Bibliometric

Handbook
SM

Web of Knowledge Research Pathways

Bibliometric Handbook: Accelerating discovery Advancing knowledge

At Thomson Reuters we provide vital decision support data, vetted by experts, delivered online and accompanied by analytical tools and other software. We are making that information increasingly intelligent: self-describing, self organizing and capable of both action and interaction. We are building digital platforms that enable global collaboration, real-time data exchange, dynamic modeling of user behavior and context-aware applications.
Intelligent information is self-aware: it understands what it is about, whom it is for, where it came from and what its meant to do. It knows how to communicate these things about itself to others, whether those others are human beings or other intelligent agents. Intelligent information isnt science fiction. For the businesses and professionals we serve, its a necessity. We bring together human expertise, rich data and the technologies of the semantic web to deliver information you can trust. Intelligent information from Thomson Reuters it provides the knowledge to act.
SCIENCE

scientific and patent information databases, 110,000 meetings and conference proceedings and over 11,000 evaluated scientific websites. Century of Science and Century of Social Science now available through the Web of Science enable bibliographic reference searching of scientific discoveries from 1864 to the present. Thomson Reuters suite of research software solutions, including EndNote, and EndNote Web represents the industry standard for publishing and managing bibliographies. These products automate the creation of bibliographies for hundreds of scholarly publications, enabling and integrating searching, writing and bibliography creation.
What is bibliometrics?

Scientific and Scholarly Research From Dr. Eugene Garfields revolutionary concept of citation indexing and searching to the latest digital information solutions, Thomson Reuters has always had the same goalto supply quality information via the most powerful platform available, delivered in a way that is clear, useful, and effective. Today, reaching that goal means offering a powerful selection of integrated information solutions that provide not only relevant data, but also unique ways to search, analyze, evaluate and collaborate. Through a suite of content, tools and services that supports the scholarly research cycle, we provide researchers with information solutions that help organizations make new discoveries, analyze trends, measure performance, collaborate and publish research. Our global customer base comprises more than 3,500 institutions, national and regional consortia, and major government funding bodies in over 100 countries. Through ISI Web of Knowledge, we facilitate search and discovery by providing researchers and scholars with an integrated collection of databases covering more than 23,000 peer reviewed scholarly journals, leading

The origins of bibliometrics go back more than a century. The statistical analysis of scientific literature began more than 50 years before the term bibliometrics even was coined in 1969. The emerging field was described in terms such as Statistical Bibliography. Since then, bibliometrics has developed into an important tool in research evaluation Today, modern bibliometrics is often used to assess scientific research through quantitative studies on research publications. Bibliometric assessments are based on the assumption that most scientific discoveries and research results eventually are published in international scientific journals so that they can be disseminated widely and enhance related scientific efforts by other researchers. It is the application of bibliometrics in science policy and in management of funding that has sparked the major interest in the field that we see today. In 1961 Dr. Eugene Garfield revolutionized scientific research with his concept of citation indexing and searching. Garfield founded the Institute for Scientific

....................................................................................................................... Bibliometric Handbook

Information (ISI) 1958. The citation data from ISI was originally published in printed format but since 1966 it has been available in machine-readable form.This development culminated in 1997 when the Web of Science was released. With the growing use of BIBLIOMETRICS, the publication and citation data of Thomson Reuters provides a consistent, authoritative and clearly defined body of data for bibliometric analysis. For more information, see http://science. thomsonreuters.com/news/2008-07/8465001/
Web of Science

Author Keywords - All author keywords are indexed and searched in a Topic search. KeyWords Plus - KeyWords Plus is unique to Web of Science and consists of words and phrases harvested from the titles of the cited articles. KeyWords Plus is searched in a Topic search. Database Year - This is the Timespan - the year in which records were added to the Thomson Reuters database (e.g., Web of Science). The database year of a bibliographic database may include data from documents with different publication years. Compare publication year. Publication Year - This is the year in which a document was published. Compare database year.

