Citation Studies of Publications in Superconductivity Research by China with Comparative Studies of Some Other Countries

Transcription

1 Citation Studies of Publications in Superconductivity Research by China with Comparative Studies of Some Other Countries A study submitted in partial fulfillment of the requirements for the degree of Master of Science in Information Management at THE UNIVERSITY OF SHEFFIELD by September / 95

2 Abstract Aims: This study aims to analyze citations statistics of original publications in superconductivity research by China, and measure the impact of China in superconductivity research with comparison to some other leading countries of the world. Background: Superconductivity research has been explosive since two scientists of IBM Zurich research center discovered which is so called as high-temperature superconductor in During , Chinese scientists contributed thousands of original articles in superconductivity research, and there is a general interest to quantify the contributions and impact of China in superconductivity research. Methodology: The methodology used for this study is citations analyses. Data source is from ISI Web of Sciences. Web of Sciences analyzing tools and Microsoft Excel are used for data analyses. Results: The original articles by China in superconductivity research are published in an extensive range of source journals and subject areas. Citations of these articles are from a broad range of countries, much more extensive range of source journals and subject areas. The total number of citations of China is far less than USA and Japan, less than Germany; however, the number of citations of China of each year has been continuously increasing, and it has been catching up or over taken England, France and Russia in many cases in recent years. Conclusion: China is one of the leading countries of the world in superconductivity research. The general impact of China in superconductivity research is far less than the super countries of USA and Japan, and less than Germany; however, China has been catching up or over taken England, France and Russia in many cases with increasing impact in superconductivity research in recent years. 2 / 95

3 Table of Contents Chapter 1 Introduction Aims and Objectives Background Research Background and Original publications Citation Analysis and Its Related Applications Implications of Interest of This Study... 7 Chapter 2 Literature Review Superconductivity Research Science Citation Index and Citations Analyses Citations Analyses in Mapping Superconductivity Research Citations Analyses in the Measure of Nations Contribution Web of Sciences, Scopus, and Google Scholar Chapter 3 Methodology Data Source: Web of Sciences Data Selection Criteria: original articles Data Analysis Criteria: non-self-citations Data Classification Scheme: JCR scheme Research Techniques and Approaches Chapter 4 Limitations of the Research Limitations by Data Sources Limitations of Methodology of Citations Analyses Limitations of Technical Analyzing Tools Limitations of Time of Study Chapter 5 Original Articles by China Growth of Articles through the Years English is the Primary Language Subject Areas Source Journals International Collaborations Chapter 6 Citation Analyses of China Citations by Document Types Citations Growth through the Years Citations by Different Languages Citations from Different Countries/Territories / 95

4 6.5 Citations by Institutions Citations by Subject Areas Citations by Source Journals Chapter 7 Comparative Studies Original Publications Publications Growth through the Years Publication Source Journals Publication Subject Areas Citations Analyses of Different Countries Citations Count, Averages, Growth, and h-index Citations Distribution in Different Countries Citations Source Journals Citations Subject Areas Key and Top Institutions Definitions and Numbers Citations of Top Institutions International Collaborations Original Publications Citations boosted? Chapter 8 Conclusions References: / 95

5 Chapter 1 Introduction 1.1 Aims and Objectives This study aims to analyze citations statistics of original articles by China in superconductivity research, and measure the impact of China in superconductivity research with comparison to some other leading countries of the world. Other objectives include finding out key institutions of China in superconductivity research and comparing the impact of top institutions of different countries, investigating international collaborations in superconductivity research and whether or not citations are boosted by international collaborations. 1.2 Background Research Background and Original publications In 1986, Bednorz and Muller, 2 scientists at the IBM Zurich Research Laboratory, reported discovery of new superconducting materials which demonstrated superconductivity at much higher degree than previously known. This finding was so important that it won the 1987 Nobel Prize in Physics, and led to an explosive further research development in superconductivity over the world (Royal Swedish Academy of Sciences, 1987). In 1987, Chinese scientists made a tiny breakthrough in finding better superconductors (Zhao et al., 1987). Since then, a lot more new contributions have been made world widely in superconductivity research. During , 125 countries contributed more than 100,000 publications in superconductivity, among which China ranks No.6 with a total number of 7,241 publications in this research area (Thomson Scientific, 2008c). 5 / 95

