Can informetrics shape biomedical research A case study

Can informetrics shape biomedical research? A case study of the HIV/AIDS research in sub-Saharan Africa Prof OB Onyancha Department of Information Science College of Human Sciences UNISA 11/7/2013 Onyancha_Inaugural lecture 1

Agenda • • • Introduction – value of research and the process Why research evaluation? Which methods of research evaluation? Informetrics and its related metrics Theoretical basis of informetrics Application of informetrics in biomedical research – HIV/AIDS case study • Can informetrics shape biomedical research? If so, how? • Conclusion 11/7/2013 Onyancha_Inaugural lecture 2

Value of Research There is an enormous amount of literature that has been published on the role or importance of research (e. g. Rantanen, 1999) RESEARCH The role of research in society is (i) to increase our knowledge of nature and society, (ii) to increase our understanding of these 3 components of life, and (iii) to provide a scientific basis for actions to make a better world (Rantanen, 1999) 11/7/2013 Onyancha_Inaugural lecture 3

Research Cycle Missing from most (if not all) models of the research cycle is what I think is the ‘final’ step in the process, namely – research evaluation? Source: Greyson, 2009 11/7/2013 Onyancha_Inaugural lecture 4

Why Research Evaluation? The increasing demand for research evaluation can be attributed to an increased emphasis on governance and accountability in both the public and private sectors. Broad purposes: ü Analysis: for example, what funding is most effective in terms of different outputs and outcomes, including the impact of research? ü Accountability: for example, for those distributing public funds who need to show they are doing the right thing. Likewise, funding organisations need to demonstrate impact to donors. ü Advocacy: for example, how the research benefits society; this would help funders wanting evidence to support their decisions, or advocates seeking evidence for their cause. ü Allocation: for example, to prioritise which projects, people and institutions are given funding. 11/7/2013 Onyancha_Inaugural lecture 5

Which methods? (Source: Geisler, 2001: 39) 11/7/2013 Onyancha_Inaugural lecture 6

Informetrics • Sample of two definitions: – Methodologies that examine “patterns that show up not only in publications but also in many aspects of life, as long as the patterns deal with information” (Diodato, 1994: ix) – “Covers that part of information science dealing with the measurement of information phenomena and the application of mathematical methods to the discipline’s problems, to bibliometrics and parts of information retrieval theory, and perhaps more widely” (Hood & Wilson, 2001: 294) 11/7/2013 Onyancha_Inaugural lecture 7

Informetrics, scientometrics, bibliometrics, webometrics, cybermetrics and altmetrics Bibliographies – largely references Whole Internet, cyberspace Informetrics Bibliometrics Cybermetrics Web presence, visibility and impact – links, pages, documents 11/7/2013 Science of Science Scientometrics Altmetrics Webometrics Onyancha_Inaugural lecture Alternative metrics – views, downloads, web citations, etc 8

Informetric methodologies/approaches • Several methods/approaches/techniques constitute informetrics methods • The methods can be divided into two broad categories – Publications count and analysis – Descriptive – Citation count and analysis – Evaluative • The most recently coined techniques or methods are link count and analysis and altmetrics– follow same approaches as above • In each of these methods, there are several techniques and measurements that can be applied to assess research 11/7/2013 Onyancha_Inaugural lecture 9

Publications count and analysis • • Publications count is the simplest informetric measurement Examples of questions that publications count is designed to answer are: – – – How many publications, citations, books, patents, etc has a particular author, group of authors, institutions and/or countries/geographic regions, produced? How much has been produced on a given topical issue, discipline, country, regional area, etc? How many publications have each been authored by how many authors? How many publications were published in a given source (journal, magazine, etc? ) In how many languages are documents published? How often does a particular word appear in a text? 11/7/2013 Onyancha_Inaugural lecture 10

Citations count and analysis Citation analysis is the examination of the frequency, patterns, and graphs of citations in articles and books. It uses citations in scholarly works to establish links to other works or other researchers. The focus areas of citation studies would include: ü what motivates an author to cite a particular work; ü the relationship between a citing work and the works cited by it; ü works cited long after their publication and works cited while relatively new; ü heavily cited works, infrequently cited works and those that have not been cited at all; ü how citation practices and patterns differ throughout disciplines or families of disciplines; ü how citation practices and patterns can be used in the evaluation of information sources; ü how citation practices and patterns can be used to enhance information retrieval systems. 11/7/2013 Onyancha_Inaugural lecture 11

Theoretical basis for informetrics • Lotka’s law states – “there a few researchers who publish a great deal and many who publish very little or nothing at all” (Ikipaahindi, 1985: 171) • Bradford’s law simply states that: – “If scientific journals are arranged in order of decreasing productivity of articles on a given subject, they may be divided into a nucleus of periodicals more particularly devoted to the subject and several groups or zones containing the same number of articles as the nucleus, when the number of periodicals in the nucleus and succeeding zones will be as 1: k: k 2 where the constant k is known as Bradford’s constant or multiplier (Ungern-Sternberg, 2000) • Zipf’s law is based on – The fact that people tend to use a “small part of their available vocabulary for most communication” (Wallace, 1989) 11/7/2013 Onyancha_Inaugural lecture 12

Application of informetrics in biomedical research 11/7/2013 Onyancha_Inaugural lecture 13

