As of 22 April 2009 this website is 'frozen' in time — see the current IFLA websites

This old website and all of its content will stay on as archive – http://archive.ifla.org

66th IFLA Council and General
Conference

Jerusalem, Israel, 13-18 August

Code Number: 054-127-E
Division Number: VII
Professional Group: Library Theory and Research:
Joint Meeting with: -
Meeting Number: 127
Simultaneous Interpretation: No

Research capacity in library and information science in South Africa - an overview

Dennis N. Ocholla

Department of Library and Information Science, University of Zululand
Kwadlangezwa, South Africa

Abstract

Uses bibliometric analyses to provide, in general, an awareness of the overall research potential and capacity in South Africa and, in particular, the potential of institutions in this country. Basically, an on-line list of masters and doctoral theses and dissertations as reflected in the South African Bibliographic Network (SABINET) from 1993-2000 is analyzed in order to determine research capacity and potential. Details regarding theses and dissertations, research themes/subjects, institutions, quantity of the reports, language medium, gender and population groups are analyzed. In the first instance, it is found that a large proportion of graduate research originates largely from LIS departments based at the formerly advantaged white universities and that females, who also publish the preponderance of articles, are the dominant contributors to this output. The University of Natal, the Rand Afrikaans University and the University of Pretoria, respectively, lead in LIS graduate research output. The number of publications arising from the 214 analyzed graduate research projects is found to be minimal while those projects that are the subject of publication are invariably linked to tertiary or university education and training. It is recommended that the universities involved should encourage their graduandi in order to ensure that quality graduate research is published. A large part of the graduate research results is consistent with a similar on-going analysis of the South African Journal of Library and Information Science (which yields the highest number of scholarly/academic articles by South African LIS authors) 1990-1998.

Paper

1. Introduction

This paper aims at reporting on an on-going investigation into research capacity in Library and Information Science in South Africa by using bibliometric analyses. The study objectives are to identify and represent research capacity in terms of: research themes but also with regard to individual research productivity, institutional productivity, size/quantity of publications, location, gender/race of researchers and size of registered research projects. For the purposes of this conference, the results of the first part of the study on the size and nature of graduate research are based on records of theses and dissertations indexed in the SABINET on-line database (as reported by gender, institutional affiliation, language of report, level of graduate research report, population group and research topic). Our definition of library and information science will include: information content, theory, technology, users, processing, centers, systems, services and management.

2. Literature Review

Society is becoming increasingly dependent on research for answers to a myriad of unresolved phenomenological and ontological issues and problems that affect life, sometimes on a daily basis. Both public and corporate organizations are spending enormous amounts of money, time and resources on research and development. Research capacity of an organization largely determines its development, influence, and wealth creation and distribution. When South Africa is positioned in Africa according to number of articles in SCI and SSCI it is singled out to be the largest research producer in Africa (1996). The measurement of research capacity is based on several, often-debatable variables. Three dominant variables, in my view, include the registered, on-going and completed research projects. The other one that is also widely favoured for research productivity analyses is based on publication output. The most commonly used indicators in this case are: journal articles, corporate publications, editorship or research reports, books or chapters in books, published proceeding size of congresses, symposia or conferences, performances and exhibitions (quality control highly debated in scholarly circles), book reviews, and patents. The final category could be the size of industrial property or intellectual property ownership. Publication counts are widely used to measure research productivity by individuals, organizations and even countries.