6 1.2.2 Citation Analysis and Its Related Applications Citation analysis has become a methodology adopted by many bibliometric researchers in different science domains since the advent of Science Citation Index (SCI). SCI is now in its web edition called Web of Sciences (WOS), which covers more than 5,800 journals from more than 100 disciplines of the world s leading research, and provides access to current and retrospective bibliographic information, author abstracts, and cited references (Thomson Scientific, 2008b). However, it has been upon the same assumptions for SCI and WOS that Garfield and the Institute of Scientific Information (ISI) or Thomason Scientific have founded the database services. The assumptions are that (although arguably): there are inherent relationships between an original cited publication and the succeeding citing publications, and in some respect, citations indicate the impact of the original cited research on the succeeding citing researches. By counting and analyzing the publications which have cited the original publications, citation analyses have been proved to be very useful at different levels in the assessment of the extent and range of impact of the original researches. With regard to the research of superconductivity, there have been already documented citation analyses which revealed the country impact in this research area. However, there has not been any of this kind analyses carried out and reported for China by now. 6 / 95

7 1.2.3 Implications of Interest of This Study In China, there have been two National Prizes in Natural Science Research (which are the highest national government prizes) awarded to the progresses in superconductivity research since 1986 (The Ministry of Science and Technology of China, 2007). On the other hand, more than a few of Chinese people including scientists have the impression that superconductivity research in China has already advanced to such level as very close to that of the leading countries (e.g. U.S., Japan, etc.) in the world in some respect (Zhang, 2006). Therefore, it might be of general interest to carry out a quantitative research by analyzing citations statistics of original publications by China with comparison to those of some other leading countries of the world, in order to disclose the actual impact and contributions of China in superconductivity research. 7 / 95

8 Chapter 2 Literature Review 2.1 Superconductivity Research Superconductivity was first discovered by H. Kamerlingh Onnes in What he found was that, when reached a critical temperature (typically a few degrees above absolute zero) specific to different metals, the electrical resistance of some metals disappeared completely in a narrow temperature range. (Schrieffer and Tinkham, 1999) This discovery is of tremendous significance in many respects. For example, pilot systems have already been developed for applications such as electrical generators and energy storage arrangements. However, since the available superconducting materials require unpractical low temperatures, the technical applications have been greatly limited. (Royal Swedish Academy of Sciences, 1987) Therefore, the drive for exploring high temperature superconducting materials is persistently demanding. By now, researches have reported materials of superconductivity at increasingly high temperatures. However, the milestone was in 1986 when Bednorz and Müller found superconductivity in an oxide material at a temperature 12 C higher than previously known (Royal Swedish Academy of Sciences, 1987). This discovery started a new era of what is now so called high temperature superconductivity research. 8 / 95

9 2.2 Science Citation Index and Citations Analyses Science Citation Index (SCI) was launched in 1964 by Institute for Scientific Information (ISI), which was founded by Eugene Garfield. Now, SCI has evolved to its web edition Web of Sciences (WOS), and ISI has been merged into Thomson Scientific. SCI/WOS aims to cover the core literatures, most of which are journals, of all disciplines of science, and expand the coverage to include other important literatures based on the ISI defined impact factors. Moed conducted a thorough investigation of SCI/WOS coverage by discipline, and concluded that all disciplines of natural science are covered as excellent or good (Moed, 2005). SCI/WOS was primarily designed for search of literatures, and citation index was meant to augment traditional natural language indexing terms. However, since it is built upon linkage of citations, SCI/WOS now has often been used for citation analyses in order to assess the contributions made by various bodies in the scholarly system to the advancement of scholarly knowledge (Moed, 2005). According to Garfield (Garfield, 1979), citations support, provide evidence for, illustrate, or elaborate on what the author has to say, and hence, the act of citing is an expression of the importance of the material. The total number of such expressions is about the most objective measure there is to the material s importance to current research. However, it has always been argued what the citations actually measure. Some researchers advocate that citations measure quality of research (Cole and Cole, 1971), authoritativeness (Gilbert, 1977), intellectual influence (Zuckerman, 1987), popularity (Borgman and Furner, 2002). Others contend that what citations measure is not clear (Cronin, 1984; MacRoberts and MacRoberts, 1989). 9 / 95

10 MacRoberts presented a complete picture about the problems in their critical review: Formal Influences Not Cited ; Biased Citing ; Informal Influences Not Cited ; Self-Citing ; Different Types of Citations (affirmative or negative); Variation in Citation Rate with Type of Publication, Nationality, Time Period, and Size and Type of Specialty ; Technical Limitations including Multiple authorship, Synonyms, Homonyms, Clerical errors, Coverage of literature,, etc. (MacRoberts and MacRoberts, 1989) However, despite the debate, citations analyses have been applied to the evaluation of research performance widely. These performance evaluations based on citations analyses have been practiced extensively at individual (most at internal circumstances, e.g. for making decisions about hiring and tenure), departmental (Bradley et al., 1992; Bishop et al., 2003) or institutional (Seng and Willett, 1995; Redman et al., 2001), and national (Braun et al., 1994a; Braun et al., 1994b; Hicks et al., 2004) levels. The above listed references are only a few examples. Meanwhile, researches have also been carried out in the attempt to validate the results of citations analyses by comparisons to peer reviews. For instances: Lawani and Bayer presented evidences based on 872 cancer papers that highly rated papers by peers are more highly cited in the publications. (Lawani and Bayer, 1983) Goldberger et al conducted a study on the Research-Doctorate Programs in the United States and identified "the clearest relationship between ratings and citations: faculty in top-rated programs cited more than faculty in lower-rated programs. (Goldberger et al., 1995) Norris and Oppenheim demonstrated again the high correlation between 2001 UK Research Assessment Exercise (RAE) scores and the citations count for archaeology departments of UK universities (Norris and Oppenheim, 2003). On the other hand, researchers who advocate citation analyses for research evaluation also observe the problems, and declare that citations 10 / 95

