Journa of the University Librarians Association of Sri Lanka. Vo. 12, 2008 Library and Information Sciences Research Literature in Sri Lanka: A Bibiometric Study Author Gunasekera, Chamani MLIS (Coombo), ASLLA Senior Assistant Librarian University of Peradeniya Peradeniya, Sri Lanka Contact Detais Tee: 00-94-81-2392482 Emai: chamanig@yahoo.com Abstract This paper presents a bibiometric study of US research iterature emanating from Sri Lanka over 141 artices from the twa LIS journas for the period from 1997 to 2007.Data coected from the two journas namey Journa of University Librarian's Association (JULA) and the Sri Lanka Library Review were anayzed to examine the type of US research, areas of LIS research and research strategies and data coection methods used by the LIS researchers in Sri Lanka. This paper aso
attempts to present an anaysis of the authorship pattern, author productivity and prominent contributors, citation pattern, ength of papers, anguage wise and year wise distribution of artices I growth and direction of US research papers over a period of ten years. The study is quantitative oriented and ongitudina in character. Sri Lanka LIS research output seems to concentrate in the area of 'US Activities" and ' Information Seeking". Some of other important findings are most artices (94%) are singe authored, author productivity is not in agreement with the Lotka's Law. There is no definite growth pattern or positive growth in the area of LIS research in the country. Engish is the most widey used anguage. Average citation per artice was 9 whie average ength of an artices was 12. Key Words: Bibiometrics, Library and Information science, US Literature, LIS Research and Deveopment Introduction Bibiometrics is associated with the quantitative measurement of documentary materias.it provides measurement that are usefu for studying schoary communication. Traditionay bi biometrc studies are used to measure the usage of materias and services within a ibrary or to anayze the historica deveopment of a specific body of iterature. History of Bibiometric studies The history of bibiometrics go back more than a century yea.. The statistica anaysis of scientific iterature began more than 50 years before
the term "bibiometrics" even was coined in 1969. The emerging fied was described in terms such as "Statistica Bibiography". (Parne,2008) The word "Bibiometrics" has been derived from Greek words, Bibio means Book and Metrikos meaning Measurement, referring to the science of measurement reating to Books.(Dhiman,2000) Aan Pritchard first introduced the term bibiometric in 1969 to denote the "the appication of mathematica methods to books and other media of communicationf', According to him definition and purpose of bibiometric is to shed ight on the process of written communication and of the nature and course of a discipine (in so far as this is dispayed through written communication) by means of counting and anayzing various facets of written communication. Quantitative anaysis can measure the growth, scattering of artices in different journas or to measure the obsoescence of iterature in different discipine. (Prtitchard f 1969) According to British standards Institution, it is the study of the use of documents and pattern of pubication in which mathematica and statistica methods have been appied (British standards institution, 1976) More recenty Sengupta had defined this term as the "organization, cassification and quantitative evauation of pubication pattern of a macro and micro communications aong their authorships by mathematica and statistica cacuus. (Senguptha,1990)
Fairthome defines bibiometrics as the ' Quantitative treatment of properties of recorded discourse and behavior appertaining to itff( Fairthome,969) According to Potter, it is the study and measurement of the pubication pattern of a forms of written communication and their authorship" (Potter,981) Many definitions covering various aspeds of bibiometrics are avaiabe, but the common feature noted in a these definitions is the emphasis on quantitative aspect. Bibiometric is a major sub-discipine of quantitative research. This too traditionay used by the ibrary and information science professionas for studying the communication process, information fows, and the others for better understanding and effective management and dissemination of information. (Rajendran, 2007) It has extensive appication in the fied of LIS particuary with regard to studying the trends in a subject. It heps in formuating need based deveopment poicy and provides objective data to inform managers to take timey decisions. (Jena, 2006) The European information science journas, bibiometrics investigations began to popuar in the 1970s and 1980s.Hungaryf eastern Germany and Switzerand were the countries which started to do research in bi biometrics since the beginning of the research era Athough famous Bradford's aw (1934) of scattering,lotkafs Law (1926) of scientific productivity are regarded as miestones in bibiometrics, but bibiometric research actuay started in ate sixties. Later in the seventies and eighties, bibiometrics research took a distinct shape and emerged as a prominent discipine. With the advent of information and communication technoogy (ICT), web technoogy and avaiabiity of different databases onine, the fied of bibiometrics gain a momentum.
