Information needs analysis of the aerospace discipline

Information needs analysis

Several techniques for information needs analysis are available (Nicholas and Herman, 2010), but for collection development they typically take one of two different approaches. One is to gather information by means of questionnaire survey, interview and other data collection methods that ask researchers first‐hand about the information sources they use in and need for their research. The other approach relies on bibliometric analysis to identify the literature actually published and cited in the discipline; the results of the citation analysis are used to determine what resources a collection must include to provide for the needs of researchers in the field. Modern studies of scholarly information needs tend to use bibliometric analysis more often than survey analysis. Nonetheless, this review first introduces some survey research to provide a sense of the difference in the nature and application of the results of the two different approaches. Although research using each approach has been conducted in several countries, the review draws primarily on Chinese language literature which is not readily available to English language readers.

Surveys are often conducted to uncover information about several aspects of LIS service and provision. A typical example of the scope of these studies is found in Westbrook's (2003) questionnaire survey. Westbrook asked women's studies scholars several questions about library use, problems and support as well as what information resources were needed and used. Wu et al. (2006) interviewed ten professors from the Department of Chinese Language and Literature and the Department of History in Taiwan, to explore how humanist scholars would accept electronic resources. In the sciences, Ye and Mie (2010) conducted a questionnaire survey at Hebei medical colleges and universities to analyze users' demands, utilization characteristics and obstacles to use. The survey approach is limited in the number of questions that can be asked about each topic, and in application by the nature of the people who choose to respond (e.g. the institution to which they belong or the geographical location). Depending on the method used and the skills of the survey personnel, results can be biased (Babbie, 2010). When an in‐depth understanding of the nature of publication in a discipline and the information needs of scholars across the discipline is required, survey research can mislead through localized or superficial research findings.

Bibliometric studies provide a way to study the literature that is actually being used and cited (Smith, 1981). To date, much of the work that uses this approach has been based on citation analysis of academic dissertations. An early study was undertaken by Kriz (1977) who analyzed citations in the masters and doctoral theses of the Department of Engineering Science at the Virginia University of Engineering in 1971‐1975 and made changes to collection development policies as a result. Almost 25 years later, Gooden (2001) undertook a similar study of the doctoral theses of the Department of Chemistry at the Ohio State University. This strand of research has also become popular in China. By means of citation analysis of the doctoral theses in organic chemistry at Beijing University, Ma (2003) was able to rank most influential journals the field. Li and Zhao (2005) ranked foreign journals by frequency of citation in applied chemistry postgraduate papers between 2001 and 2004 at Jinan University in order to investigate library collection adequacy, and citation analysis of economics and management doctoral theses 1987‐2002 at Nankai University resulted in decisions being made about information resource provision by the university library (Liu, 2005).

Availability of large‐scale full‐text databases of periodicals has resulted in an increase in citation analysis of published works in recent years. Mo and Sun (2004) examined the citations, formats and languages of 1,748 references cited in 222 research papers published in China by senior researchers at Guangxi University between 1995 and 2002, and concluded that library collections of foreign books and reference works should be strengthened. An analysis of the papers published by Chemistry teachers at Zhongshan University and indexed in the ACS, Elsevier, RSC and Wiley full‐text journal databases was undertaken by Xu and Zheng (2010). Yan and Sun (2007) analyzed the references in 172 research papers published in 2005 by researchers in the Department of Engineering Mechanics at Dalian University of Technology and found not only that the references cited by researchers were dominated by foreign journals but also that electronic resources had become an important component of the library's resources.

The reviewed citation studies share the limitation that they are based on resource utilization by scholars in a single institution or at a small number of institutions at which collection development decisions are to be made. Although such studies have wider coverage than surveys, they share risks that the scope of the discipline as a whole is not well mapped, and the needs of research and teaching that is broader than that already conducted at the institution and reflected in current use is not identified. This paper therefore takes a different approach to information needs analysis by citation analysis, and considers the discipline as a whole by analyzing publications by the leading researchers in the discipline.

