Julie Sharp

Due Date 5/9/03

Final Assignment

Part 1: Final Assignment

I give permission for my final project to be made available through the LIS Learning Showcase web server.  Have a great summer!

Bibliometric Analysis of Geological Journal using ISI

 

GJ ranks 22 out of 36 in current geological articles.  Compared to 35 other geology journals in JCR, Geological Journal ranks in the top 10 for the impact factor.  GJ ranks 16 out of 36 in total cites.  GJ ranks 19 out of 36 in the immediacy index.  The statistics indicate that certain authors are cited many times in one article, but not many times overall.  There is not an abundance of numbers of documents in Geological Journal written by the same author.  It may be a sign of striving for a diversity of ideas in the journal. The journal is published quarterly, a respectable, productive level of frequency.  There are 11 geology journals that have rankings of >10 in terms of half-life.  GJ is one of these journals.  Based on these factors, GJ is a productive journal.

 

In total cites, Geological Journal ranks 3,468 out of 5,752 journals on ISI WOS.  In terms of articles published, GJ produces many articles.  Judging by the number of references in the articles, one can assume that they are well-documented.  The keywords listed for the articles indicate that the articles are about diverse regions and topics within geology. 

Introduction

The web page for Geological Journal states the following regarding their aims and scope:

In recent years there has been a growth of specialist journals within geological sciences. Nevertheless, there is an important role for a journal of an interdisciplinary kind. Traditionally, GEOLOGICAL JOURNAL has been such a journal and continues in its aim of promoting interest in all branches of the Geological Sciences, through publication of original research papers and review articles.

The Journal has a particular interest in publishing papers on regional case studies, not only from the British Isles but from any global locality, which have conclusions of general interest. Such papers may emphasize aspects across the full spectrum of geological sciences.

GJ publishes papers about sedimentology, palaeontology, geology, geophysics, geochemistry, metamorphic geology and igneous geology.

Simple Productivity Analysis:

Document types:

·        Geological Journal publishes articles, biographical items, discussions, editorial material, items about individuals, letters, notes and reviews.

·        From 1997-2002, 124 articles and 4 editorials were cited.

Authors in GJ

 

The top five most cited authors are in the table:

 

Name of most cited author

Number of documents in Geology Journal written by author

Number of articles in Geology Journal written by author

FYI: Number of volumes in Geology Journal in which author is cited (not the number of hits)

1

Rong, JY

1

1

2

2

Vidal, O

1

1

1

3

Collins, AS

1

1

1

4

Wilson, P

1

1

2

5

Vennin, E

1

1

1

Table 1-1

Searching of ISI indicates that all the writers are prolific and publish frequently, particularly Rong, Jy.  Each of these authors has one article in GJ and is cited at least once. 

Citivity (References Counted or Number of Citations Made)

Geological Journal
Volume 37, Issue 3, 2002.

Number of References

 
Article 1

72

 
Article 2

55

Article 3

51

Article 4

26

Total References in volume

204

Mean (average)

51

Median

53

 

 

Table 1-2

The citivity table indicates that the articles are well-researched.

 

2d. Subjects (i.e. keywords)

 

Geological Journal
Volume 37, Issue 3, 2002.

Author Keywords:

 
KeyWords Plus:

Article 1

Central Anatolia, Nigde, ignimbrites, petrography, geochemistry, mineral chemistry, petrogenesis

 
FRACTIONAL CRYSTALLIZATION, TRACE-ELEMENT, MAGMA GENESIS, ROCKS, PLAGIOCLASE, ARC, IGNIMBRITES, ISLANDS, CONSTRAINTS, SUBDUCTION

Article 2

Tibet, Tethys, Mesozoic, Paleogene, sedimentary facies, paleogeography, tectonic reconstruction, Jinsajiang suture, Bangonghu-Nujiang suture, Shuanghu suture

 
NORTHERN TIBET, SUTURE ZONE, GEOTRAVERSE, EVOLUTION, PLATEAU, LHASA, COLLISION, HIMALAYA, GOLMUD, BLOCK

Article 3

Coralliferous Group, Silurian, sedimentology, sequence stratigraphy, Pembrokeshire

 
WELSH BORDERLAND, SOUTHWEST WALES, EVOLUTION, CANADA, BEDS

Article 4

graptolites, chitinozoans, conodonts, Silurian, USA, Lagerstatten, biostratigraphy