Search - Search by Topic, Author, Group Author, Source Title, Publication Year, Address, and Conference. Abstract - This is a brief summary or description of the essential content from the source document. All author abstracts are indexed in English-language author abstracts in their entirety (where supplied by the publisher). Abstracts are made available in several product formats. Addresses - All author addresses are indexed and searchable. Reprint authors are identified and their e-mail addresses are provided when available. Advanced Search - Use 2-character tags, Boolean operators, parentheses, and set references to create your query. Results appear in the Search History at the bottom of the page. Article Type - In addition to research articles, many journals publish review articles and short items such as letters and editorials. To see the complete list of article types in your subscription dataset, select the Article Type report. Cited Reference Search - All cited references from each article are indexed and searchable. Search by Cited Author, Cited Work, and Cited Year. Remember: Secondary Cited Authors are automatically searched in Web of Science source records within your subscription. Chemical Search - This is only available in the Chemical Databases within the Web of Science. It can: Search for chemical compounds and reactions that match a structure query that you create. Search for data associated with compounds and reactions. Search for compound or reaction data without doing a structure search.

Author Names - All authors are indexed. Search using last name and up to five initials. This may be of interest to identify prolific authors in a specific country or at a specific unit. Web of Science contains the names of all authors associated with a source document such as: Journal articles, Proceedings papers, Letters, Meeting abstracts and etc. When searching the Author field for records, be aware that the system displays author names in various fields within a record. Journal author names appear in the Author(s) field within a record. Book author names appear in the Book Author(s) field within a record. Group Author- A group author is an organization or institution that is credited with authorship of a source publication such as a journal article or a book. The name of a group author (or book corporate author) may take the place of a list of author names if the number of author names is very large (that is, over 400).

Search History - This is a search query or sequence of queries that you save to either the ISI Web of Knowledge server or to the hard drive of your computer. There is no limit to the number of search histories you can save.

....................................................................................................................... BIBLIOMETRIC HANDBOOK

View / Manage Marked Lists - This page allows you to view and/or delete your marked records. Search Operators - Boolean search operators AND, OR, NOT, and SAME may be used to combine terms in order to broaden or narrow retrieval. Search using AND, OR, NOT,and SAME (same sentence) to create logical search statements. Truncation Symbols - Use truncation to retrieve plurals and variant spellings. * = zero to many characters; ? = one character; $ = zero or one character UT - This is a unique identifier for a journal article assigned by Thomson Reuters which is part of the bibliographic information within the Web of Science record. For example - 000078066600003 DOI - The Digital Object Identifier (DOI) is a system for permanently identifying and exchanging intellectual property in the digital environment. Example: DOI: 10.1134/S1061920808010020. See http://www.doi.org for more information. Titles - The full title is indexed and searched in a Topic search. Foreign language titles are translated into English. Cited References - Web of Science indexes extensive information about each article indexed in its products including the articles cited reference list (often called its bibliography). This information allows a user to search the citation indexes for articles that cite a known author or work. All cited references are searchable via the Cited Reference Search interface. References that appear in blue serve as links to other Web of Science source records. These links are limited by your subscription. References appearing in plain black text may be: References to books or other types of documents not indexed in Web of Science References to articles outside of your subscription limits Cited reference variants or works that were cited incorrectly by the source publication Cited Author - In the Web of Science, only the first cited author is indexed and displays with the reference. Secondary cited authors are searchable, but only for those records that are source records within your subscription limits. Distinct Author Sets - This Web of Science feature is a discovery tool showing sets of papers likely written by the same person. We recommend that you use this feature as a tool to narrow your search rather than as a definitive list of a specific authors works. Times Cited - The number of times a particular work or source article appears as a cited reference in other source articles from the more than 8,600 journals indexed for the Web of Science. Cited Work - All Cited Works are indexed. The full work title and article title will display for citations that refer to source records in Web of Science. Refine Results - Use Refine up to 100,000 results to find the top 100 Subject Categories, Source Titles, Document Types, Authors, Publication Years, Conference Titles, Countries, Institutions, and Languages.