High level research and high quality publication

In general, the leading researchers in a discipline can be recognized by the number of citations to their publications: the researchers who are the most highly cited (the most productive) are considered to be the leading researchers. This observation is recognized in citation databases. The Thomson Reuter's journal impact factor (JIF) is an indicator of how highly cited the papers published, in aggregate, in a journal are. It is calculated as the ratio of the number citations to papers in the journal in a period to the number of papers published in the journal in the period. Thus, just as citations to an individual's publications provide a generic indicator of an individual's influence in the discipline, the JIF provides a generic indicator of a journal's influence in the discipline. Thomson Reuters provides a JIF for the “most important and influential journals” for information retrieval (Testa, 2012). These journals form the core of its Web of Knowledge databases. Details of the journals themselves (including the JIF) are available from the Journal Citation Reports (JCR). Works that are published in the JCR journals, along with works that are cited in papers that are published in the JCR journals are indexed, and in many cases available in full text, in the Web of Science databases. Cited references are an important part of a complete paper not only because they show how the paper builds on previous findings but also because they reflect researchers' and readers' requirements for literature resources. This paper uses these resources to identify the structure of the literature of the aerospace discipline and the resources necessary to meet the information needs of high level researchers working in the field.

Data preparation

The basis for analysis was a set of scholarly publications selected primarily from the 27 journals included in the Web of Knowledge category “Engineering, Aerospace” in 2009 or 2010. (From here on, these journals are called the JCR journals.) Two databases were prepared: the Publications database consists of works about aerospace research; the second database (the Source database) consists of the works cited in the publications.

The publications were drawn from Science Citation Index Expanded (SCIE) for the period 2009‐2010. SCIE is the Web of Science database that indexes and provides full‐text access to journals in the “Engineering, Aerospace” category. To ensure comparability of the research, two journals that were indexed in only one of the two search years were included manually in the search. The keyword search term “aerospace engineering” was also used to identify journal papers in the discipline that were published outside the 27 JCR aerospace engineering journals. (Theses, editorial material and corrections were excluded.)

Data cleaning was necessary to standardize journal titles. Acta Astronautica, for example, appeared as both ACTA ASTRON and ACTA ASTRONAUT. Similarly, institution names were standardized, e.g. only one form of boeing Co and Boeing Co was used. The final set of publications consisted of 4,290 journal papers.

The Source database was derived from the works cited as references (sources) in the papers in the Publications database. This involved extracting the details of each publication's reference list from its record in SCIE. There were 17,015 references in all.

Data analysis

RQ1. Productivity

The 20 top‐ranking institutions, by number of SCIE‐indexed works in the field, are shown in Table I. These core research institutions in the aerospace discipline published 25.5 percent (1,094) of the 4,290 works in the Publications database.

Most of the institutions are in either the USA (nine) or China (five). A comparison of these institutions shows that, while there are many types of high‐level research institutions in the field in the USA; Chinese aerospace research institutes are concentrated in polytechnic colleges and universities. The structure of the aerospace research discipline in the two countries is, in fact, different. In China, there is only one non‐university research institution, the Chinese Academy of Sciences, whereas research in the USA is done by many types of organization.

Table II extends the information in Table I for Chinese institutions for which ten or more papers were included in the database; this table shows that both the Chinese Academy of Sciences and the National University of Defense Technology (NUDT) participate in Chinese aerospace research, although their output is not among that of the 20 most productive institutions internationally. Government authorities with strong aerospace research in the USA include NASA (the National Aeronautics and Space Administration) and the US Air Force Academy (USAF). Aerospace research is also undertaken by a number of enterprises such as the Boeing Company, as well as NGOs (non‐governmental organizations) such as the Aerospace Corporation whose publications were included in the database, but not in sufficient number to be included among the most productive institutions by publication in Table I.

Table II shows that researchers from the most productive institutions in China published 11.3 percent of the papers in the database. Grouping all the papers by country to rank the top countries by productivity shows that the USA (34.1 percent) and China (including Hong Kong and Taiwan) (15.2 percent) together published just under half of the aerospace research papers (Table III) in 2009‐2010. India (3.3 percent) was also ranked in a relatively high position, highlighting the scientific research strength of two developing countries (China and India) in the field.

