過去關於科學計量領域的計量分析結果：Wouters and Leydesdorff (1994)根據Price指標的分類，指出科學計量學並未成為一門硬性的社會科學(hard social science)，Schoepflin and Glänzel (2001)則認為這個領域的異質性很高，每一個次領域都有它本身的特性。
本研究針對以下的問題進行探討：確認Scientometrics期刊上1978到2001年發表論文資料的主題，分析這期間論文的分布情形，不同國家在不同主題上的貢獻，具有主要生產力的機構，並藉由合著關係發掘國內與國際間的合作情形。本研究將論文的主題分為科學計量評估(scientometric assessment)、引用與叢集分析(citation and cluster analysis)、科學計量分布(scientometric distribution)、科學的歷史(history of science)、科學合作(scientific collaboration)、科學計量學的理論研究(theoretical studies on scientometrics)，不在上述主題的論文則歸類為其他。
結果發現：論文數最多的主題是科學計量評估(scientometric assessment)，這個現象反映出科學政策的制定逐漸運用科學計量工具的事實，其次是理論研究(theoretical studies)。在前期(1978-1986年)，科學的歷史(history of science)方面的論文較多，其次是引用與叢集分析(citation and cluster analysis)；科學計量分布(scientometric distribution)在前期與中期(1987-1994年)都相當重要，但後期(1994-2001)逐漸減少；後期具有最重要地位的主題則是科學合作(scientific collaboration)。
美國是目前生產力最高的國家，共占17.7%的論文，主要的8個歐洲國家則共佔47.6%，但美國在論文所佔的比例逐年減少，加拿大與前蘇聯有同樣的情形，但荷蘭、印度、法國和日本在上升中；從每一個機構平均發表的論文數可以看出這個領域的生產力相當分散，1317篇論文的作者資料共來自1538個機構，但是有1109篇論文是單一機構發表，兩個或以上的機構發表的論文只有208篇；在1538個機構中，發表超過15篇或以上論文的機構共有8個，匈牙利和荷蘭各有2個，其餘的4個機構分別位於印度、比利時、英國和美國；雖然目前的論文以單一作者為主，論文的平均作者數僅為1.73，但多位作者的論文雖然僅占18%，但正逐漸增加。
The study indicates that the US share of papers is constantly on the decline while that of the Netherlands, India, France and Japan is on the rise.
The research output is highly scattered as indicated by the average number of papers per institution.
The scientometric output is dominated by the single authored papers, however, multi-authored papers are gaining momentum.
However, Wouters and Leydesdorff [1] presented a combined bibliometric and social network analysis of papers published in first 25 volumes of Scientometrics, and concluded that scientometrics has not become a hard social science as reflected by the values of Price Index.
In another study, Schoepflin and Glänzel [2] point out that the field of scientometrics is heterogeneous, and each sub-discipline has its own characteristics.
The objectives of the study are:
(i) to identify the scientometric themes on which papers have been published in volumes 1(1978) to 50 (2001), and to find out as to how the emphasis on different themes have changed during different periods;
(ii) to examine the distribution of output of different countries during 1978 -2001, and to analyse the change in the trend, if any;
(iii) to study the relative research emphasis of different countries on different scientometric themes;
(iv) to identify the most productive institutions, and to study the scientometric themes they have dealt with;
(v) to study the pattern of co-authorship and the pattern of domestic as well as international collaboration.
The entire data set was classified into seven groups:
scientometric assessment;
citation and cluster analysis;
scientometric distribution;
history of science;
scientific collaboration;
theoretical studies on scientometrics.
Papers which could not fit into these categories were kept under ‘others’.
An analysis of the data indicates that about one-third of the papers published in Scientometrics deal with scientometric assessment which mainly include cross-national, national and institutional assessment, besides evaluation of journals, bibliometric performance indicators, funding and performance, and S&T indicators. This was followed by theoretical studies (Table 1).
From the values of the Activity Index presented in Table 1, it is observed that the priorities of different themes kept changing during different periods. For instance, during 1978-1986 ‘history of science’ followed by ‘citation and cluster analysis’ were the areas of maximum emphasis.
Studies dealing with ‘scientometrics distribution’ got almost the same priority during 1978-1986 and 1987-1994, but emphasis on this theme has gone down considerably in the last block.
During 1994-2001, studies dealing with ‘scientific collaboration’ got maximum priority followed by ‘scientomeric assessment’.
Major contribution (>=2%) of the total output came from 13 countries listed in Table 2. The distribution of papers presented in Table 2 indicates that USA tops the list of publications which are 17.7 per cent of the total world output.
The values of the Activity Index for different countries (Table2) indicate that during the last two blocks, i.e. 1987-1994 and 1994-2001, the productivity of the USA has declined considerably. Similar is the case with Canada and the former USSR.
Further analysis of data presented in Table 2 indicates that Scientometrics is getting Euro-centred, as 8 countries of Europe listed in Table 2 have contributed 47.6 per cent of the total output. The share may be greater, if the output from other European countries not listed in Table 2 is included.
The total output of 1317 papers published in 50 volumes of Scientometrics came from 1538 institutions. 1109 papers were published involving only a single institute and the rest 208 involved collaboration either with 2 or more institutes.
Number of such institutes which published 15 or more papers is only 8 and their share in the total output is 259 (19.66 %). Of the 8 prolific institutions, two are from Hungary, two from the Netherlands, and one each from India, Belgium, UK and USA.
The results presented in Table 5 indicate that slightly more than half of the papers were single authored and the rest were written by either two or more authors. The share of multi-authored papers (>=3) is much less (18%) only as compared to single or two authored papers.
A study carried out by Cunningham and Dillon [7] for authorship pattern in library and information science indicates average number of authors per paper for information science is 1.17. In scientometrics the average number of authors per paper is 1.73 which indicates a better collaboration than library and information science.
The values of DCI for Spain, France, India and Japan were much higher than the world average indicating a good domestic collaboration. However, except Spain all these countries had very low values of ICI, which indicates that these countries have a poor international collaboration. On the other hand UK, Hungary, Belgium, Canada and Germany had good international collaboration as reflected by the values of ICI.
The focus of scientometric studies is shifting from the history of science and scientometrics distribution to scientific collaboration and scientometric assessment.
Scientometric assessment constitutes about 34% of the total output of the papers which is the highest among all the themes. Emphasis on scientometric assessment studies reflects the growing realisation of its utility as a tool for science policy making.
沒有留言:
張貼留言