11 are only rough indicators which could be used to measure the works among a relatively large sample of scientists rather than a particular individual s work (Cole, 1989). Fortunately, while the problems of citations analyses in the evaluative context are magnified in the circumstances of individual evaluations, these problems could be confidently canceled out while the evaluation is conducted at an aggregate level of the joint performance of scientists. There are other applications of citations analyses as well, which comprise mapping science activities, identifying trends or research fronts, and tracing the history of research (Garfield, 1998). Moed summarized citations analyses into five broad themes: (1) The assessment of the contribution made by various bodies in the scholarly system to the advancement of scholarly knowledge; (2) Analyses of the global scholarly system, including studies of its various characteristics and their relationships to research performance; (3) Analyses of scholarly fields; (4) Analyses of the science-technology interface and the economic contributions of science; (5) Assessment of educational, social and cultural contributions of basic research. (Moed, 2005) This study will involve analyses of nations contributions of science in the field of superconductivity research, and may also reflect related scientific activities such as the international collaborations in the global context. 11 / 95

12 2.3 Citations Analyses in Mapping Superconductivity Research There are different approaches in mapping the research activities of superconductivity by citations analyses. The following are only some examples. Nadel compared the citation and co-citation data of the core explanatory theory of superconductivity (BCS theory) and the historically identified events, and indicated the correspondence in between (Nadel, 1981). Arunachalam and Singh analyzed publications in superconductivity research of , and their citations over the following twelve years, with the attempt to identify the geographical origin, language, and journal-wise distributions of the papers, the citedness and the distribution of citations as a time series. Their main findings included: English is the main language for communications of superconductivity research; USA, USSR and Japan are leading countries in the research; the highly cited papers received most of their citations during the years shortly after publication, and the citations declined with years (Arunachalam and Singh, 1984). Garfield associated the key research papers identified by the high number of citations, with the milestones of the progresses of superconductivity research, and indicated high coincidence (Garfield, 1988). He also used citations count to identify the high impact individuals, institutions, and journals, as well as to indicate research front in the field of superconductivity research. (Garfield, 1990) Chu concerned about the formal and informal structure of communications between Chinese and non-chinese scientists in 12 / 95

13 superconductor research. She identified notable similarities in formal communications: analogous cited cores, identical publication sources, and comparable intellectual structures of co-citation data (Chu, 1992a); and the lower visibility of Chinese scientists compared to their counterparts in the informal communications (Chu, 1992b). Chu also studied the formal communications among Chinese superconductor researchers in a later research, based on citation statistics from Chinese Science Citation database (Chu, 1998). Cardona and Marx studied the impact of the works by Vitaly Ginzburg, one of the most influential scholars in superconductivity research, with particular attention to his most frequently cited articles and books and their impact on recent research (Cardona and Marx, 2006). 13 / 95

14 2.4 Citations Analyses in the Measure of Nations Contribution There are also many examples where citations analyses have been used to measure nations contributions in both specific subjects and more general science and technology. The following are only a few examples. Arunachalam and Singh analyzed superconductivity research papers by Israeli scientists indexed in Physics Abstracts , as well as the relevant citations records, and concluded that Israel performed quality research in the field of superconductivity, almost as good as the leading country USA. (Arunachalam and Singh, 1985) This research, as to our knowledge, has been the only attempt thus far to measure country performance in superconductivity research by now. Jain and Garg studied the patterns of publications and citations of laser research by India during with data from Journal of Current Laser Abstracts and Science Citation Index, and were convinced that Indian performance in laser research was part of the mainstream science (Jain and Garg, 1992). Uzun et al explored Turkish physics publications that appeared in the list of SCI source journals during , and discovered that papers from Turkey that appeared in the American and European journals are cited at rates higher than the corresponding average' papers (Uzun et al., 1993). Kim measured research performance of Korean physicists with data from SCI , and compared Korean-authored papers versus internationally co-authored papers. Based on the citations count, Kim revealed that papers with U.S. co-authors received the highest citation rate, and there had been significant differences between the average 14 / 95