 
CHITINOZOAN, BIOZONATION, SAAREMAA, ESTONIA, WALES

Table 1-3

Volume 37, Issue 3, 2002 of Geological Journals contains no important duplicate keywords .  The lack of important duplicate keywords is indicative of the nature of the journal.  Many of the terms refer to very specific, unique geographic locations.  Authors are writing about different places, reducing the likelihood they would use identical words.  The articles deal with very specialized geological information.  The locations in the articles are diverse with diverse characteristics.  Thus, the terms used are diverse.  This bibliometric analysis suggests that a lack of duplicate keywords is an indication of the unique nature of the articles found within a journal.

Journal Citation reports

As this table indicates, for the year 2001, the impact factor is 1.476 and the immediacy index is 0.136.

Mark

Rank

Abbreviated Journal Title
(linked to full journal information)

ISSN

2001
Total Cites

Impact
Factor

Immediacy
Index

2001
Articles

Cited
Half-life

 1

GEOL J

 0072-1050

 447

 1.476

 0.136

 22

 >10.0

Table 1-4

Step 3: Conclusion

In total cites, Geological Journal ranks 3,468 out of 5,752 journals on ISI WOS.  Many articles in the journal are very specific to certain geographic locations and presumably much of the information is brand new.  This would result in fewer cites. 

In terms of articles published, GJ produces many articles.  Judging by the number of references in the articles, one can assume that they are well-documented.  The keywords listed for the articles indicate the specialized nature of the articles.  ISI keywords plus help make the articles easier to find in searches.

GJ ranks 22 out of 36 in current geological articles.  Compared to 35 other geology journals in JCR, Geological Journal ranks in the top 10 for the impact factor.  GJ ranks 16 out of 36 in total cites.  GJ ranks 19 out of 36 in the immediacy index.  These statistics are very respectable.

The statistics indicate that certain authors are cited many times in one article, but not many times overall.  There is not an abundance of numbers of documents in Geological Journal written by the same author, indicating potential diversity of ideas within the journal. 

The journal is published quarterly, a respectable, productive level of frequency.  There are 11 geology journals that have rankings of >10 in terms of half-life.  GJ is one of these journals.  Based on these factors, GJ is a productive journal.

 

Part 2: Final Assignment:

Short bibliometric analysis for Sara Fabrikant [Blue tables are for Fabrikant, yellow is for GJ.]

Citation Identity, Citation Set and Citation Image

The following tables are created from using ISI for the years 1955-2003.  As expected, the same names appear in Fabrikant’s citation identity, citation image and citation set.

The table indicates Fabrikant’s citation identity, specifically, in this case, the authors listed in ISI as cited by Fabrikant more than twice.

Fabrikant’s Citation Identity (authors listed in ISI as cited by her)

Number of times cited by Fabrikant

BUTTENFIELD BP

7

MARK DM

5

FABRIKANT SI (self-cited)

4

GOLLEDGE RG

4

KUIPERS B

4

FURNAS GW

3

SKUPIN A

3

TUFTE E

3

Table 2-1

Buttenfeld is Fabrikant’s greatest influence according to the table.

The table indicates Fabrikant’s citation set of 15.

Citation Set (number of authors citing Fabrikant).  Each cell designates an individual journal article

Number of Citations
  1. Ancona D;
  2. Freeston M;
  3. Smith T;
  4. Fabrikant S

1
  1. Borner K;
  2. Chen CM;
  3. Boyack KW

2
  1. Cai GR

2
  1. Cartwright W

1
  1. Fabrikant SI;
  2. Buttenfield BP

3
  1. Kwan MP

1
  1. MacEachren AM

1
  1. Skupin A

2
  1. Skupin A

1

 

14 total citations

Table 2-2

Fabrikant has influence in her field and is cited by other authors.

The table indicates Fabrikant’s citation image.

 

Authors citation image, i.e., author’s cocited with Fabrikant, S*

Ancona D;

Freeston M;

Smith T;

Buttenfield BP

 
Table 2-3
Fabrikant writes articles with authors whom she cites and who cite her.