Bibliometrics is a field that has attracted a great deal of attention among scholars of all disciplines. The existence of an International Society for Scientometrics and Informetrics Congress (organized to foster studies in this unique field for quantitative analysis and measurements of publications) is a manifestation of its popularity. Widely viewed (Broadus 1987; Sengupta (1992), bibliometrics tends to focus on the study and analyses of information resources by country, authors, languages, words, type, research methodologies, articles, obsolescence and distribution of authors in literature. When Alan Prichard (1969) (some of us prefer Otlet) proposed a replacement of 'statistical bibliography' with 'bibliometrics,' he perhaps foresaw that this quantitative method of study held potential for rapid growth. By this time, evidently, 'statistical bibliography' had produced exceptional results, as witnessed by Lotka's law on 'Author Productivity' (1926), Zipf's law on 'The Distribution of Words in Text' in 1933 (Zipf 1935), Bradford's law on 'The Scatter of Literature', and "The Decline in the Use of Materials as They Age' or 'half -life' of literature proposed by Burton and Kebler (1960). The quantity of research and publications based on the bibliometric method as well as its application, as reflected in standard bibliographic databases, suggest that the methods have been recognized and still enjoy significant support for scientific enquiry. For instance, a BUBL search (http://bubl1.lib.strach.ac.UK/) with 'bibliometrics' yields 81 matches found in 7678 documents from 200 current (last ten years) LIS journals. In Africa the bibliometric method has not been widely popularized as would be expected. There is evidence of its application by West African scholars, particularly in Nigeria(Aina 1997 & 1998) and Ghana ( Alemna 1994 & 1996).There is not enough information available to suggest its use for library and information science research in South Africa during the covered period. However, we expect fresh publications from a recent doctoral work completed at the University of Natal (Jacobs 1999).

Boon and Zyl did a recognizable study on research capacity in library and information science in South Africa that covered the period 1979 -1989 (1990) ten years ago. The authors focused on the quantity, relevance and quality of research in South Africa. They identified the disciplines researched frequently as well as those less researched and concurred with similar studies elsewhere that quality of research in "LIS tends to be mediocre rather than excellent or poor'. There are significant commonalties between the current study and Boon's and Zyl's study in terms of the use of database records of completed masters and doctoral research as well as the South African Journal of Library and Information Science for determination of quantitative and qualitative indicators. However, differences in methodology, scope and focus are observed in both instances.

3. Methodology

A database on Union Catalogue of Theses and Dissertations produced by the South African Bibliographic Network (SABINET) on-line (http://www.sabinet.co.za) from 1993 to 2000 was analyzed. This database provides titles of research, authors, level of research work, affiliation and broad subject such as library science, information science, computer science and commerce. Information on narrow subjects such as those indexed in table 6 are not provided. In addition, information on gender, language and population groups are omitted. A Boolean search by library and information science yielded 151 records while that done by information science not library science produced 81 records from 1993 - 2000. All information science dissertations and theses in question were for degrees/qualifications in fields other than library and information science and had a bearing on one of the following fields: computer science, commerce, business management, political science and information systems. Although qualifications reflected the aforementioned fields, they dealt with familiar areas such as information systems, IT, IM, and information needs assessment that also featured in the degrees offered in library and information science. After deleting duplications and irrelevant records, 218 records were analyzed by topic/theme, affiliation, and level of thesis or dissertation, language, gender and population group. A table consisting of relevant rows and columns representing the indicators was designed where the data was first recorded in frequencies using special symbols made for recording purposes. The data items were then quantified and summed up in each case and the results represented systematically and ranked by frequencies, in subsequent tables (see tables 1-7) using a word-processor and excel software programs. The results have been verified by triangulation based on data obtained from the analysis of publications in the South African Journal of Library and Information Science covering 1990-1998.The results obtained from graduate research analysis are largely consistent with those in the journal. Unfortunately details from the journal analysis could not be made available for this paper due to technical constraints.

4.0 Results

The results appearing in this section represent the research capacity based on 218 records in the Union Catalogue of Theses and Dissertations produced by SABINET on completed graduate masters and doctoral research in library and information science from 1993 - 2000 as at 20/5/2000. The records have been analyzed by level of graduate work (e.g. masters and doctoral), by gender (male or female), by population group (emphasis falls on black involvement in research as opposed to other racial groups), by language, by affiliation and, finally, by subject orientation. Table eight collates tables 1-6 for comparative data. A summary of the results appears on tables one to seven.

