從碩士論文口試委員探討臺灣圖書資訊學界社會網絡及研究主題多樣性

Abstract

圖書資訊學是一個實務發展先於理論的學科，因而與其他領域有頻繁的知識交流，以往關於跨領域學科互動的研究多基於單向引用關係，缺乏實際互動的交流研究，故本研究從圖書資訊學科的碩士論文審查行為切入，試圖探討國內圖書資訊學與其他學科交流的現況。本研究以論文審查委員共現關係建置社會網絡，以論文委員參與的論文主題與合作對象屬性當作多樣性來源，應用三種指標來計算多樣性，並將網絡特性與多樣性結合，期待能以不同方法探究圖資界教師的合作模式與其研究主題、合作對象屬性多樣性的關係。 研究結果顯示，近六年碩士論文研究主題以「使用者群服務」與「資訊系統與檢索」兩者最多，論文審查委員之學科專長前三名為圖書資訊（22%）、電算機（17.8%）與教育（16.8%），顯現出與其他學科的高度交流。社會網絡則以交大圖資凝聚力最高，其他七校不相上下；臺大圖資明顯傾向邀請校內教師擔任論文委員，而輔大圖資的合作對象專長多樣性最高，分布也最平均。以網絡中心性當作依變項，多樣性與其他屬性當作自變相的SPSS迴歸分析結果，三種多樣性指標具有一致性，惟雪農指標的模型預測力中表現最突出，影響中心性變化的的主要因素有：參與論文數量、論文主題多樣性與年資。 本研究證實論文委員的社會網絡特性與多樣性有關，但是在知識交流方面，僅限於其他學科向圖書資訊學流動，未能完全了解圖資學跨領域互動情形，故未來若要進行相關研究，應考慮納入圖書資訊學教師審查其他學科論文行為，並加入教師的期刊論文、國科會計畫等研究主題多樣性，以期更完整了解圖書資訊學科的知識交流與知識多樣性程度。

Citation analysis has long been adopted to visually present a knowledge domain. Drawing from thesis committee membership data, the study took a social analytical approach to analyze the social and intellectual structure of the field of LIS (Library and Information Science in Taiwan). Specifically, the study attempt to study: firstly, the faculty collaboration network; and secondly, the degree and nature of inter- disciplinary collaboration, as reflected in the composition of the thesis committees. A total 751 Master’s theses published by the eight LIS in the periods from 2006 to 2011 were analyzed. Data extracted included the composition of each committee member- ship and research topics covered in these theses. Furthermore, data about the faculty’s background were also collected, including gender, seniority, department affiliation, highest degree received, with which the faculty’s degree of interdisciplinary collaboration, as reflected in the thesis committee they appeared, could be analyzed. With the theses committee composition data, networks of faculty collaboration were generated where each faulty member were represented by a node, and the strength of their collaboration was represented by the frequency of their co-appeared in these committees. Various network measures, such as centrality, clustering coefficient, E-I index were used to study the cohesion of each department and the LIS filed as a whole. Three diversity measures: Shannon, Simpson, and Gini coefficient, were used to assess the degree of interdisciplinarity of each department and faculty. Regression analyses showed that the number of these advised, seniority, and thesis topic diversity were the significant predictor of a faculty’s centrality in the LIS network.