Sort Results- Sort up to 100,000 records by: Latest Date (default), Times Cited, Relevance, Publication Year, Source Title, First Author, and Conference Title. Output Records or Save to Endnote Web - Output records, add to your Marked List, or save to EndNote Web. Quickly print, e-mail or save to a temporary marked list (500 records maximum), or save permanently to EndNote Web (10,000max). Click more options to save a range of records, adjust your saved fields, or export directly to ResearchSoft reference software you have installed on your desktop.

Funding Acknowledgment Web of Science Funding Acknowledgement field captures funding agency names and grant numbers. Many funding bodies mandate that articles based on the research they have supported must include an acknowledgement. Funding bodies can use this info to: Track the research output and influence for any funding body, a specific grant, or research program; Identify the strategic scope of a funding body; Identify vested interests. This can also be utilized by researchers to identify future funding opportunities; support an existing grant application by showing related information and evidence of previous performance. Example: National Eye Institute / R03 EY014357. This study was supported in part by the National Eye Institute (R03 EY014357) and the American Brain Tumor Association, Kathy Murphy Translational Research Grant...

Analyze Results - Like Refine, you can analyze up to 100,000 results. With Analyze you can out-put the results to Microsoft Excel to create your own graphs.

....................................................................................................................... Bibliometric Handbook

Citation Counts - These are an indicator of an articles impact and usefulness to the research community; they are the mode by which peers acknowledge each others published scientific research papers. The value of a citation is only as important as its source. Clearly a citation from a prestigious peer review journal has more value than a citation from non-scholarly material. See the URL for more information - http://thomsonreuters. com/products_services/science/academic/citation_ indexing/ Second-Generation Citation Counts -These are the sum of the citation counts of all the papers citing a target paper. This is a measure of the long-term impact of a paper which is similar in effect to the Google PageRank. Like first-generation citation counts, second-generation citation counts need to be normalized by field.

Citation Report - The Citation Report is only available in Web of Science. The Citation Report provides aggregate citation statistics for a set of search results. These statistics include: The total number of times all items have been cited The average number of times an item has been cited The number of times an item has been cited each year The average number of times an item has been cited in a year

Related Records- This is an exclusive value-added feature in Web of Knowledge citation indexes. Related Records enhances the power of cited reference searching by linking and displaying all the articles within each index that have cited references in common. When you find one record of interest, you can easily find additional articles that are related to your subject, items not easily accessible with traditional search techniques.

Citation Map - It presents the citation relationships (cited references and citing articles) between a paper and other papers in an interactive graphical tree format. For more information, please see the URL below http://wokinfo.com/products_tools/multidisciplinary/ webofscience/citmap/

H Index- The h-index is based on a list of publications ranked in descending order by the Times Cited. The value of h is equal to the number of papers (N) in the list that have N or more citations. This measure attempts to reflect both productivity (number of papers) and impact (number of citations) in one number.

....................................................................................................................... BIBLIOMETRIC HANDBOOK

This metric is useful because it discounts the disproportionate weight of highly cited papers or papers that have not yet been cited. For example below, the h-index of 3 indicates that in the dataset, 3 papers were cited at least 3 times each. The 3 articles with 3 or more citations appeared above the green line. Like citation counts and citations per paper, the H-index depends on time period and discipline. A proper use of the H-index will ensure that the time period of analysis and field of research is similar among the researcher/ institutions/research groups being compared. H-index Without Self Citations - An h-index based on times cited values calculated by subtracting the number of self citations from the times cited number. Thomson Reuters Full Text Links These links are designed to bring the literatures full text directly to the researcher as efficiently as possible. It provides the users with bi-directional links for navigation back and forth between records in ISI Web of Knowledge and corresponding full-text documents and references your library subscribes to. Whether your access is through the Internet, an institutional intranet, or a combination of both, Thomson Reuters Links enables links to your information resources no matter where they reside. The various links available are Full Text Publisher Links, OpenURL Links Resolver, OPAC/Holdings Links, and Patent Links. These links can be requested by opening an e-Ticket at http://science.thomsonreuters.com/ techsupport/