Keywords From Fabrikant  [a brief analysis of GJ’s keywords as subjects or tools follows the analysis for Fabrikant]

The subject of the author keywords and the keywords plus from the geologist Fabrikant’s articles fit the disciplines of Geographic Information Science, Library Science, and Computer Science.  The fact that the terms can be categorized as subjects within all these fields is evidence of the relationships between these fields.  Two major conclusions can be drawn.

The keywords are usually tools in the disciplines, but not in a traditional sense.  They are not tools in the sense that they are an instrument or apparatus used to perform an operation.  They are, however, useful in the practice of a vocation or profession and the “means to an end,” which is another definition of tool.   They are a means to convey information within the three disciplines listed.

Table with keywords and their brief definition indicate the keywords fit the disciplines of Geographic Information Science, Library Science, and Computer Science. 

The terms are also tools of the disciplines in most cases.

Author Keywords

Discipline and Tool: Geographic Information Science

Discipline and Tool: Computer Science

Discipline and Tool: Information Science

Brief Definition of Term

information retrieval

Yes

Yes

Yes

Descriptive term meaning data retrieval

semantic information spaces

Yes

Yes

Yes

Spatialized or graphic database depictions are known as semantic information spaces (Fabrikant)

spatial metaphors

Yes

Yes

Yes

Spatial metaphors are the graphic or visual depiction of something in a graphical user interface.

Spatialization

Yes

Yes

Yes

Graphic or spatial depiction of data (Fabrikant, S.I.)

Visualization

Yes

Yes

Yes

Graphic or spatial depiction of data (Fabrikant, S.I.)

Keywords Plus

Discipline and Tool: Geographic Information Science

Discipline and Tool: Computer Science

Discipline and Tool: Information Science

Brief Definition of Term

VIRTUAL ENVIRONMENTS

Yes, but it is a general term

Yes, but it is a general term

Yes, but it is a general term

Computer or software generated environments

RETRIEVAL

Yes, it is from the discipline, but it is a general term./No, retrieval is too general to really be a tool.

Yes, it is from the discipline, but it is a general term./No, retrieval is too general to really be a tool.

Yes, it is from the discipline, but it is a general term./No, retrieval is too general to really be a tool.

Descriptive term meaning getting

SEMANTIC SPACES

Yes

Yes

Yes

Spatialized database depictions are known as semantic information spaces (Fabrikant

Table 2-4

Keywords from Geological Journal

Volume 37, Issue 3, 2002 of Geological Journals contains no important duplicate keywords .  Table 1-3 is shown again below indicating this.  The lack of important duplicate keywords is indicative of the nature of the journal.  Many of the terms refer to very specific, unique geographic locations.  Authors are writing about different places, reducing the likelihood they would use identical words.  Most importantly, the articles deal with very specialized geological information.  The locations in the articles are diverse with diverse characteristics.  Thus, the terms used are diverse.  This bibliometric analysis suggests that a lack of duplicate keywords is an indication of the unique nature of the articles found within a journal.

 

Geological Journal
Volume 37, Issue 3, 2002.

Author Keywords:

 
KeyWords Plus:

Article 1

Central Anatolia, Nigde, ignimbrites, petrography, geochemistry, mineral chemistry, petrogenesis

 
FRACTIONAL CRYSTALLIZATION, TRACE-ELEMENT, MAGMA GENESIS, ROCKS, PLAGIOCLASE, ARC, IGNIMBRITES, ISLANDS, CONSTRAINTS, SUBDUCTION

Article 2

Tibet, Tethys, Mesozoic, Paleogene, sedimentary facies, paleogeography, tectonic reconstruction, Jinsajiang suture, Bangonghu-Nujiang suture, Shuanghu suture

 
NORTHERN TIBET, SUTURE ZONE, GEOTRAVERSE, EVOLUTION, PLATEAU, LHASA, COLLISION, HIMALAYA, GOLMUD, BLOCK

Article 3

Coralliferous Group, Silurian, sedimentology, sequence stratigraphy, Pembrokeshire

 
WELSH BORDERLAND, SOUTHWEST WALES, EVOLUTION, CANADA, BEDS

Article 4

graptolites, chitinozoans, conodonts, Silurian, USA, Lagerstatten, biostratigraphy

 
CHITINOZOAN, BIOZONATION, SAAREMAA, ESTONIA, WALES

Table 1-3

 

Works Cited

Fabrikant, Sara, I. The Geography of Semantic Information Spaces.  [Online]. An extended abstract.  http://www.giscience.org/GIScience2000/papers/016-Fabrikant.pdf.  Accessed 5/3/03.