Bibiometric studies an LXS research Literature The iterature of ibrary and information science as a discipine has been extensivey researched from a variety of perspectives. The systematic study of subject trends in L IS iterature has been justified as a sef - refexive exercise to determine the historica roots of present ibrary practice (AtkinsJ988) and to predict research trends in future. Wersing and Neveing (1976) presented one eary refexive anaysis of LIS.Their normative view of US has been contrasted with empirica anaysis where subject cassifications, and methodoogica approaches were derived from an examination of iterature sampes.(cano,999) Various contact anayses of LIS research pubications have been made by many schoars around the word. These content anaysis of US research pubications can be roughy divided into two types in severa ways: One method distinguishes between anayses, anayzed sampes pubished in the short term (generay in one year) and in the ong term: these anayses can aso be divided at the nationa and internationa eves: or, in the comprehensive fieds of US and in the imited fieds of LIS. Athough these studies have some kind of deficiencies, they a have contributed to the deveopment of this research subject.(cheng,1996) For this study mainy used as a base of the research artices of two we known schoars in this fetid Jarvein and Vakkari in 1990.In genera, the content anaysis mode of Jarvein and Vakkari consists of three parts: 1. The distribution of the artices over the topic 2. The approaches: viewpoints on information dissemination and socia eves; and,
3. The methods: research strategies and data coection methods (Cheng,1996) It is difficut to identiw the categories of the authors of LIS artices in Sri Lanka. Therefore ony two parts of Jarvein's and vakkari mode use for data anayzing in this study. 1. The distribution of the artices over the topic 2. The methods: research strategies and data coection methods At present, it is difficut to derive an adequate understanding of the deveopment of particuar US subject areas, their theoretica foundations, and preferred methodoogica approaches because resuts reported in the iterature are not cornparabe.(jarvein and Vakkari, 1990) Though bibiometrics has quite ong history in the word very few studies on bibiometrics in US research iterature found in Sri Lanka. Objectives of the study The aim of this study is to study the growth and reated aspects of US research iterature for the period from 1997 to 2007 in Sri Lanka.The objectives of the study are: 1. to examine the characteristics of US research, research strategy used and data coection method 2. to assess the growth of iterature (22)
3. to study the Authorship pattern, authorship productivity, 4. to study citation pattern and sef citations 5. to find ength of papers 6. to find most productive authors 7. to study author's affiiation Methodoogy and data coection Criteria for journa seection In order to examine the characteristics and trend of the LIS research in Sri Lanka two reputed journas have been seected and the criteria for seection these two journas was reguarization of pubication which was necessary for ongitudina data anaysis. This study focused the artices pubished in the Journa of University ibrarian's Association (JULA) pubish by the University ibrarian's Association in Sri Lanka and the Sri Lanka Library Review pubish by Sri Lanka Library Association. The database of the study comprised 141 artices pubished in these two journas from 1997 to 2007.In the study each individua artice was scanned, checked, examined and tabuated for necessary data in to separate sheets in terms of subject of artice, research strategy and date coection method used,names of authors, number of authorship, author's institutiona affiiation, number of references, author's sef citation, ength of artice, year of artice etc. Cassification Scheme The cassification scheme for topics of US by Jarvein and Vakkari (1990) was constructed of US and it's swbfied (Appendix 1) was adapted in