For more information on the benefits of ResearcherID, please refer to http://wokinfo.com/researcherid/ benefits/

ResearcherID Badge - The badge advertises the members ResearcherID profile on their Web page or Blog. The Badge creates a hovering display of recent publications, and allows viewers to also link to the members full profile in ResearcherID.

Citation Metrics - ResearcherID generate citation metrics with times cited information for items added from Web of Science H-index Citation distribution per year Total Times Cited count Average Times Cited count

Promote Scholarly Output

Web Services

Web Services - Web Services allows for a real-time lookup of Web of Science source record information against Web of Science using the institution subscription entitlements. When a match is found, Web Services will return Web of Science article information such as: article title, authors, source data, author supplied keywords, and more. The institution can use this service to link into Web of Science from their library web page or institutional repository and the following: Access to formatted, current information for enhancing the institutions repository. Automatic, real time querying of multiple records eliminating the need for manual searches. Deliverability of real-time Web of Science information into multiple institutional applications such as department web pages or institutional repository. Web Service Lite: This is a DIY service which the appointed administrator can perform an API query which returns five fields (data elements) from Web of Science to your system. Citation counts are not included. Authors All authors, book authors, and corporate authors ....................................................................................................................... Bibliometric Handbook

ResearcherID - Name ambiguity is a recurring issue that impacts research accuracy and quality, career advancement and tenure, global collaboration among researchers, and identification and attribution of funding for institutions and individual authors alike. By assigning unique identifiers to registered participants, ResearcherID helps to solve the common problem of author misidentification. But thats not all ResearcherID does, and its not a tool only for published authors.

 Article Title Source Includes the source title, subtitle, book series and subtitle, volume, issue, special issue, pages, article number, supplement number, and publication date Keywords all author supplied keywords UT A unique article identified provided by Thomson Reuters ResearcherID Batch Upload service This is a DIY service which the appointed administrator will undertake the batch uploading ResearcherID for the whole institution. This ResearcherID Web Services alleviate involvement from the busy researchers at your institution at the upload phase and increase visibility of their work by creating ResearcherID profiles for them and suggesting publication lists. Then use ResearcherID Download to track publication output and properly quantify research performance at your institution. ResearcherID Upload is a web-based service that enables administrators to mass create ResearcherID profiles and upload publication data for some or all of the accounts you create for faculty, researchers, etc. at your institution. For more information on this, please refer to the URL below. http://researchanalytics. thomsonreuters.com/solutions/researcherid/ ResearcherID Download Service This service is available after the Upload service is completed at the institution. With ResearcherID Download you can gather even more information about your researchers and their publishing performance to make strategic decisions for your institution. ResearcherID Download is a subscriptionbased web service that enables you to query ResearcherID for researchers at your institution and return publication data for them, including times cited counts where applicable, as well as return institution affiliation for researchers at the requesting institution. Academic administrators and librarians are the primary audience; however, the service would need to be implemented by an IT librarian/staff as use of the service requires familiarity with XML and web service protocols. Article Match Retrieval - Thomson Reuters Links Article Match Retrieval Service is tool for academic administrators, librarians, and journal publishers to perform a real time query against the Web of Science or the Journal Citation Reports for matching records. If your institution subscribes to Web of Science or Journal Citation Reports you can now gain easier access to data such as citation counts to enhance your institutional repository and publisher pages. The service is not meant for use by end-users. This tool allows for a real-time lookup of bibliographic metadata such as DOI, author, source title, etc., against