4.1. Research capacity by levels of theses or dissertation

TABLE ONE: RESEARCH CAPACITY BY LEVELS OF GRADUATE WORK (N=218)

YEAR MASTERS DOCTORAL

1999 12 6

1998 10 5

1997 25 8

1996 31 5

1995 43 7

1994 36 4

1993 25 1

Total 176 38

A large part of graduate research is dominated by masters' work (82%) as opposed to doctoral work (18%). Admission to masters in South Africa normally takes place during the sixth year of university education for LIS graduates and fifth year for non-professional degrees. The duration for masters is a minimum of one year full-time or two years part-time. Doctoral degrees take a minimum of two years full-time and four years part-time after masters or equivalent qualification. Masters work can be offered by both course work and mini-dissertation or by thesis/dissertation only. Doctoral work is still offered by thesis only. While high standards are achieved in both cases, the variations found in masters' work have an influence on the depth and breadth of the research report or dissertation and thesis. There has been a decline in masters and doctoral research from 1998. The causes of the decline are worthy of investigation.

4.2 Research Capacity by Gender.

It is of interest to establish the ratio of male to female in regard to graduate research.

TABLE TWO: RESEARCH CAPACITY BY GENDER (N=218)

YEAR MALE FEMALE

1999 6 12

1998 6 9

1997 10 23

1996 13 23

1995 24 26

1994 17 23

1993 14 12

Total 90 128

Females (58.7%) dominate both masters and doctoral research output. Female research output in library and information science is higher than 80% while less than 5% in information science technology oriented subjects (computer science, business management and commerce). More males (over 80%) conduct work in information science technology oriented disciplines such as IT and information systems and tend to obtain most information oriented qualifications/degrees in computer science, business management, information systems, commerce, education and political science.

4.3 Research capacity by population group

Population group in this case refers to race. The nascent democracy in South Africa is yearning for black empowerment through redress. Affirmative action is one of the means employed to achieve empowerment. Analysis of research capacity by population group is therefore essential. Table three provides a summary in this category.

TABLE THREE: RESEARCH CAPACITY BY POPULATION GROUPS (N=218)

YEAR BLACK WHITE & OTHERS

1999 6 12

1998 8 10

1997 9 24

1996 4 32

1995 6 44

1994 1 39

1993 1 25

Total 34 184

The major population groups in South Africa are categorized in terms of Black, White, Indian and Coloureds. Whereas I intended to provide a breakdown of research capacity by all four-population groups, their identification by names provided in the publication proved to be difficult. However, it was possible to identify African names and this made it possible for analysis of the records by black as opposed to other population groups. Indians and Coloureds sometimes use western names that confuse identity with the whites unless individuals are personally approached in order to establish the population group to which they belong. This may be regarded by some to be an unethical form of approach. The number of blacks involved in graduate research is insignificant, although numbers are slowly increasing (15.6%). There are several career opportunities in other challenging, prestigious, high-income earning fields that were not easily accessible for blacks in previous times. These include fields such as medicine and engineering. Whether new opportunities in these fields have any influence on career options remains debatable. Attainment of final qualifications in these areas may continue to represent a "long walk" for blacks. It is, however, clear that graduate research largely remains a field occupied by whites. A detailed breakdown by all population groups including non-South Africans may reveal a clearer picture for manpower planners.

4.4. Research Capacity by Language<

The South African constitution permits the usage of eleven official languages including Afrikaans, isiNdebele, siSwati, Setswana, isiZulu, English, Sesotho, Sepedi, Tshivenda, Xitsonga and isXhosa. It is interesting to establish whether the eleven languages are used in graduate research and, if not, which languages dominate graduate research and why. Table four provides a summary in this regard

TABLE FOUR:RESEARCH CAPACITY BY LANGUAGE USE (N=218)

YEAR ENGLISH AFRIKAANS

1999 17 1

1998 15 0

1997 29 4

1996 25 11

1995 44 6

1994 26 14

1993 19 7

Total 175 43

Presentation of dissertations and theses in English (80.28%) is increasing as opposed to Afrikaans (19.72%) which is in decline. There is no evidence that other nine remaining languages are used for graduate dissertations and theses. The demise of the African languages in graduate research sets a social agenda to be tackled by those who are keen on the promotion of African languages and the African renaissance. More South Africans seem to write theses and dissertations in English, perhaps because of its popularity with regard to publication and scientific communication.