15 citations per paper of different countries (Kim, 2001). Sometimes researchers tend to introduce or employ specific scales in the measurement of country performance: Arunachalam and Doss analyzed papers indexed in three years of Biochemistry and Biophysics Citation Index. They used the sum of Impact Factors (IFs) of the journals in which papers had been published as a rough measure to scale institutions research contribution. They identified great portion of papers with international co-authors, and found that international collaboration raised the IFs value of Israeli publications (Arunachalam and Doss, 2000). Garg investigated publications of laser research by India and China at conferences and in journals during 1993 and 1997, and used the following measures to quantify the research impact: Normalized Impact Factor, Normalized impact per paper, Proportion of papers in high quality journals, Publication Effective Index (Garg, 2002). These measures were used again later in Kumar and Garg s analyses of publications by India and China in computer science during , to reflect the research impact of these two countries in computer science (Kumar and Garg, 2005). There are also researches which involve comparative studies of many nations performances in science and technology. King analyzed the publication and citations numbers of different nations, and quantified the nations impact in science and technology during He compared 31 countries which contributed altogether more than 98% of the world s total number of highly cited papers, defined as the most cited 1% papers by field and year generated from Thomson ISI. He used indicators as: total number of publications and citations, the share of top 1% highly cited publications, the number of citations per paper, publications and citations per researcher, and 15 / 95

16 sometimes combined with financial/economical indices such as GDP, HERD (Higher Education funding of R&D), etc. Based on share of top 1% of highly cited publications during , he produced a list of ranks for different nations, among which China was at the place of 19th. Based on the share of citations, he also identified the national strengths in different disciplines in seven categories: clinical medicine; preclinical medicine and health; biological sciences; environment; mathematics; physical sciences; and engineering (King, 2004). Guan and Ma measured China s impact in semiconductor research based on the publications and citations data of from ISI Web of Sciences database. They focused on five Asian countries/regions, but also included Germany and USA for comparative studies in a broader extent. They counted publications and citations, adopted the average number of citations per article published in a country, the uncitedness, and the percentage of highly cited papers as measures for the impact analyses, which revealed that Chinese scientists in semiconductor research have a low international visibility. They also studied Leimkuhler curve and Gini indices for the inequality of citations times and informatric productivities among different countries (Guan and Ma, 2007). 16 / 95

17 2.5 Web of Sciences, Scopus, and Google Scholar For decades, citations analyses had solely depended on the citations data source from Science Citations Index (SCI) or Web of Sciences (WOS). However, Scopus from Elsevier and Google Scholar from Google started to provide citations data since By now, there have been many other databases with citations launched and in use, however, WOS, Scopus and Google scholar remain the principle ones when referred to citations databases, and some researches have already been carried out to compare the performances of their citations data sources. Bar-IIan et al. have proposed a set of measures to compare these citations databases, and also tried applying these measues to the publications of highly cited Israeli scientists for feasibility tests. The results indicate high similarities between the rankings of the ISI Web of Sciences and Scopus and lower similarities between Google Scholar and the other tools (Bar-Ilan et al., 2007). However, Falagas, et al s study in biomedical information indicates that Scopus offer about 20% more citations than Web of Sciences (Falagas et al., 2008). On the other hand, Scopus has been pointed out to have more coverage in journal titles while at the meantime have its limitations of publication years which include merely 1995 onwards. (Falagas et al., 2008), and Google Scholar has been identified to have its different indexing strategies of including books and proceedings, which often attract more citations than journal articles. (Bar-Ilan et al., 2007) Meho and Yang s case study in citations of library and information science advocates that Google Scholar has its strength in the coverage of conference proceedings as well as international, non-english language journals (Meho and Yang, 2007). With regard to database reliability, Jacso has reported with that the citations count from WOS gives out only the exact match, however, with 17 / 95

18 facilities to allow the searchers to visually scan and check the imperfect but close matches (Jacso, 2006a). Bar-IIan have uncovered that citations from Google Scholar often include both preprint and journal version of the same paper, which contribute to duplicate count of citations (Bar-Ilan, 2008). Jacso have also reported with figures about the duplicated, non-scholarly, and error citations in Google Scholar, which are meant to cause the phantom or inflated citations count (Jacso, 2006b). There is also a detailed and justified research reported by Bakkalbasi, et al., which counted citations of articles from two different disciplines of oncology and condensed matter physics, and of two different years of 1993 and The results revealed that: all three tools, i.e. WOS, Scopus and Google Scholar, returned some unique citations; however, WOS returned the most citations for condensed matter physics of both 1993 and 2003 and for oncology of 1993; while for oncology 2003, Scopus returned the most citations, and Google Scholar returned the most unique citations. Furthermore, it was concluded that the question of which tool provides the most complete set of citing literatures may depend on the subject and publication year of a given article (Bakkalbasi et al., 2006). Bakkalbasi, et al s conclusion is of great significance because it could be used to well explain the inconsistency in different research results, while it is has been well supported by the major findings with regard to the differences of different databases reported by other researches. It could also be served as evidence that WOS should be an appropriate data source for this specific study, since superconductivity is a branch of condensed matter physics, for which WOS has reported to be the most suitable data source for citation analysis. 18 / 95