order to examine the nature of US research in the country. The data were coected and cassified using the cassification scheme devised by them. The main casses of LIS topics as foows: 1.The Profession 2.Library History 3.Pubishing and book history 4.Education in LIS 5. methodoogy /Anaysis of US 6. Library and information service activities 7. Information storage and retrieva 8. Information seeking 9. Scientific and professiona communication 10.Other LIS topic (other aspect of LIS) 11. Other Study ( other discipines Since the focus of this paper is not anayzing the contents of the artices, the author used ony the cassification headings mentioned above. The cassification scheme divides LIS topics into 11 major casses with their respective subcasses.(appendix I) Each artice was cassified under ony one main subcass. When an artice deat with more than one topic, ony its main topic is considered. In addition to the cassification scheme,jarvein and Vakkari aso presented a cassification of research strategies. M 10. Empirica research strategy M 20. Conceptua research strategy M 30. Mathematica /ogica method M 40. System/software anaysis design
M 50 Literature Review M 60 Discussion paper M 70 Bibiographic method M 80 Other method M 90 No method/not appicabe Each methodoogica category was subdivided into more specific method, for exampe, the empirica research strategy subdivide into 11 more specific methodoogies such as Historica method,survey Quaitative method, Evauation method etc.(appendix 11) method, The cassification of research methods scheme is further deveoped as the scheme of data coection methods. In this scheme 3arvein and Vakkari identified 10 data coection methods such as: C. Questionnaire, interview C2. Observation C3. Thinking Aoud C4. Content anaysis CS. Citation anaysis C6. Historica source anaysis C7. Severa method of coecting C8. Use of data co ected earier C9. Other method of cot ecting C10. Not appicabe Each 'artice pubished in the two LIS journas chosen was read and cassified according to theme, methodoogy empoyed and data coection method used and noted down according to the year of pubication.
Authorship as productivity variabe The foowing variabes were identified per artices.. Name of author 2. Number of author per artice 3. Author's affiiation 4. Author's se If-citations 5. Most productive authors in the fied Other variabes In order to presenting comprehensive overview of US research iterature foowing variabes were aso identified per artice and noted down according to the year. 1. Literature growth (Year wise distribution of paper) 2. Citation pattern (No.of citations per artice and sef citations of author) 3. Length of artice (Pages of artice) 4. Language of artice Resuts and discussion Growth of LIS iterature Tabe 1 depicts the growth of research iterature of US.
Tabe : Year wise distribution of papers Year No. of artices Cumuative of % of artices artices 1997 9 9 6 1998 20 29 14 1999 10 39 7 2000 15 54 I1 200 7 61 5 2002 5 76 2003 5 91 2004 13 104 9 2005 S 12 116 8,5 2006 12 128 8.5 2007 13 141 9 Tabe I does not indicate any definite growth pattern of the iterature. During the ten year period the high decine of iterature has been observed in year 1997 and 2001. The reason may be ony one journa pubished (Library review) in these years. The highest number of artices (20) was pubished in year 1998 and there after a gradua decine is seen. However in ast four years the growth of iterature static with 12/13 artices per year which appear to be insignificant in terms of word US research iterature. Language of pubication It is aso important to note the anguage of pubication. As usua Engish is the predominant anguage of pubication. (Tabe 2)
Year 1997 1998 1999 2000 200 2002 2003 2004 2005 2006 2007 Tota Tabe 2: Language wise distribution Language Engish Sinhaese Tami Tota 7 16 9 14 6 12 12 I1 8 11 11 117 3 2 2 3 2 18 1 6 9 20 I0 5 7 15 5 13 12 12 13 141 h - Tabe 2 indicates that out of 141 artices, Engish occupies the first position with 117 artices.(83 %) Distribution of research artices over topics Tabe 3 presents the distribution of artices over the topics.
The Tabe 3 shows that the highest number of artices covered within the period of this study is under the heading of Library and information service activities with 44 artices (31 /0). Information seeking is the second most popuar topic with 26 artices whie other aspect of US is in the third with 24 artices. he anaysis reveas that the information retrieva is in fourth pace with 9 artices. Further tabe 3 indicates that there is no definite growth pattern for any topic covered by the artices. Research strategies used in the artices The distribution of the research strategies in the artices from 1997 to 2007 is shown in the tabe 4.