4.5 Research Capacity by Affiliation
The breadth and depth of doctoral theses and dissertations are normally superior compared to masters studies. This may be indicative of the level and perhaps the quality of research undertaken. For purposes of identification with regard to research partnership and associates as well as affiliations for graduate studies, information appearing in table five on research capacity by affiliation may be extremely important.

TABLE FIVE: GRADUATE RESEARCH CAPACITY BY INSTITUTION AFFILIATION (n=218)

NAME OF INSTITUTION 1999 1998 1997 1996 1995 1994 1993 TOTAL

University of U M U M U M U M U M U M U M U M

Natal-
Pietermaritzburg
(UNP) 5 5 1 6 1 7 1 3 2 13 1 3 - - 11 37

Rands Afrikaans
University
(RAU) 2 1 - - 3 5 2 5 2 5 - 12 - 4 9 32

Pretoria
(UO) - 3 3 1 1 3 1 3 1 3 1 4 1 4 9 20

South Africa
(UNISA) - - 1 1 1 1 - 2 1 1 1 2 - 5 4 12

Cape Town
(UCT) - - - - - 3 1 11 1 2 - 3 - 2 2 21

Potchestroom
University
(PUC) - 1 - - 1 1 - 4 - 2 - 2 - 2 1 12

Witwatersrand
(WITTS) - 1 - 2 1 - - - - 15 - 5 - 6 1 29

Stellenbosch
(Stell) - - - - - 2 - 2 - 1 1 3 - 1 1 9

Zululand
(UZ) - - - - - 2 - - - 1 - - - 1 - 4

Orange
Free State
(UOFS) - - - - - 1 - - - - - 1 - - - 2

Vista
University
(VU) - - - - - - - - - - - 1 - - - 1

Western
Cape
(WCU) - - - - - - - 1 - - - - - - - 1

Total 7 11 5 10 8 25 5 31 7 43 4 36 1 25 38 180

A large part of graduate research in library and information science originates from the University of Natal in Pietermaritzburg 48 (22 %), the Rand Afrikaans University 41 (18.8%) and the University of the Witwatersrand 30 (13.76%) that largely offers information science oriented research in Business Management. Following those three closely are the University of Pretoria 29(13.3%), the University of Cape Town 23 (10.6%) and the University of South Africa 16 (7.4%). Potchefstroom 13 (6. %), Stellenbosch 10 (4.6%) and the remaining universities make an insignificant contribution. In terms of advanced research leading to doctoral work, the leading universities, respectively, are Natal (Pietermaritzburg) (11), RAU (9), Pretoria (8), UNISA (4) and Cape Town (2). Most of the Rand Afrikaans University's doctoral are in information science based qualifications in computer science, commerce and information systems.

It is believed that the success of most of the aforementioned universities is attributable to a strong graduate tradition and to the maintenance of a stable graduate supervisory capacity. Most senior academics are still found in the aforementioned universities, which carry on with the graduate tradition. Universities that specialize in graduate library and information science education such as the University of Natal (Pietermaritzburg) and the University of Cape Town are likely to lead in research and graduate work. Between 1993 and 1999 information science based masters and doctoral research in commerce, computer science, geography and environmental studies, political science, education and music were obtained at the following universities: Pretoria (7-D, 4-M), Rand Afrikaans (4-D, 9-M), Witwatersrand (I-D, 28-M), Cape Town (8-M), Natal (I-M), UNISA (M-4), Potchefstroom (M-2), and the Orange Free State (1-M). The University of Witwatersrand (it does not offer library and information science degrees) has a strong base for information oriented education in business management while the University of Natal (Pietermaritzburg) leads in both graduate masters and doctoral output in Library and Information Science.

4.6. Research capacity by subject orientation and institutional affiliation. It is of particular interest to establish research subject orientation as a whole as well as by institutional affiliation in South Africa in order to determine the gap for manpower development as well as popular research topics for research partnership and graduate enrolment. Table six provides an indexed list of the research subjects and affiliations.