the Web of Science database (using the institutions subscription entitlements). If a match is found, the service will return Times Cited information as well as links to view the full record, related records page, or citing articles page in Web of Science. An institution can use these links as a way to link into Web of Science from their library web page or institutional repository. Subscribers to Journal Citation Reports can use this service to retrieve links to the JCR record for a given journal. For more information on this, please refer to the URL - http://isiwebofknowledge.com/ directlinks/amrfaq/ Web Tools - These are useful web tools users can add to their desktop and initiate a search without opening their browser. Add Web of Knowledge to your web browser search bar and search this vast resource from your browser. Add the ISI Web of Knowledge search box to your web page and allow visitors to search directly from your website. Add Search Web of Knowledge to your iGoogle page. Add the ISI Web of Knowledge search widget to your widget dock. For more information on this, please refer to the URL below. http://isiwebofknowledge. com/webtools/

Bibliometric terminology

Bibliographic Coupling - A bibliographic coupling analysis connects publications that share items in their reference lists (see figure below), that is, refer to the

....................................................................................................................... BIBLIOMETRIC HANDBOOK

same publications. The assumption behind bibliographic coupling is that publications within the same subject share core material and the more like the publications, the more like the reference lists. One specific trait of this method is that it makes it possible to find conceptual connections between articles that are so new that they havent had the time to be cited yet. An example on how bibliographic coupling can be used is the case where you find a highly relevant article that refers to an important previous article. You may then do a bibliometric search to see what other articles that refers to this previous article to see if there are newer relevant articles on the same subject as the first article. Co-Citation Analysis - A co-citation analysis studies reference-pairs, i.e. papers cited (referred to) by the same publication. Please see figure below. By doing a co-citation analysis you may find older articles that are related to each other, even though they do not refer to each other. This type of analysis will usually generate clusters with highly cited articles since two highly cited papers are more likely to be co-cited in several reference lists than two lowly cited ones. Citation Analysis This is the examination of the frequency and pattern of citations in articles and books. Scientiometrics - This is the measurement and analysis of science. In practice, scientometrics is often done using bibliometrics that is the measurement of (scientific) publications. Scientiometrics is also the title of a journal on the subject. It is published by Springer, indexed in Web of Science and Henry Small is on the editorial board. Self-Citation - Self Citation is the act of citing ones own publications. However, it can also apply to a journal that cites itself or any other type of entity. Self citation itself is not necessarily a bad thing; it is natural for a researcher to cite his or her own prior work as their research evolves. However, the artificial use of self citation for the purposes of inflating citation counts is bad; in general this is a fairly rare occurrence but can be a problem for Journals that are trying to inflate their own Impact Factor. Negative Citation - Researchers cite each other for a variety of reasons but in general citations are positive in nature, for example where a researcher is paying homage to prior research that they used as part of their own work. However, sometimes researchers will cite a work for negative reasons; maybe they disagree with the conclusion or were unable to re-produce the research results. It is not possible to track the nature and context of a citation in Thomson Scientific products. However, when looking at large citation sets such as countries or institutions, the negative citations can be considered statistically irrelevant. Studies have shown that negative citation rates are around 4% and negative citations will drop out of the scope relatively quickly
Journal Citation Reports

Impact Factor The journal Impact Factor is the average number of times articles from the journal published in the past two years have been cited in the JCR year.

In simpler terms, it means An Impact Factor of 1.0 means that, on average, the articles published one or two year ago have been cited one time. An Impact Factor of 2.5 means that, on average, the articles published one or two year ago have been cited two and a half times. For more information on this, please refer to: http://thomsonreuters.com/products_services/science/ free/essays/ http://thomsonreuters.com/products_services/science/ free/essays/using_impact_factor/ Journal Immediacy Index This indicates how quickly articles in a journal are cited. The Immediacy Index is calculated by dividing the number of citations to articles published in a given year by the number of articles published in that year.

