Citation analysis

The citation analysis is a field of bibliometrics, which deals with the study of citations, ie the relationships between cited and citing works, and its application as a bibliometric research method.

As bibliometric parameters are different indicators are used, which are usually calculated using statistical methods, with certain assumptions about the Zitierverhaltens of authors can be made:

• In general, a scientific work is considered influential when it is quoted by other authors often. The indicator here is used to Zitationswert, sets the citation frequency in relation to the total number of the quoted posts, where also its temporal evolution can be Also taken into account.
• Assuming a relationship between several works that are cited together or another quote together, so you can make with the help of the cluster analysis groups of authors and / or publications that may deal with a branch of science.
• If the possibility of substantive evaluation of publication consists, then, the correspondence analysis is a suitable tool for the creation of science country maps, you

Since the establishment of scientific publications is largely standardized and next references other data such as abstracts, full text, descriptors and addresses contain the authors, the citation analysis is increasingly used. For research and analysis of citations, there are citation databases, provide the information required in an open format ready. Since it is not often the results of citation analyzes expressed in simple numbers, and methods of information visualization are used to model the complex Zitationsbeziehungen.

Objectives of citation analysis are

• Professional content of relations among persons, institutions, publications, scientific disciplines and their impact (Impact)
• Research focuses on scientific work and their development over time
• Activities and topicality of research groups within the sciences as well as "hot" topics and discussions
• Groundbreaking publications that form the basis for further research.
• Relationships between research and industry ( citations from scientific literature in patents )
• Use as an evaluation tool for peer review
• Aspect for assessing the quality of scientific work as a basis for public research institutions

History

Zitationsgraphen

A lot of citations and related publications can be described mathematically as a graph, then as Zitationsgraph (also Citespace ) is called. Come to analyze, among other methods of network analysis are used. In the form Zitationsgraphen publications (usually scientific articles ) nodes between which citations span the edges. Alternatively, all articles of a journal to summarize a node to compare different journals. Other possibilities for the construction of graphs of citation data is information on the co-authorship and Kozitation. The most important proposals based on citation relations between two publications A and B are:

• Citation: An article A cites an article B
• Kozitation: Two items A and B are cited together by an article C
• Bibliographic coupling: Two items A and B cite a common item C

From studies of Zitationsgraphen various regularities and structures on the publication behavior of authors can be read. It turns out that Zitationsgraphen scale-free and form small-world networks. Cluster analysis and other techniques of network analysis communities can be found by are often mutually citing researchers.

In the webometrics ( Sitationen ) are examined analogous to Zitationsgraphen linking graph of Web pages and hyperlinks.

Sources of error

The results of the citation analysis can be distorted by corrupted indicators such as the rate of publication of authors. Thus authors are cited more often by a large number of publication, which is reflected in the statistics, but this does not mean that they or their contributions are more significant than others. Some authors publish the same scientific results several times under different titles or in different languages.

Statistical artifacts caused by the fact that in publications with multiple authors, not all authors are considered. The weighting of the individual authors work share can be utilized in the analysis hardly, as the author's name is often stored for technical reasons alphabetically. Also, typing and transcription errors that arise in the collection of names in heterogeneous character set systems that fall into this category.

Incomparable comparisons then arise if, for example, rankings are created for magazines that neglect available information such as publishing cycles, rate of publication. Also quotes from review articles that merely provide an overview can be filtered without substantive evaluation by statistical automatisms hardly.

Unauthorized use of mathematical or statistical methods whose conditions are not fulfilled or violated without making appropriate corrections, leading to inefficiencies. These can be an insufficient number of documents to be like in the case of Multidimensional Scaling or the lack of a normal distribution or an invalid data independence in regression analysis.

While quoting his own works or as a favor for reasons deemed to be frivolous, however, if there are factual or more strategic reasons, can often be difficult to identify.

Not taken into account differences in the type of referencing in various disciplines.

Problem

Since it is assumed that good works are cited more often, the number of citations that undergoes a scientific contribution, often used as a measure of its quality. However, this common practice is not without problems, as sometimes the content review threatens to become neglected and give purely statistical criteria decisive. The fact that a particular author is often cited, can have a wide variety of reasons and not allowed in any case, a statement about the quality of his contribution. Also, it may form so-called Zitierzirkel come ( " you are quoting me, cited ' I love you " ), which distort the results ( boshafterer Name: Zitierkartell ).

Specific criticisms of citation analysis

• The value measured by the number of citations in scientific papers Impact of a work ignores the impact that a work may have in other areas ( for example in the industry that do not publish their results )
• By Matthew Effect frequently cited work without looking at the contents are in turn cited more often. A study by M. V. and V. P. Roychowdhury Simikin suggests that only about a quarter of the cited papers are read by the authors at all
• Some citations are only added to increase the impact factor of a journal or an author. Also publications be optimized to achieve a high number of citations, rather than to optimize the quality.
• The importance of articles in professional journals and proceedings and monographs varies significantly depending on the subject area. In citation databases, these differences are not taken into account, so that individual areas of expertise are undervalued or overvalued.
• The most cited papers are often exceptions that are cited mainly because it is customary to cite this work. Other works similar in influence other hand, are not cited explicitly, because their contents have become a matter of course.
• The calculation of the impact factor is always time-related with a maximum period of 6 to 8 years (two PhD - periods), which is reached on average after the first two years of Zitationspeak. However, depending on the activity of the research field, this value can vary greatly, which can lead to errors in the Berechunung of Impact Factors and half-life of the citation. Similar errors can occur when the periodicity of the publications is not considered.
• Different citation individual works are treated in citation databases as different publications. For example, the famous Paper Initial sequencing and analysis of the human genome, the human genome project in 2001 in the ISI was initially listed with different authors, making it did not show up among the most cited papers.
• Some non-English language journals will also appear in an English edition. However, this is often mistakenly counted as independent magazine.

Science Impact Index

Despite the undoubtedly legitimate criticism of individual citations has the Science Impact Index ( SII ) has proven as a measure of research quality of a scientist as a valuable objective guidance. The SII is a citation. It corresponds to the number of scientists who cite the work of an author within a year in a journal with an impact factor (IF). It also books, book chapters and publications are recorded in journals without IF, namely, when they are quoted in IF journals. At least in scientific and medical fields of SII can at scientists who for seven years had been at least possibilities to explore, such a reliable positioning among peers and such a stable prediction of the following years, as it does not reach through other dimensions and reviewers so far be.