Table 6: Research capacity by subject orientation

SUBJECT FREQUENCY INSTITUTIONAL AFFILIATION

Information Technology 47 WITS,UCT, RAU,PUC, UP, UNISA, UNP

Management/information management 42 UP, UNP, PUC, UZ, RAU, UNISA, PUC, UCT, UOFS,STEL, UWC

Information systems (includes, EIS, GIS, etc) 30 WITS,RAU,UCT,UNISA,UP,UN

User studies/information needs assessment 26 UCT, PUC, UNISA, UNP, UOFS, RAU, VISTA, UZ, STELL

Information dissemination/services 24 UNISA,RAU,UNP,UOFS, VISTA,UCT,PUC

LIS -education and training 19 RAU,UNP,UNISA,PUC,UP

Library and Information theory 13 UNP,UCT,STELL,UP,RAU

Readership/information seeking 13 UNP,UCT,RAU,UCT,STEL,UZ

Information collection development/management 12 RAU,STELL,UNP,UNISA,UCT

Historical librarianship 9 UNISA,STELL,UNP,UCT

Children's literature 8 UCT,PUC,RAU,STEL

Information security management 7 UNISA,UZ,RAU

Community information services 7 UNISA,UNP,UZ,UP

Cataloguing/classification 6 UNP,RAU,UCT,PUC

Marketing information system/publicity & promotion of information services 5 UNP,RAU

Legal /ethical issues 4 UNP,UP,PUC,RAU

Records management/archiving 4 UNP,UZ,RAU,UNISA

Databases/CD -ROM/ bibliographical 3 UP

Searching and retrieval 3 UOFS, RAU

Information literacy 3 UP,RAU,UNP

Information sources 3 UNP,RAU,UP,

Interlending/resource sharing 2 UNP,RAU

Information mapping/information audit 2 RAU

Publishing 2 RAU

Health information systems and services 2 UP

Information brokerage/consultacy 1 UP

Legal information systems/services 1 UP

Economics information 1 RAU

Indexing 1 PUC

The ranked list of subject orientation for graduate research shows fairly diversified distribution that may be divided into three zones with an interval of 20. Zone one constitutes those subjects that appear between 40 to 60 times. The highest distribution is concentrated in information technology and management. The second zone ranges from 20 to 40 and includes information systems, user studies and information needs assessment and information dissemination. Information Technology and Information Systems largely emanate from non-library and information science degree/qualification programmes and departments such as computer science, commerce and business management. However, there is evidence that collaborative research between LIS departments and the aforementioned programs are ongoing. Zone three ranges from 1-20 as ranked on table six. Most of the 32 subjects ranked in this zone are offered for LIS degrees. Information security management, legal/ethical issues, databases; information mapping/audit and economic information are shared between degree qualifications.

Table 7: GRADUATE RESEARCH CAPACITY INDICATORS COMPARED (MASTERS AND DOCTORAL THESES AND DISSERTATION 1993-2000) N=218

YEAR 1999 1998 1997 1996 1995 1994 1993 Total

MASERS 12 10 25 31 43 36 25 182

DOCTORAL 6 5 8 5 7 4 1 36

FEMALE 12 9 23 23 26 23 12 128

MALE 6 6 10 13 24 17 14 90

BLACK 8 5 9 4 6 1 1 34

WHITE
AND
OTHERS 10 10 24 32 44 39 25 184

LANGUAGES
Eng
Afr 17
1 15
0 29
4 25
11 44
6 26
14 19
7 175
43

UNIVERSITY/INSTITUTIONAL AFFILIATION (by number of masters and doctoral research)