In simpler terms, we can say An Immediacy Index of 1.0 means that, on average, the articles in the journal have been cited one time within the same year. Journal Cited Half-Life Helps evaluate the average age range of articles cited from the journal Applications: You use the Journal Cited Half Life to see if articles from a journal that were published a long time ago are still being cited. This shows you if the journal has a good track record and was producing good articles in the past. The 5-Year Impact Factor - The traditional Journal Impact Factor is based on two years of cited journal content cites in the current year to journal material published in the prior two years. The 5-Year Impact Factor is based on cites in the current year to journal material published in the prior five years. This is suitable for the following: Journals where body of citations may not be large enough to make reasonable comparisons Publication schedules may be consistently late Journals that may take longer than two years to disseminate and respond to published works

Journal Citation Reports - Within the Web of Science Thomson Reuters captures bibliographic information and cited references for journal articles. This data can be processed to create a variety of metrics for the evaluation of journals. - Journal statistics provide a systematic and objective way to measure the influence of journals, identify trends, categorize and characterize journals.

....................................................................................................................... Bibliometric Handbook

Eigenfactor Metrics - Like the Impact Factor, the Eigenfactor Score and Article Influence Score use citation data to assess and track the influence of a journal in relation to other journals. Eigenfactor Metrics are available only for JCR years 2007 and later. You can learn more about Eigenfactor Score and Article Influence Score at www.eigenfactor.org. Eigenfactor Score - Eigenfactor score is calculated by the ratio of total number of citations for the JCR year to total number of articles published in the last 5 years. Eigenfactor will include the citations from both Science and Social Science Citation Index as well. But eliminates the self citations which means if the article from the journal has cited some other article which belongs to the same journal will not taken in to account. Say for example, if the journal A has published 150 articles for the last five year and has 50 citations (excluding the self citations) for the JCR 2008. Then the Eigenfactor ratio will be 50:150. Eigenfactor score would be 0.333 Comparing the journal CELL and INT J ROBUST NONLIN, the Journal CELL has better Eigenfactor than the INT J ROBUST NONLIN The box plot (see top screenshot) depicts the distribution of Impact Factors for all journals in the category A, B and C. The horizontal line that forms the top of the box is the 75th percentile as labeled. The horizontal line that forms the bottom is the 25th percentile as labeled. The horizontal line that intersects the box is the median Impact Factor for the category. The cross represents the mean value. The maximum and minimum values are represented by the horizontal lines above and below the box. These values are no more than 1.5 times the span of the inter-quartile range, which is the range of values between the 25th and the 75th percentiles. These lines are commonly referred to as whiskers. An open circle represents an outlier, which is a single value greater or less than the extremes indicated by the whiskers. In this case, the journal does not apply.
Essential Science Indicators

Article Influence Score - Article Influence Score is calculated based on article level whereas Eigenfactor is calculated based on journal level.The formula appears to be same but Article Influence Score will include self citations in to account (as it is a article level calculation). Say for example, if the journal A has published 150 articles for the last five year and has 150 citations (Including the self citations) for the JCR 2008. Then the Article Influence Score will be 1. The average Article Influence Score is 1.00. A score greater than 1.00 indicates that each article in the journal has above-average influence. A score less than 1.00 indicates that each article in the journal has belowaverage influence. Based on the Article Influence Factor, the journal CELL is better than the journal INT J ROBUST NONLIN. Immediacy Index- The average number of times that an article published in a specific year within a specific journal is cited over the course of that same year. This calculation, published in the Journal Citation Reports, is one developed by ISI as an indicator of the speed with which citations to a specific journal appear in the published literature. Such information is useful in determining which journals are publishing in emerging areas of research Impact Factor Box Plot - The Impact Factor box plot depicts the distribution of Impact Factors for all journals in the category. The horizontal line that forms the top of the box is the 75th percentile. The horizontal line that forms the bottom is the 25th percentile. The horizontal line that intersects the box is the median Impact Factor for the category. The cross represents the mean value. In terms of quartile for journal - JOURNAL OF EVOLUTIONARY BIOLOGY, it is in Q1 in Category of Ecology; Q2 in Category of Evolutionary Biology and Genetics & Heredity.