- The most popuar research strategies used by US researchers were conceptua research method and the discussion paper which was in equa number ( 52) of artices. Tabe 4 further shows that a continuous increase in the use of the survey method from the year 2003. Mathernatica/Logica method and content or protoco anaysis method are ess ikey to be used in research iterature in Sri Lanka. Data Coection method used for LIS research Data coections methods used for the research strategies are given in tabe 5.
Tabe 5 shows that the Thinking aoud is the most frequent data coection method was used for LIS research in Sri Lanka. The Questionnaire, interview and the Severa method of coecting were aso used for the US research. Surprisingy, 23% of studies (33) were not appied any data coection method for the study. (Not appicabe) Authorship pattern and author productivity Tabe 6 shows the authorship pattern of the artices. Tabe 6: Authorship pattern Year /Singe Joint Three Four Tota 1997 9 9 1998 1999 2000 2001 2002 20 I0 13 7 15 2007 8 3 Tota 131 7 2 2 20 0 5, 7 15 15 13 1 12 12 13 14 1 Tabe 6 reveas that 131 artices (93%) out of 141 are of the singe authorship. Seven artices (S0/u) were written by two authors and 2 artices (1.3%) by three authors. Ony one artice was written by mutipe authors (4 authors) which is 0.7% of the tota contribution.
Author Productivity Lotka's Law has been used to measure the productivity of authors. It states that ' the number of authors making 'if' contributions is about /n2 of those making one; and the proportion of a contributors, that make a singe contribution, is about 60 percent." This means that in a given subject out of a authors, about 60 percent wi have just one pubication and 15 percent wi have two pubications, 7 percent of authors wi have three pubications and so on. According to Lotkas's Law of scientific productivity, ony six percent of the authors in a fied wi produce more than 10 artices.tabe 7 presents author productivity. Tabe 7: Author Productivity I Se. no No, of artices No. of authors Observed (O/o) 1 45 63 2 2 11 15 3 3 7 10 4 4 3 4 5 5 2 3 6 7 6 9 2 3-8 10 -- 1 A tota of 72 authors contributed 141 artices over the period of study. Of these authors, 45 contributed one artice each, 11 contributed two artices each, 7 contributed 3 artices each and 3 contributed four artices each. Two sets of two different authors contributed 5 and 6 artices each. Lasty two sets of different authors contributed 9 and 10 artices each.
Tabe 7 reveas.that the percentage of authors contributed ony one artice each was 63O/0 (45) which is arger than the origina 6O0/o Lotka's data. On average 1.98 (Approx.2) artices pubished by each author. The Largest number of artices were contributed by an author is 10. As can be seen 15% of authors contributed to two artices which exacty match with the Lotka's Law. But in genera the resuts do not match with Lotka's Law. Ranking of most productive author Tabe 8 presents the most productive authors in the US research area. Tabe 8: Rank ist of productive authors in the fied I Name Of Author 1 No.of contributions Rank Ruwan Gamage 5 4 Pradeepa Wijetunga I Sumana Jayasuriya 1 4 I S I I Sriyani Peram 4 I Geetha Yapa 7 authors I1 authors 45 authors I S 41 3 6 2 7 1 8
Tabe 8 shows that the most productive authors are N.U.Yapa who contributed 10 artices and Sriyani Ieperuma with 9 artices. Foowed by W.R.G.Siva and M.B.M.Farooz both with six artices each. Padeepa Wijetunga and Ruwan Gamage are ranked fourth with 5 artices each. Foowed by Sumana Jayasuriya,Sriyani Perera and Geetha Yapa with 4 artices each. Seven authors who contributed 3 artices each have taken the sixth pace of ranking. Eeven authors with 2 artices and asty 45 authors with 1 artice each. It is noted that the most of the top six rank authors were affiiated with higher education institutions. Author's Institutiona affiiation Institutiona affiiation of authors is divided into three categories namey higher educationa institutions, government institutions and private organizations. Tabe 9 presents the distribution of authors by institutiona affiiation. Tabe 9: Institutiona Affiiation Sequence.no 01 02 03 04 Institution Higher Education institutions (Universities) Government institutions Private organizations Not mentioned (Not affiiated) Tota No. of contribution 93 33 6 9 14 -
