficients in Measuring Human Mental Models

指導教授 : 梁曉帆

摘要


在人類的日常生活中,每個人總是有一套自己的理解方式或推理過程對遇到的事物有所反應,這是因為每個人的心智模式都不一樣。由於現今許多產品或系統的設計講求以使用者為中心,因此在「人性因素」為重要考量的這個時代,要如何有效地衡量人們的心智模式更顯得重要。本研究以八種常見的相似度衡量,來探討經由何種相似係數計算公式所得到的群集結果較能將人類的想法呈現出來。研究方法為先請受測者使用卡片分類法依據自己心中的想法將卡片中的項目做分類,收集到資料後,經由在各種不同相似係數的計算方式之下,觀察所得到的各種相似矩陣與分群結果,將相似係數有差異的部份詢問受測者,請受測者依直覺回答不同項目間的相關程度,再分別與不同係數法所計算出來的相似係數做比較,並用統計方法檢定其差異。而研究結果發現Jaccard、Dice、Anderberg、Ochiai與Ochiai II這五種係數法用來衡量心智模式都能夠代表人們心中的想法。本研究的結果應可提供日後有關於人類認知方面的研究,對於卡片分類的分析能選擇較有效的相似係數計算方式。

並列摘要


Individuals have their own way of understanding the world and constructing their mental models based on their knowledge and experience. Recently many product or system emphasizes user-centered design and "human factors" has become an essential issue, therefore, it is very important to understand and effectively measure human mental models. In this study, eight common similarity coefficients were used to investigate which ones can effectively measure human mental models. The card sorting method was used to calculate the similarity coefficients, subjects were asked the proximities between sorting items through a questionnaire, and then compares the differences between each coefficient and the proximities of subjects answered. Results showed that the Jaccard, Dice, Anderberg, Ochiai and Ochiai II are better formulas of similarity coefficients that can represent human mental models. The findings of this study should be useful for providing an effective suggestion of measuring human mental models.

參考文獻


[14] Finch, H., Comparison of Distance Measures in Cluster Analysis with Dichotomous Data, Journal of Data Science 3(2005), 85-100.
[4] 李素卿,認知心理學,五南圖書出版公司,2003
[5] 呂昌鴻,「探討運轉員對於龍門核電廠系統架構之心智模式」,國立台北科技大學,碩士論文,2009
[8] Anderberg, M.R., Cluster Analysis for Applications. Academic Press, New York. 1973.
[9] Capra, M.G., Factor Analysis of Card Sort Date: An Alternative to Hierarchical Cluster Analysis, Proceedings of the Human Factors and Ergonomics Society 49th Annual Meeting—2005,691-695.

延伸閱讀