Fabrikant, S. I.  “Visualizing Region and Scale in Information Spaces.” [Online]. Proceedings, The 20th International Cartographic Conference, ICC, Beijing, China. August 6-10, 2001.  Accessed 5/3/03.

 

The Usefulness of Bibliometrics

Bibliometric methods are useful in gaining an understanding of patterns of scholarly communication in a chosen discipline.  Good interpretation is required along with mere statistics.  Citation analysis can work in conjunction with an overall knowledge of a discipline to confirm or revise existing thoughts on a scholar’s citation set or the number of authors citing a scholar, which is indicative of their importance in the field.  Also, scholarly patterns within indexed journals can be studied. 

A scholar’s influence in their field can be measured by the number of times other authors cite the scholar.  Biblometric methods allow scholar’s rates of productivity to be measured and compared.  Scholar’s egotism or the number of times a scholar cites oneself also can be measured.  This self citation may have a valid source, however.  Scholars build upon their own work, making self citation a common occurrence.  When measuring a scholar’s influence, their self citation rate should be evaluated. 

When used to gain an understanding of an academic discipline, one must always bear in mind the content of the articles, however.  A scholar could be cited frequently regarding the perceived error of their work.  What Brooks calls “negative citation” could create flawed impressions of the importance of a scholar.  An author that writes many articles may be cited more frequently than others even if the quality of the articles is not exceptional.  A gifted, yet less ambitious scholar who works in isolation, unpublished, will never be cited, but may be more talented.  Hopefully, that is an uncommon scenario. 

Bibliometric methods are an adjunct to a familiarity with the scholarly literature through reading within a chosen field.  In absence of training within a field, bibliometric methods can be used to draw conclusions that must be tempered by the thought that bibliometrics are not a substitute for familiarity with the literature gained through reading and study.  Bibliometric methods allow a generalist to gain an awareness of trends within a given field and draw conclusions that are in many cases correct. 

Howard White points out that ISI indexing leaves out important documents such as a scholar’s books, chapters within books, publications and forms of publication.  This can lead to an ISI–based citation identity that leaves out a scholar’s major works  Multi-authored works, increasingly common, create another inaccuracy for ISI citation identity.  The first listed author receives credit for citations to a multi-authored article.  White also notes that older articles and author names create problems.  He refers to “author homonyms – the same name for different individuals” and “allonyms,” which are his term for “alternate names for the same person;” they can cause citations to be missed.  He concludes that “citation identities are only as good as the editorial decisions that inform them.”  Identities based on ISI data are “sketches subject to revision,” not “definitive portrayals” (White).

            Citation analysis allows productivity and influence to be examined in a quantifiable way.  This lends an additional measure of fairness when scholars are evaluated for raises, tenure or other professional awards.  The great strength of bibliometrics is that it replaces impressions regarding scholarly output with statistics proving scholarly output. 

            Bibliometric analysis allows researchers to analyze a scholar’s citation image or the group of authors with whom one is cited.  This allows inferences to be made about their areas of specialty and opinions.  It can be used to evaluate one’s importance within a field if the scholar frequently coauthors works with respected scholars. 

Citation image is an important bibliometric measure because it charts the development of a scholar over time.  It deals with not only an author’s rate of productivity, but their evolution.  What Koehler terms “idea migration,” changing developments or interests can be tracked.  Trends can be verified. 

Cluster analysis also reveals a great deal in various disciplines.  It reveals the most important journals and scholars.  The laws of bibliometrics reveal something about human behavior, the tendency of scholars to repeatedly use the same sources maybe when others also would work.  Informed interpretation is key to useful bibliometric analysis.

The Future of Webometrics

The web is a system of Internet servers that support specially formatted documents.  Webometrics is defined as:

the use of traditional informetric methods as a starting point for analysis on the web as generally conceivable for any kind of statistical aspects (language, word, phrase frequencies), characteristics of authors, their productivity and the degree of their collaboration, as well as citation analysis for the distribution over authors, institutions, and for the measure of growth of a subject or a database, and concomitant growth of new concepts, definition and measurement of information and types and characteristics of retrieval performance measures (Boudarides). 