Above tabe reveas that out of 141 artices, the majority are written by authors affiiated to the universities with a tota of 93 (66%) artices. Foowed by government organizations 33 artices (24O/0) and ony 6 artices (4%) by the private organizations. Institutiona affiiation of 9 authors (6%0) coud not be ascertained. Citation pattern The distribution of citation pattern is given in tabe 10. Tabe 10: Citation pattern Year No. of Cumuative No. of Cumuativ Average Cumuat artices tota citations e tota of citation ive of artices citations per artice average of citation 5 1997 1998 1999 2000 2001 2002 2003 2004 2005 9 20 10 15 7 15 15 13 12 9 29 39 54 61 76 91 104 116 25 207 134 77 33 99 156 114 139 25 232 366 443 476 575 731 845 984 ; 2.78 10.35 13.4 5.13 4.71 6.6 10.4 8.78 11.58 2.78 8 9.38 8.2 7.8 7.57 8.03 8.13 8.48 2006 12 128 116 -p--. 1100 9.67 8.59 2007 13 141 129 1229 9.92 8.72
The tabe 10 presents the year wise distribution of citations in the artices and the average of citations thereof, tota number of citations and average citations per artice. Tabe 10 reveas that tota of 1229 citations distributed among 141 artices. It is aso seen that the number of artices are increasing but not in a uniform manner. Simiary, the number of citations per year is varying from year to year. The highest number of citations per artice is in the year 1999 and the owest number of citations per artice is in the year 1997. Author sef- citation The frequency of author sef citations in the references of the artices are identified in tabe. Tabe 11: Author sef citation Sef citation ' Frequency Percentage 2 Yes I1 8 No 130 92 Tota I 141 100 Out of 141 of tota artices ony 11 (8%) contained author's sef citations. This indicated that some of the contributors of the two journas under study are quite productive and are continuousy working towards contributing more artices to LIS journas. Length of artices Tabe 12 contains distribution of artices according to their ength.
Tabe 12: Distribution of papers according to ength h Year No. of Cumuative No. Cumuative Average Cumuative. artices tota of of tota of pages average artices pages pages Per no. of artices pages 1997 9 9 61 61 6.78 6.78 1998 20 29 226 287 1,3 9.9 1999 0 39 114 401.4 10.28 2000 15 54 181 582 12.07 10.78 2001 7 61 76 658 10.86 10.79 2002 15 76 22 1 879 14.73 11.57 2003 2004 13 13 91 104 177 149 1 1056 1205 13.62 11.46 11.6 11.59 2005 12 116 150 1355 12.5 11.68 Tabe 12 reveas that the average ength of artices is 12.28 pages. Further it has been observed that the average ength of artices varied from a minimum of 6.78 pages to a maximum of 16.08 pages. The foowing concusions are drawn from the study. 1. It is seen that there is no definite growth pattern or positive growth of LIS research artices pubished from 1997 to 2007.
2. The subject wise anaysis of the contents of the artices reveaed that Library and Information services activities (LIS) was prominenty studied (3I0h) in the LIS fied.. LIS research iterature in Sri Lanka seems to concentrate mosty in the area of US activities, Information seeking, other aspects of LE and Information storage & retrieva. In combination they covered over 81% of research artices. In genera the findings of the study indicate that the US research in the country has not changed greaty during the ten years period. 3. The research strategies most frequenty used by LIS researchers were conceptua research and the discussion papers whereas the east frequent were empirica, mathematica and historica methods. These resuts indicate that the seection of research strategies by LIS researchers in Sri Lanka is more towards the survey types. 4. The most frequent data coection method used for the LIS research in Sri Lanka was "Thinking Aoud". 5. Singe authorship is prominent in the LIS fied. 6. It was found that the author productivity is not match with the Lotka's Law. The most proific author is in the fied was N.U.Yapa who contributed 10 artices. Majority of top ranked authors are affiiated to the Universities. 7. The average citations per artice are 9 (8.72).