19 Chapter 3 Methodology 3.1 Data Source: Web of Sciences Although there are other sources with citations statistics, such as Google Scholar, Scopus, etc., the previous research by Bakkalbasi, et al which is metioned in Chapter 2 has implied that Web of Sciences should be the most suitable data source for this specific study. 3.2 Data Selection Criteria: original articles 1986 was the starting point of the explosive development of research in high temperature superconductivity, as it has been reported in the section of literatures review was the last year that the complete year-round publications are available. It is intended that the inclusion of citations data of the complete set of publications should reveal a holistic picture of the impact of the superconductivity research of China in the new high temperature era. The following search terms are used for the complete set of publications in superconductivity related research of China during within the database of Web of Sciences. TS=(superconduct*) AND CU=(China) Timespan= Databases=SCI-EXPANDED Where superconduct* is used for key word search, * is used as truncation to represent superconductor(s), superconducting, superconductivity, etc; and the country is limited to China, time span is limited to / 95

20 The result reveals that 7,241 papers including articles, notes, letters, reviews, editorial materials, meeting abstracts, corrections have been published. Table lists the relevant record count and the percentage rate of different document types of publications. Document Type Record Count % of 7241 ARTICLE % NOTE % LETTER % REVIEW % EDITORIAL MATERIAL % MEETING ABSTRACT % CORRECTION % Total: % Table Document Types of Publications of China Since this study aims to evaluate the original important contributions by China in superconductivity research, and since articles are of the most important original publications here above, it is decided that the publications and citations analyses are based on articles only. 3.3 Data Analysis Criteria: non-self-citations Citations analyses are primarily based on the number of citations excluding self-citations. However, in the comparative studies where h-index is used, h-index based on non-self-citations is not available. In this case, the h-index based on the total numbers of citations is used in order to obtain some overall understanding about the situations. Although it is arguable that self-citations may also reflect the impact of a publication, it is evident that this impact is limited within the authors research group. Since this study aims to evaluate the impact of original research by China in the broad extent of the academic world, it is reasonable to exclude self-citations in citations analyses of China as well as the comparative studies of citations of different countries.. 20 / 95

21 3.4 Data Classification Scheme: JCR scheme JCR classification scheme is used for subject analyses of original publications as well as citations of publications. Since JCR is used by the database of Journal Citation Report, which is a by-product of Web of Since, and since this study is of citations analyses carried within WOS database, it will be more consistent using JCR classification scheme for data analyses. 3.5 Research Techniques and Approaches WOS has provided excellent citations analyzing tools, i.e. analyzing publications and citations by countries, document types, institutions, languages, publication years, source titles, subjects, etc. It also presents citations reports, provided with detailed results of non-self citations. These have been used for most of our data analyses. Besides, Microsoft Excel is used when there is the need or the WOS analyzing tools become inefficient, e.g. to compare the data sets of different countries, to sort out the publications by reprint authors, etc. The following major aspects are looked while data are analyzed: trend of growth or decrease of publications and citations over the years; coverage of source journals of publications and citations; impact factors of source journals; subject areas of original publications and citations, key institutions in terms of the original publications and citations of the top institutions of different countries; International collaborations and whether or not they have helped to boost the number of citations. Comparative studies of different countries are emphasized on similarity or dissimilarity in patterns rather than bench marking, although at some point this might seem evident and inevitable. 21 / 95

22 Chapter 4 Limitations of the Research This study is not complete and has inevitable limitations in terms of data sources, methodology, technical analyzing tools, and time of study. 4.1 Limitations by Data Sources Although Web of Sciences is supposed to provide the best data sources for citations analyses in this study, it is not perfect. There are some inevitable faults in the data sources sometimes, such as errors of data entries, etc. For example, Web of Sciences analyzing tools sometimes may provide the results with a note such as 12 records (0.0763% of total) do not contain data in the field being analyzed, which is obviously a kind of faults due to the errors of data entries. It is believed that such minor faults of data may not affect the overall results of this study, although it is definitely better to have a data source without faults. Besides, since this study is carried out in the middle of 2008, the citations data for some of the 2007 original articles are less than one year, which makes the citations analyses incomplete in this case. 4.2 Limitations of Methodology of Citations Analyses The most distinct limitations of citations analyses with regard of this study should be self-citations and reasons of citing. In this study, self-citations are excluded, and self-country citations are also excluded in the comparative studies. However, reasons of citing do affect the way how to perceive the citations of an original article, as some original articles might have presented incorrect theories or inaccurate experimental 22 / 95