UNP 10 7 8 4 15 4 - 48

RAU 3 - 8 7 7 12 4 41

WITS 1 2 1 - 15 5 6 30

UP 3 4 4 4 4 5 5 29

UCT - - 3 12 3 3 2 23

UNISA - 2 2 2 2 3 5 16

PUC 1 - 2 4 2 2 2 13

STEL - - 2 2 1 4 1 10

UZ - - 2 - 1 - 1 4

UOFS - - 1 - - 1 - 2

VISTA - - - - - 1 - 1

UWC - - - 1 - - - 1

5. Conclusions

It is observed that a large proportion of graduate research is produced in LIS departments located in previously advantaged universities and that white females dominate in numbers among those graduating with masters and doctoral degrees. The same gender group publishes the preponderance of articles. The masters and doctoral research output by African South Africans is alarmingly small. Whereas the University of Natal (Pietermaritzburg) leads in both masters and doctoral research output for graduate degree/qualifications in library and information science. Other rankings appear in tables 5 and 7. The publications arising from the 218 analyzed completed research work are minimal and those publications are linked to tertiary institutions or to university education and training institutions. Where there is no pressure to publish for career growth graduates publish less or nothing. Arguably, research is complete only when it is published. The tradition that requires research supervisors or promoters to produce joint publication with their students from the research project or encourage their students to publish should be encouraged. This would attract good research work, where the authors are reluctant to publish because they are not under any pressure or obliged to do so as opposed to university academics that is compelled to publish. Specialization and location of senior academics that determine graduate research levels, topics, output and other staff related variables, favour previously advantaged universities.

Information reflected in this ongoing study should, in my view, provide data for determining graduate research capacity in South Africa; for knowledge and comparison of graduate research potential; selection of graduate institutions by those aspiring for masters and doctoral research; for student exchange; for research partnership and for determination and developments in graduate research. The rapidly growing research capacity in the country offers promising opportunities for research and professional linkages as well as manpower training at graduate levels for the people of Africa. South African LIS institutions have the potential to play a leading role with regard to LIS research in Africa.

The work done by SABINET with regard to the establishment of a database on graduate research in South Africa is commendable. A current data base on ongoing (registered) and completed LIS research for non-degree purposes is essential to supplement the SABINET. I believe that the forthcoming first biannual DISSA conference to be held in Pretoria during October will set a strong agenda in this regard.

6. References

Journal of Information Science

Aina, L. O. (1998), Manuscripts Submitted for Publication in the Information Profession in Africa.Comparative of Rejected and Accepted Papers. Journal of Documentation, Vol.54, Number 2: 250 -251.

Alemna, A.A. (1996) The periodicals Literature of Library and Information in Africa:1990-1995. International Information and Library Review, Vol.28 No 2: 93 -103.

Alemna, A.A. and Badu, E.(1994) The Nature and Trends in Research and Journal Literature in English Speaking Africa. International Information and Library Review, Vol.26, No 1: 19-31.

Boon, J.A.; van Zyl Rhona E., (1990) State of the art of research in Library and Information science in South Africa: a macro analysis. South African Journal of Library and Information Science Vol. 58 No. 2: 155-166.

Bradford, Samuel C., (1934) Sources of Information on Specific Subjects. Engineering 137: 85 -86.

Broadus, Robert N., (1987) Towards a definition of 'Bibliometrics. Scientometrics (12):373-379.

Burton, R.E.; Kebler, R.W., (1960) The 'Half-Life' of Some Scientific and Technical Literature. American Documentation 11: 18 -19.

Jacobs, Daisy (1999) A bibliometric study of the publication patterns of social scientists. Unpublished Ph.D thesis, University of Natal , Pietermaritzburg.

Lotka, Alfred J, (1926) The Frequency Distribution of Scientific Productivity. Journal of the Washington Academy of Science 16: 323.

Pritchard, Allan (1969) Statistical Bibliography or Bibliometrics. Journal of Documentation, 25:348 - 349.

Sengupta, I.N., (1992) Bibliometrics, informetrics and librametrics: an overview. Libri 42:75-98.

South African Science and Technology Indicators 1996 (1996). Pretoria, Directorate for Science and Technology Policy, 134.

Zipf, George K., (1935) The PsychoBiology of Language. Boston, Houghton Mifflin Company.

Professor Dennis Ocholla is Professor and Head of the Department of Library and Information Science at the University of Zululand, South Africa.

Latest Revision: May 15, 2000

66th IFLA Council and General Conference