8. The artices have an average of 12 (12.28) pages which shows the reasonabe ength of a research artice. References Atkins,S.E.(1998) Subject Trends in Library and Information Science Research 1975-1984.Libraty T / 36 P 633-658 British Standards Institution (1976) British standards G/msaty of Documentation Terms. Prepared under the Directives of the Documenta tion Standard Committee p.7 Cano,V.(1999) Bi biomet tic Overview of Library and Information Science research in Spain. Journa/ of the American suciev for InfomaYon science I 50(8) 675-680 D himan,a.k.(2000) Ethno botany journa: A ten year Bi biometric study. Iasfic Buetin, 45 (4) 177-182 Fairthome, R.A. (1969), Empirica by paraboic distribution (Bradford, ZipF-Maud/bart) for bibiometric description and prediction. Journa of Don/mentatiion, 125(1) 3 19 Huanwen, Cheng (1996) A bibiomettrc study of Libraty and Infirmattion #search in China: proceedings of the 6Pd IFU Genera conference hed at the Beifing, China Jarvein, K and Vakkari, P.(1990) Content anaysis or research artices in ibrary and information science. Libmy and Information science, 12 p 395-42 1
Jena, Kama Lochan (2006) A bibiomehic anaysis of the Journa " Indian Journa of Fibre and Textie research. Annas of Libraw and 1nformat.n studies, 53(1) 22-30 Patra, Swapan Kumar et a1.(2006) Bibiometric study of Literature on Bibiometrics. DESIDOCBu//eun of Information TTedK/qy, 26(1) 27-32 Parne, Staffan (2008) Bibiometrics ; Background. Swedish University of Agricuture Sciences Libraries.Avaiabe from: http://www.bib.su.se/bibiometri/ebakqrund.htm (Accessed on 10.06.2008) Potter, W.G. (1981) Lotka's Law Revisited. fibray Trend, 30(1) 31-35 Pritchard,A.(1969) Documentanm, 25(4) 348-349 Statistica Bibiography or Bibiometrics? Journa/ of Rajendran, P. and Parihar, Y.S. (2007) A bibiometric study of Laser iterature in India 1995-2005. Annas of ibrayandnformation Studis 54(2) 112-118 Sengupta, I.N. (1990) Bibfiometrics and its appficatcin. New Dehi, Atantic Pubishers 165-191 Wersig, G. and Neveing. U.(1976)The Phenomenon of interest to information science. Te Informafion scientist, 9 p 127-140
APPENDIX THE CLASSIFICATION SCHEME Professions Library history Pubishing and book history Education in LIS Methodoogy/Anaysis of US Library and Information service activities Circuation or interibrary oans o Coections o Inf. or ref. service o User education CI o Buidings or faciities Administration of panning o Automation (except when concerned with some particuar activity) CI o o Other L& service activities Other L& service activities Severa interconnected L&I activities Information storage and Retrieva Cataoguing o Cassification and indexing (process or anguages) CI CI Information retrieva Bibiographic databases or bibiographies o Non bibiographic data bases (textua, numeric...)
information seeking Information dissemination o The use/users of information channes/sources o The use of L&I services (no other channes considered) o o o Information seeking be havor (focus on persons) Information use (whether (and how) used) Information management Scientific and professiona communication Scientific or professiona pubishing o Citation patterns and structures o Other aspects of communication Other US Topic Other study (other discipine) APPENDIX I1 M ETHODS - RESEARCH STRATEGY Empirica research strategy Historica method Survey method Quaitative method Evauation method Case or action research method Content or protoco anaysis Citation anaysis Other bibiometric method Secondary anaysis
Experiment Other empirica method Conceptua research strategy Verba argumentation, criticism Concept anaysis Mathematica or ogica method Systern/soffware ana ysis/design Literature review Discussion paper Bibiographic method. Other method Not appicabe, no method