23 results and hence citations of such articles are only to point out the mistakes or inaccuracies, which is on the contrary of the idea that citations reflect the impact of the original articles on later researches. Since this study is intended to reveal the impact of China in superconductivity research, in which case, the impact is supposed to be perceived in the positive way, and hence it is regretful that there is by no means to distinguish the reasons of citing. However, for the comparative studies of other countries, since this is the same problem for all countries under study, it is believed that this problem won t affect much the relative results of different countries. 4.3 Limitations of Technical Analyzing Tools The main analyzing tools used for this study are Web of Sciences analyzing tools and Microsoft Excel. The manual works on Microsoft Excel are based on the data deprived from Web of Sciences analyzing tools. Web of Sciences analyzing tools are not able to generate citations report for more than 10,000 original articles, and therefore, the comparative studies of citations analyses have to be carried out based on original publications of and with subject focus in three top areas. This has caused inconsistency of the comparative studies, as the publications analyses of original articles are of but citations analyses are of different years. 4.4 Limitations of Time of Study This study has been carried out within only 2 months, and with such a limited time, there couldn t be thorough and complete investigations in every possible aspects. 23 / 95

24 Chapter 5 Original Articles by China In order to obtain an overview of the original articles by China, a series of statistical analyses are presented in this chapter. Besides, comparative studies of original articles by different countries are presented in Chapter Growth of Articles through the Years Table-5.1 lists the record count and percentage of different years of the original articles in superconductivity research by China. Publication Year Record Count % of % % % % % % % % % % % % % % % % % % % % % % Total: % Talbe-5.1 Articles of the Years 24 / 95

25 It is clearly seen that there is a continuous growth in the number of articles published through the years (Chart-5.1.1). Detailed analysis reveals that the rapid growth started from 1999, as there were only less than 30% articles published during the beginning half of the period, while on the other hand there were more than 70% articles published during the later half of the period (Chart-5.1.2) , 6.78% 2003, 7.51% , 8.06% Chart Original Articles: growth through the years 2006, 7.88% 2007, 8.74% : 29.34% 1991, 3.41% 1992, 2.75% 1993, 2.81% 1994, 4.00% , 3.80% 2002, 5.58% 2001, 5.22% 2000, 6.28% 1999, 3.77% 1998, 4.03% 1997, 6.82% 1990, 2.62% 1989, 2.36% 1988, 1.88% 1987, 1.35% 1996, 3.97% 1986, 0.39% Chart-5.1.2: % of Original Articles of the years 25 / 95

26 5.2 English is the Primary Language Analysis by languages reveals that English is the primary language used for original articles in superconductivity research by China, as 95.71% articles are in English (Table-5.2). Further analysis by publication years identifies a general trend of growth in English articles published through the years (Chart-5.2). Language Record Count % of 6977 ENGLISH % CHINESE % GERMAN % JAPANESE % Total: % Table-5.2 Analysis by Language Chart-5.2 English Articles: growth through the years Since English is the language of science, and it is the most commonly used language in the academic world, articles published in English usually have greater influence than those in other languages. With increasing number of articles published in English, China has obviously become more visible and influential in the world wide community of superconductivity research. 26 / 95

27 5.3 Subject Areas Analysis by subject areas has disclosed that all original articles by China in superconductivity research are published in 69 subject areas. This indicates that superconductivity research in China has covered a broad range of subject areas. Table-5.3 lists the focus original subject areas of the original articles by China in superconductivity research, which are defined as that each of these subject areas has included at least 1% original articles by China. It is seen that there are 19 such subject areas, with the top 3 subject areas as: applied physics, condensed matter physics, and multidisciplinary physics, which have published most original articles. This indicates that superconductivity research of China is very much focused in these physics subject areas. Subject Area Record Count % of 6977 PHYSICS, APPLIED % PHYSICS, CONDENSED MATTER % PHYSICS, MULTIDISCIPLINARY % MATERIALS SCIENCE, MULTIDISCIPLINARY % ENGINEERING, ELECTRICAL & ELECTRONIC % PHYSICS, MATHEMATICAL % CHEMISTRY, PHYSICAL % METALLURGY & METALLURGICAL ENGINEERING % MULTIDISCIPLINARY SCIENCES % PHYSICS, NUCLEAR % PHYSICS, PARTICLES & FIELDS % NUCLEAR SCIENCE & TECHNOLOGY % PHYSICS, ATOMIC, MOLECULAR & CHEMICAL % PHYSICS, FLUIDS & PLASMAS % CHEMISTRY, MULTIDISCIPLINARY % MATHEMATICS, APPLIED % THERMODYNAMICS % OPTICS % MATHEMATICS % Table-5.3 Focus Original Subject Areas of China 27 / 95

