本研究針對台灣各大學人文及社會科學領域發表於Web of Science之SSCI、A&HCI引文索引資料庫所收錄之文獻，以書目計量學探討台灣各大學人文及社會科學領域學術生產力之研究，並針對人文及社會科學領域不同學門其學術生產方式加以分析。 依據科技部之學門分類，將人文學與社會科學領域分為21類，分別統計各領域產出相關之情形，以管理學一產出6,876篇最高，而管理學一領域同時也是產出學校數（122所）最多的領域，且以平均產出56.36篇也是最高，顯示管理學一領域相關科系為目前社會科學領域產出之主要科系。人文學與社會科學各領域文獻單一作者與多作者文獻篇數作統計結果發現，各領域間以多作者共同產出文獻為主，僅少部分領域單一作者比例高於多作者比例，以各領域平均作者人數而言，如以四捨五入後為1位者，僅文學一與哲學兩個領域，在全部21個領域中最為接近單一作者產出。其餘19個領域平均作者人數皆大於1.5人。 本研究結果為台灣各大學人文及社會科學領域學術生產力提供一個概括呈現，並協助高等教育評鑑對於全臺灣各大學社會科學領域，學術生產力概況有一背景瞭解，作為日後之評鑑參考。 / This study aims to investigate the scholarly productivity of the humanities and social sciences in Taiwan’s universities by employing a bibliometric methodology. The research data are retrieved from SCI and A&HCI database via Web of Science. In addition, this study reveals the academic production regime in different disciplines under humanities and social sciences. According to the academic classification by the Ministry of Science and Technology in Taiwan, this study is conducted to divide the humanities and social sciences into 21 disciplines, collect the statistical data, and analyze the output in all the disciplines as well. The findings are as follows. Manegement I, with an output of 6,876 articles, leads in quantity of academic papers. Furthermore, Manegement I is also the most highly productive discipline in the fields of humanities and social sciences by the research papers from 122 colleges. As for numbers of average output papers, the discipline Manegement I, again, indicates that it is currently the main department in social science fields; an average output of 56.36 papers is the highest productivity. Compare the multi-author documents with the sole author articles in humanities and social science fields; it is statistically found that, the majority are multi-author papers among the disciplines. Only a small proportion of single-author articles in some areas, the ratio are higher than that of the multi-author documents. In the analysis of the average numbers of authors after rounding to the nearest integer, Chinese Literature and Philosophy, with the average number of author--1, are the only two disciplines which are close to single-author output in all 21 disciplines. The average numbers of authors in the 19 remaining fields are more than 1.5 persons. The results of the study may draw an outline for the scholarly productivity of the humanities and social sciences in Taiwan’s universities. This thesis, on the other hand, may suggest an overview for the further evaluation of the higher education in Taiwan, by building background knowledge for the disciplines under the humanities and social sciences in the universities.

本研究旨在探討1975至2006年間，台灣地區12所大學學術生產力之概況，透過各校被WOS資料庫蒐錄的文獻數量，進一步觀察不同領域中學校、學門，以及個別系所學術生產力的分佈情形，並以書目計量學三大作者學術生產力定律，加以驗證機構學術生產力之可行性，最後利用文獻成長模式檢視12所大學文獻成長情形，進而呈現其研究主題之逐年變遷。 研究結果得知台灣地區學術生產力普遍存在不均現象，多集中於台灣大學；而在自然科學與社會科學領域學術生產力之表現，最多產學門分別為電子與電機工程學門，以及管理學門；研究結果亦顯示合作研究已成為當前學術生產力的主要模式。其次在定律驗證，得知本研究與三大作者學術生產力定律均不相符，究其原因在於高生產力機構數量遠多於低生產力機構，並呈現機構之間學術生產力的極端現象，而與定律不相吻合。最後以文獻成長模式印證台灣地區12所大學整體學術生產力成長情形，可知其融合線性成長、指數成長以及邏輯斯第成長等模式，且目前仍處於蓬勃發展的階段；而近30年來研究主題的變遷，是以電子與電機工程和管理主題為研究主流。 本研究結果在為台灣地區學術生產力提供一個概括呈現，並協助高等教育評鑑對於不同領域及學門的學術生產力概況有一背景瞭解，以作為日後對於不同學科的評鑑參考。 / The main purpose of this study is to explore the scholarly productivity of 12 universities in Taiwan from 1975 to 2006 based on the theoretical perspectives of the bibliometrics, such as productivity of academic discipline, productivity of school, productivity of department, and the growth of literature. The distributions of institution productivity were examined using the Lotka’s, Price and 80/20 Laws. A total of 130018 bibliographic records were retrieved from the SCIE, SSCI, and A&HCI databases. The results of this study revealed that the productivity of institution in Taiwan existed uneven ordinarily, mostly concentrated in National Taiwan University. In scientific field, the most productive academic discipline was Electronics and Electrical Engineering. In social scientific field, the most productive academic discipline was Management. However, the results of this study also showed that coauthorship was major research model. Furthermore, the examination of Lotka’s law was found to be inapplicable. The KS Test is also utilized to test the invalidity of observed distributions. Price’s square root law and 80/20 law were found to be inapplicable to the distribution of institution productivity. The reason was the number of high productivity institutions was more then the number of low productivity institutions. Finally, the growth of literature in 12 universities had been identified. It was demonstrated that the growth of literature was fitted by linear, exponential, and logistic models. Moreover, the overall growth rate is in a vigorously up-growing trend. Nearly 30 years, the mainstream of research was electronics and electrical engineering and management. The results of this study may apply the overview of the scholarly productivity based on 12 universities in Taiwan from 1975 to 2006 and be the background knowledge for evaluation of high education, in order to understand what difference among different academic disciplines.