Research Fronts- A research front is a group of highly cited papers, referred to as core papers, in a specialized topic defined by a cluster analysis. The cluster analysis is based on the number of times the papers are cited together, or co-cited. Research fronts offer an alternative classification scheme for highly cited papers since the assignment of papers to a research front is not based on the journal categories used in Essential Science Indicators or on the words appearing in the papers, but purely on citation patterns. This co-citation methodology was developed by Thomson Reuters and is being used an information science tool to identify core sets of articles, authors, or journals of particular fields of study. For more information on this, please refer to: http://garfield.library.upenn.edu/small/smalllist.html http://sciencewatch.com/about/met/rf-methodology/ http://thomsonreuters.com/products_services/science/ free/essays/research_fronts/ Citation Thresholds - The citation thresholds are set to select roughly the same proportion of entities from each field. For scientists the top 1% of names is selected for each of the fields. This percentage of inclusion translates into a specific citation frequency cutoff for each field. For institutions this percentage is also 1%, but for countries and journals 50%. Highly cited papers are selected based on the top 1% by field and by year. Hot papers are in the top 0.1% for each field and bimonthly period. The Citation Thresholds are updated bimonthly and can be found at: http://sciencewatch.com/about/met/thresholds/ Baselines - Baselines are measures of cumulative citations per paper across large groups of papers that provide expected citation rates for groups of papers in a specific field and year. Since citation frequency is highly skewed, with many infrequently cited papers and relatively few highly cited papers, average citation rates should not be interpreted as representing the central

....................................................................................................................... BIBLIOMETRIC HANDBOOK

tendency of the distribution, but rather as guidelines or benchmarks. Similarly, percentiles, or other fixed percentage cuts, indicate the citation rates for specific top segments of the citation distribution. Thus baselines provide comparative averages and percentiles provide comparative proportions. Hot Papers - Hot papers are papers that receive citations soon after publication, relative to other papers of the same field and age. Hot papers are selected based on a current two month citation window looking back only two years. A 0.1% threshold is applied to each field and bimonthly period over the two year span Highly Cited Paper - Generally, citations to papers peak in the second, third, or fourth year after publication, but some papers continue to be cited for many years. A few papers can exhibit delayed recognition. The patterns can vary greatly depending on the type of paper, the field, and the nature of the finding reported. Papers reporting discoveries, for example, can rise quickly and then fall as the discovery is further elaborated in other articles. Papers reporting methods or techniques can gradually increase in citation frequency over several years as the methods diffuse throughout the community and prove their utility. For selection criteria used to determine a highly cited paper, please refer to http://sciencewatch. com/about/met/core-hcp/ Paper Counts - These measure productivity, which are the most basic bibliometric measure and provide the raw data for all citation analysis. Ranking institutions in terms of paper counts helps to compare the productivity and volume of research output among various institutions. The number of researchers at an institution should be taken into account when comparing publication counts across institutions. Characteristics of the papers, such as document type, publication year, and categorization method, should also be considered.
Research Analytics

One can compare an articles citation count to this norm by forming a ratio of actual citations to expected citations. A ratio greater than 1 indicates that the articles citation count is better than average. Citation Frequency Distribution - The citation frequency distribution shows the number of papers cited different numbers of times on a scatter plot with lines. The horizontal axis shows the number of citations on a linear scale. The vertical axis shows the number of articles on a log scale. The Articles line tracks the change in the number of citations in relation to the number of papers. In most cases, the number of citations per paper decreases as the number of papers increases.