Journal publication by Chinese aerospace research institutions (RQ1.1)

This analysis was restricted to the 25 journals that were included in JCR in both 2009 and 2010. Table IV shows that 509 (78 percent of the 651) journal papers from Chinese research institutions were published in JCR aerospace engineering journals in the period. Chinese research institutions have paid close attention to the important journals of this discipline, and the authors' scientific research achievements meet the requirements for high‐quality journals.

On the other hand, 191 (37.5 percent) of the papers seen in Table IV were published in the Chinese Journal of Aeronautics, which has a JIF of 0.294 and is ranked 20th of the 25 aerospace engineering journals by JIF. The papers published by Chinese research institutions in the aerospace discipline and indexed in SCIE are therefore likely to have limited international influence.

RQ2. Citation analysis

After removal of non‐journal citations from the Source database, 14,618 references remained. Journals therefore accounted for 85.9 percent of the total citations, making journals the document type to which the scientific researchers in the aerospace field pay the most attention.

The most highly cited journals consisted of a set of 32 journals that were cited 35 times or more. This set included the 25 JCR 2009‐2010 aerospace engineering journals, which were cited 2,873 times (accounting for 19.7 percent of the total number of cited papers). The aerospace discipline journal papers indexed in SCIE have not been intensively cited in the core journals of this discipline.

Among the highly cited journals were the Annual Review of Fluid Mechanics and another 18 journals that are not among the journals in the aerospace engineering category in JCR. These journals, listed in Table V, are in fields that include fluid mechanics, applied mechanics and engineering, computational physics, engineering thermo‐physics, automatic control, signal processing and control and other basic disciplines or the sciences in general. The higher number of papers cited from these journals (80 percent) indicates that aerospace studies and aerospace research strongly depend on high quality journals in related disciplines and the sciences in general.

Journal citation analysis for Chinese aerospace researchers (RQ2.3)

A subset of the Source database was taken for the references cited in the 509 papers published by Chinese aerospace researchers. These papers have been cited for 1,500 times in 809 journals. A total; of 14 JCR journals were cited in 281 (18.7 percent) of the papers.

Highly‐cited journals not indexed in JCR included the Journal of Computational Physics and the other journals ranked in the top 13 in Table V. Journal information sources used by Chinese aerospace researchers are consistent with those used by researchers used in the discipline overall. There was no bias for journal papers in the Chinese language, although Chinese scientists did cite fewer papers from a smaller set of source journals.

RQ3. Online database coverage

Seven electronic databases were checked for coverage of the 32 most highly cited journals. ScienceDirect provided the best coverage, indexing ten of the journals; ScienceDirect should be used as key literature resource provision for aerospace research. A number of papers were indexed in three additional databases – Academic Source Premier‐ASP, the IEEE/IEE Electronic Library and AIAA Meeting Papers – which together covered 13 journals. These databases should also be included in the key literature provision for aerospace research. Although AIAA Meeting Papers covered only six journals, it indexed 1,994 (13.7 percent) of the total number of cited papers, by far the highest number of papers indexed in any of the databases. AAIA Meeting Papers was therefore the most important literature resource for aerospace researchers.

Three open access (OA) databases also provide coverage of aerospace journals: HighWire Free Online Full‐text Articles, PubMed Central and the Directory of Open Access Journals. Full text of papers indexed in two of the 32 journals (neither of them aerospace core journals) was available only from one of these databases and not from any of the reviewed commercial databases.

Limitations

The research has some limitations. Selection of papers published in 2009 and 2010 means that the results may be influenced by the topics and projects of interest in those years. In a fast developing field, any metric study of past publications runs the risk of missing a new sub‐field. Nonetheless, the high citation of both discipline‐specific and related discipline journals confirms the breadth of the disciplinary underpinning of aerospace research. Broad coverage of related disciplines should provide a sound basis for ongoing study.

Attribution of author's institution and country based only on first author is a limitation that should be addressed in future work. Work that was conducted across institutions or countries could not be identified with this method. In addition, commercial partners are not often named as first authors so commercial influence on aerospace research may be stronger than identified here. Furthermore, published work does not capture confidential research, so some aspects of aerospace research may be beyond from the subject coverage of the journals listed in this paper.