The future of webometrics applying bibliometric citation and cluster analysis to the web is a developing science.  Web Impact Factors are a common area of study.  Web Impact Factors or WIFs are the ratio of the number of links to a site, divided by the number of pages at the site (Smith). 

Initially, one might think that web-based citation analysis would have the same or better future as citation analysis of print materials since the Web makes information retrieval easier.  There are factors working against this, however.  The constantly changing face of the web, its size, its continual growth, and the fact that automated methods must be used to create analysis because of the size complicate matters.

In theory, webometrics should work well, yet the variety found within network structures is a complicating factor.  How deep within a file structure a web crawler will search is an issue that confronts webometrics.  As the web quickly changes, with information being added or taken away, the accuracy of webometric analysis is constantly being impacted.  Any webometric analysis is like a snapshot of the changing web. 

The Internet, a network of networks, with many web servers is more like a living organism than print materials, which are published and stay the same.  With print, new material is published, but the information in old sources remains the same, making bibliometric analysis more consistently accurate.  The technology of web crawlers changes frequently, too, which means not only the results, but the tools of webometrics are in flux, creating inconsistencies in the information gathered.

Sometimes, the complex programming problems/issues that only appear intermittently are not obvious except to the programmers and others who have tested the code.  Alistair Smith in his exploration of Australasian Web structures gives evidence of the inconsistencies that can occur with webometrics.  He uses different, yet equivalent Boolean searches that should return identical results, yet the results are different (Smith).  It should be noted that his study was in 1999.  Some of these issues may have been resolved.

One of the limitations that bibliometrics and webometrics share is the limited number of source materials.  Since ISI and other databases limit the sources they index, they miss some citations.  As a practical matter, a researcher attempting a webometric analysis is limited by the amount of sources for a webometric analysis.  The number of sources to search--webpages--is greater, and the need to limit their number of sources is greater, which eliminates many potential sites, skewing data.  Researchers have to work with a smaller portion of the web than would be ideal. 

Finding relevant results, not accurate counters of results are the purpose of search engines.  Thus, it is possible that their performance is more accurate with qualitative rather than quantitative information.  Difficulties arise when counting web links.  Prime notes, “many pages on the same host share the same external linkages.”  Nearly duplicate websites in a different language exist.  This also must be accounted for so that links are not counted too many times.  The webometric analyst must be careful to not overcount.  Also, as anyone knows who uses the web, most search engines will return irrelevant, inaccurate search results, at times.  Conversely, relevant websites may not be found by search engines.

A more recent webometric study by Vaughn Hysen’s in 2002 show that progress is occurring within webometrics despite the challenges.  He examines web impact factors or WIFs.  He states that web inpact factors or WIFs reveal that journals with higher journal impact factor scores do indeed attract more links to their Web sites.  Nagging problems remain, however, in that the particular search engine used for data collection could affect the conclusion of a study (Hysen).

As Prime observes, “web visibility and academic performance” are “completely different.”  The techniques and tools will inevitably improve, but many issues remain regarding the reliability of webometric data.

 

Works Cited

Boudourides, Moses A., Beatrice Sigrist and Philippos D. Alevizos.  Webometrics and the Self-Organization of the European Information Society.  [Online.]  Accessed 5/1/03.  http://hyperion.math.upatras.gr/webometrics/

Koehler, Wallace.  “Information Science as “Little Science”: The implications of a bibliometric analysis of the Journal of the American Society for Information Science.”  Scientometrics. 51(1): 117-132, 2001.

Prime C, E. Bassecoulard E, M. Zitt. “Co-citations and co-sitations: A cautionary view on an analogy.”  Scientometrics 54 (2): 291-308 2002.

Smith, Alistair.  ANZAC webometrics: exploring Australasian Web structures.  [Online.]  Accessed 5/1/03. http://www.csu.edu.au/special/online99/proceedings99/203b.htm

Vaughan L, Hysen K.  “Relationship between links to journal Web sites and impact factors.”  ASLIB Proceedings.  54 (6): 356-361, 2001. 

White, Howard D.  “Authors as citers over time.”  Journal of the American Soceity for information Science and Technology.  52, (2): 87-108, 2002.