28 5.4 Source Journals Analysis by source journals has disclosed that all original articles by China in superconductivity research are published in 422 source journals. Table lists the core original source journals of China, which are defined as that each of these journals has published at least 1% original articles by China. It is seen that there are 18 such core original source journals, and they have published more than 2/3 of the original articles by China in superconductivity research. Source Title Record Count % of 6977 PHYSICA C % PHYSICAL REVIEW B % SUPERCONDUCTOR SCIENCE & TECHNOLOGY % SOLID STATE COMMUNICATIONS % CHINESE PHYSICS LETTERS % IEEE TRANSACTIONS ON APPLIED SUPERCONDUCTIVITY % CHINESE PHYSICS % JOURNAL OF PHYSICS-CONDENSED MATTER % PHYSICS LETTERS A % ACTA PHYSICA SINICA % COMMUNICATIONS IN THEORETICAL PHYSICS % INTERNATIONAL JOURNAL OF MODERN PHYSICS B % JOURNAL OF SUPERCONDUCTIVITY % JOURNAL OF APPLIED PHYSICS % CHINESE SCIENCE BULLETIN % APPLIED PHYSICS LETTERS % PHYSICAL REVIEW LETTERS % CRYOGENICS % Total: % Table-5.4.1: Core Original Source Journals of China Since Journal Impact factor is a well known indicator which implies the general academic influence of an academic journal, it is attempted to look at the impact factors of these core original source journals of China in order to have a general impression about their academic influences. The result reveals that the average journal impact factor of these journals is (Table-5.4.2). 28 / 95

29 Abbreviated Journal Title Impact Factor APPL PHYS LETT CHINESE PHYS CHINESE PHYS LETT CHINESE SCI BULL COMMUN THEOR PHYS CRYOGENICS IEEE T APPL SUPERCON INT J MOD PHYS B J APPL PHYS J PHYS-CONDENS MAT PHYS LETT A PHYS REV B PHYS REV LETT PHYSICA C SOLID STATE COMMUN SUPERCOND SCI TECH AVERAGE: Table Journal Impact Factors 1 Since most articles are published in top 3 physics subject areas: applied physics, condensed matter physics, and multidisciplinary physics, the attempt is made to look at the average journal impact factors of these three physics subject categories. This is for comparison between the two average impact factors in order to reveal the relative academic influence of the core original source journals of China. Analysis of data of impact factors from Journal Citation Reports reveals that the average journal impact factor of the above mentioned three physics subject categories is (Thomson Scientific, 2007; Thomson Scientific, 2008a). The closeness of the two averages (2.011 vs ) may lead to the conclusion that the core original source journals of China are at about the average level in terms of academic influence. 1 Note: Journal Impact Factors are not available for the following two journals: ACTA PHYSICA SINICA and JOURNAL OF SUPERCONDUCTIVITY 29 / 95

30 5.5 International Collaborations Analysis by countries reveals that a substantial amount of original articles by China in superconductivity research are of the results of international collaborations, with a total number of 53 collaborative countries which count for nearly 30% collaborative research activities with regard of the total number of articles (Table-5.5.1). Country/Territory Record Count % of 6977 USA % JAPAN % GERMANY % AUSTRALIA % FRANCE % SOUTH KOREA % ENGLAND % ITALY % CANADA % BELGIUM % RUSSIA % NETHERLANDS % SINGAPORE % SWEDEN % SWITZERLAND % SPAIN % BRAZIL % POLAND % AUSTRIA % ISRAEL % DENMARK % NEW ZEALAND % NORWAY % INDIA % BYELARUS % PAKISTAN % SLOVAKIA % UKRAINE % CUBA % CZECH REPUBLIC % FED REP GER % SCOTLAND % SIERRA LEONE % 30 / 95

31 USSR % BULGARIA % BURUNDI % GREECE % IRELAND % MALAYSIA % MEXICO % THAILAND % TURKEY % ARGENTINA % ARMENIA % FINLAND % JORDAN % MOLDOVA % PORTUGAL % REP OF GEORGIA % ROMANIA % SERBIA MONTENEG % SLOVENIA % UZBEKISTAN % Total: % Table Collaborative Countries of China Since the reprint authors are usually responsible for the articles, and since this study aims to evaluate the original contributions of China, therefore, it is of obvious interest to have a look at the total number of reprint authors of China in the international collaborations. Detailed investigation into the database has revealed that 88% of the reprint authors of the complete set of articles are of China, while only 12% are of other countries (Talbe-5.5.2). Country/Territory Record Count % of 6977 China % Other Country % Total: % Table Number of Reprint Authors in Total Number of Articles 31 / 95

32 In Table-5.5.2, the countries of reprint authors are defined by the authors affiliations rather than nationalities. This is because the researches are actually carried out in the authors work places, and it is fair to attribute the research results to the countries of the research institutions. Compare Table and Table-5.5.2, it could be reasonably concluded that China has been playing an active role in the international collaborations in superconductivity research, and has been the main contributor of most international collaborative researches in which it is involved. However, there are some difficulties in estimation of the exact percentage rate of the contributions of China because the record count in Table includes duplicate numbers as sometimes more than two countries may work together for one research project. 32 / 95