The median is the median number of citations received by articles in the dataset. Half the articles in the dataset received fewer than this number of citations, and half received more than this number of citations. Cites per article is the average number of citations per article. It is calculated by dividing the total number of citations received by all articles in the dataset by the total number of articles in the dataset. The h-index is a measure of the average number of citations received by articles in the dataset. Journal Expected Citations - JXC is an abbreviation for journal expected cites Average number of citations to articles of the same article type from the same journal in the same database year. One can compare an articles citation count to this norm by forming a ratio of actual citations to expected citations. A ratio greater than 1 indicates that the articles citation count is better than average. Disciplinarily Index- This is a measure of the concentration of a set of papers over a set of categories. The index ranges from 0 to 1, where the higher the number, the more concentrated the set. For example, an index of .9 indicates a high level of concentration. This index is based on the Herfindahl index, which is commonly used in economics to assess market share. Category Actual/Expected Cites - The category actual/ expected cites is the ratio of total citations received by articles in the dataset divided by the number of total category expected cites for all articles in the dataset. If the number of total citations is the same as the category expected cites, then the ratio equals 1. In cases if the needle points to the right of 1, then the number of total citations is greater than the total number of citations expected for the categories represented in the dataset. If it points to the left, then the number of total citations is smaller than the total number of citations expected for the categories represented in the dataset. This ratio provides a normalized measure of category-level performance for a multidisciplinary set of papers.

InCitesTM - This is a citation-based evaluation tool for academic and government administrators to conduct analyses on their institutional productivity and benchmark their output against peers and aspirational peers in a national or international context Category Expected Citations - CXC is an abbreviation for category expected cites. This is an average number of citations received by articles of the same article type from journals in the same database year and same category. If a journal is assigned to more than one category, the category expected cites is the average for the categories.

....................................................................................................................... Bibliometric Handbook

When an article appears in a journal that is classified into more than one category, the category expected cites value used is an average of the category expected cites for all categories in which the article belongs. Journal Actual/Expected Cites - The journal actual/ expected cites is the ratio of total citations received by articles in the dataset divided by the number of total journal expected cites for all articles in the dataset. If the number of total citations is the same as the total journal expected cites, then the ratio equals 1. If the needle points to the right of 1, then the number of total citations is greater than the number of total citations expected for the journals represented in the dataset. If it points to the left, then the number of total citations is smaller than the total number of citations expected for the journals represented in the dataset. This ratio provides a normalized measure of journal-level performance for a set of papers from multiple journals and categories. Global Comparisons - Global Comparisons is divided into Institutional Comparisons and National Comparisons. These comparisons enable you to evaluate the research performance over time of an institution, country, or territory in the context of selected field of research. Research Performance Profiles - Research Performance Profiles is a collection of reports that provide aggregate metrics generated from data in a subscription dataset. In addition, customized reports based on selected criteria may be generated through the Create a Custom Report feature.

Journal Selection Process

Each year Thomson Reuters evaluates approximately 2,000 journals for possible coverage in Web of Science. Today Web of Science covers over 9,000 international and regional journals and book series in every area of the natural sciences, social sciences, and arts and humanities. For more information, please see the URL http://thomsonreuters.com/products_services/science/ free/essays/journal_selection_process/
Conference Proceedings Selection Process

Conference proceedings are published as part of scholarly journals or in books. Each year Thomson Reuters evaluates fast-moving areas of the physical sciences, particularly engineering and biomedical sciences. For more information, please see the URL http://wokinfo.com/products_tools/multidisciplinary/ webofscience/cpci/cpciessay/

....................................................................................................................... BIBLIOMETRIC HANDBOOK

Science Head Offices Americas Philadelphia +1 800 336 4474 +1 215 386 0100 Europe, Middle East and Africa London +44 20 7433 4000 Asia Pacific Singapore +65 6775 5088 Tokyo +81 3 5218 6500 For a complete office list visit: science.thomsonreuters.com/contact
AG1006000ap Copyright 2010 Thomson Reuters