33 Chapter 6 Citation Analyses of China By the time of this study, there have been 29,838 total number of citations to the 6,977 original articles in superconductivity research by China published during , among which the number of non-self-citations is 15,735 (Thomson Scientific, 2008c). A series of detailed citation analyses based on the non-self citations are carried out in this chapter. In addition, the comparative studies of citation analyses of other countries are presented in Chapter Citations by Document Types Table-6.1 lists the citations count by different document types. It is seen that most citations come from articles, and some are from reviews, notes, news items and other publications. The citations of articles and notes generally indicate the impact of original articles on the new original researches, and those of reviews and news items usually indicate the important interest in the original articles recognized by research peers in the academic communities. Therefore, it is impressive that the original articles have received 14,497 article citations and 842 review citations. Document Type Record Count % of ARTICLE % REVIEW % NOTE % LETTER % EDITORIAL MATERIAL % CORRECTION % CORRECTION, ADDITION % NEWS ITEM % MEETING ABSTRACT % Total: % Table-6.1 Analysis by Document Types: total non-self-citations 33 / 95

34 6.2 Citations Growth through the Years Citations through the years have been listed in Table-6.2. It is clearly seen that there has been continuous growth in the number of citations through the years, with the rapid growth started from 2001 (Chart-6.2). The continuous growth in citations numbers through the years is partly because citations accumulate by years, and within each year, citations include not only current articles but also previous ones which may date back to a few years or even many more years ago. However, the continuous and rapid growth in citations since 2001 is also partly attributed to the continuous and rapid growth of the original articles from 1999 (Chart-5.1.1). The delay of years is because citations are always one or two years behind the original publications. On the other hand, the continuously increasing number of citations also indicates that the general impact of the original articles by China in superconductivity research has been increasing gradually. 34 / 95

35 Publication Year Record Count % of % % % % % % % % % % % % % % % % % % % % % % % Total: % Table-6.2 Citation numbers through the years: non-slef-citations / Chart-6.2 Citations Growth through the years: non-self-citations

36 6.3 Citations by Different Languages The original articles by China in superconductivity are published in only 4 different languages: English, Chinese, Germany and Japanese. However, citations of the original articles are of 12 different languages (Table-6.3). Language Record Count % of ENGLISH % CHINESE % RUSSIAN % JAPANESE % GERMAN % FRENCH % SPANISH % PORTUGUESE % UKRAINIAN % CZECH % KOREAN % WELSH % Table-6.3 Analysis by languages: total non-self-citations There is no surprise that English is the primary language of citations as it is the language of science and also the primary language of the original articles. It is not unexpected either, that many citations are in Chinese, Japanese and German because these are the languages which are also used by the original articles. However, it is impressive that some of the citations are in other languages which are not used by the original articles. This indicates that the original articles by China in superconductivity research have their impact beyond the language barriers. 36 / 95

37 6.4 Citations from Different Countries/Territories Citations of the original articles in superconductivity research by China are from a total number of 104 countries/territories all over the world. Table-6.4 has listed the important citing countries which have contributed at least 1% of the total number of citations. It is seen that China itself contributed the most citations with 23.67% of total, followed closely by USA (21.70%) and Japan (15.89%) not far away from. However, the biggest citing group is the European countries, as Germany, Russia, France, Italy, England, Switzerland, etc., have added up to over 50 % citations in total. Country/Territory Record Count % of CHINA % USA % JAPAN % GERMANY % RUSSIA % INDIA % FRANCE % ITALY % SOUTH KOREA % ENGLAND % SWITZERLAND % POLAND % AUSTRALIA % CANADA % AUSTRIA % SPAIN % SLOVENIA % BRAZIL % UKRAINE % SWEDEN % NETHERLANDS % BELGIUM % ISRAEL % TURKEY % Table-6.4 Important Citing Countries: total non-self-citations 37 / 95

38 Compare to the countries involved in the international collaborations with China, the total number of citing countries is almost doubled. 50 more countries other than the original collaborators of China, have contributed to the citations. This indicates that the original articles by China have their impact in a broad range of countries. On the other hand, detailed analysis of the citations by important collaborative countries reveals that the majority of citations come from the collaborators of China, as they have contributed 70% of the total number of citations (Chart-6.4). This indicates that the most distinct impact of the original articles in superconductivity research by China is contained within the collaborative countries of China. Chart-6.4 Citations Distribution in different countries: total non-self-citations The important collaborative countries are defined as that the total number of collaborative articles by each of these countries in collaborations with China is at least of 1% of the total number of original articles by China. 38 / 95