A Comparative Study of the Literature on HIV/AIDS in Kenya and Uganda: A Bibliometric Study We expected more HIV/AIDS research on Kenya, throughout the entire period of study, considering Kenya’s continued increase in HIV infections besides the country’s favourable research environment. The study found that Although Uganda has, for a long time, been unstable politically from the persistent fights—a situation that could have hindered effective research—research activities regarding HIV/AIDS were focused in the country more than they were in Kenya. 11/7/2013 Onyancha_Inaugural lecture 14

Productivity, and Scientific Impact of Sources of HIV/AIDS Research in Eastern and Southern Africa Journals are major players as far as research dissemination is concerned The study concluded that the coverage of HIV/AIDS sources published in Africa in the MEDLINE and ISI databases is minimal Most journals that are gaining in popularity are those that are electronically available; this perhaps explains why some African journals, which are largely available only in print, are not easily internationally visible — thus affecting their impact To achieve wider circulation, visibility and impact, we recommend that such journals be published online 11/7/2013 Onyancha_Inaugural lecture 15

Authorship patterns of the literature on HIV/AIDS in Eastern and Southern Africa: an exposition of the responsible authors, institutions and countries Foreign authorship dominated the scene and that majority of the publications were published in foreign countries South Africa Kenya E&S African countries Rest of Africa Foreign A total of 120 countries (43 African and 77 foreign countries) – including the countries No. of % of Countries Total within the scope of this study authored at 13 15. 29 16 18. 82 56 65. 88 least 1 paper 12 16. 90 13 18. 31 Uganda 11 15. 94 14 20. 29 The most productive foreign country was the USA (2429), followed by England with Tanzania 12 17. 65 17 25. 00 1412 papers. These two countries Zambia 10 17. 24 13 22. 41 recorded a total of 3841 postings which Zimbabwe 13 23. 21 12 21. 43 accounted for 54. 55% of the total foreign Malawi 12 23. 53 8 15. 69 hits (i. e. 7041) 10 Ethiopia 11/7/2013 26. 32 3 7. 89 46 64. 79 TOTAL 85 71 44 63. 77 69 The most productive regional (African) countries were led by South Africa, 39 57. 35 68 Kenya, Uganda, Tanzania, Malawi, 35 60. 34 58 Zambia, Zimbabwe, Ethiopia, 31 55. 36 56 Botswana, and Cameroon 31 60. 78 51 25 Onyancha_Inaugural lecture 65. 79 38 16

Country-wise collaborations in HIV/AIDS research in Kenya and South Africa 75 countries The two countries largely collaborate with foreign countries, with the 56 countries USA emerging as the strongest collaborator Kenya exhibited stronger links with foreign countries while South Africa had much of its papers published through internal collaborations Impact-wise, Kenya’s papers produced a higher average impact (i. e. 7. 32 in single and 20. 18 in co-authored papers) than South Africa’s papers (4. 16 in single and 8. 05 in co-authored papers) Co-authored papers yielded a higher average impact than the single-authored papers 11/7/2013 Onyancha_Inaugural lecture 17

Changing patterns and trends in author co-authorship networks of HIV/AIDS research in Eastern and Southern Africa Author co-authorship Networks, 1981 -2005: 15 2 18 1981 -1985 2001 -2005 11/7/2013 Onyancha_Inaugural lecture 18

Subject content analysis of the HIV/AIDS research in eastern and southern Africa No. Main MESH Terms 19921995 - 199820012004 Year No. of Terms 1994 1997 2000 2004 -2005 2003171242005 No. of Hits 1 1 25524 1 2001 -2003 1983 -1985 Rank 1 HIV Infections Acquired Immunodeficiency 2 1 Acquired Immunodeficiency 2 Syndrome Sarcoma, Kaposi 29 2 22 3 Retroviridae Infections Adolescence 3 3 Adolescence 1986 -1987 Rank 5 Pregnancy Complications, 21 4 Antibodies, Viral 1 Acquired Immunodeficiency 5 Infectious Deltaretrovirus 6 Anti-HIV Agents Syndrome 2 HIV Seropositivity 7 Pregnancy 14 3 Antibodies, Viral 8 Risk Factors 7 4 Adolescence 9 Disease Transmission, 5 HIV Antibodies Vertical 6 HIV 15 3 No. of 15 Hits 1424 13 172 3815 36 10 34 32 30 3 HIV-1 4 4 10 HIV Seropositivity 11/7/2013 6 4 7 Onyancha_Inaugural lecture 2 1998 -2000 1995 -1997 3 1992 -1994 1989 -1991 4 1986 -1988 8 1983 -1985 1980 -1982 6 3 196362 16139 98973 66344 2090 5 453 127 6 5 6 7 14 11 8 11 13 9 7 7 10 2 3 4 9 19

An informetric investigation of the relatedness of opportunistic infections to HIV/AIDS Rank 1 Cytomegalovirus 2 Pneumocystis carinii 3 Toxoplasma 4 Mycobacterium aviumintracellulare 5 Mycobacterium tuberculosis 6 Herpes simplex Year of Publication 11/7/2013 222 213 370 340 HIV/AIDS 511 694 692 943 62 121 230 364 383 260 51 1471 0 0 142 244 465 343 72 1266 0 7 49 149 254 324 84 867 32 ’ 82’ 84 69 ’ 85’ 87 98 ’ 88’ 90 96 ‘ 91’ 93 104 ‘ 94’ 96 159 ‘ 97’ 99 33 ‘ 00’ 03 591 TOTAL Onyancha_Inaugural lecture 939 813 910 545 225 119 3871 3665 20

Is HIV/AIDS in Africa distinct? What can we learn from the literature? 11/7/2013 Onyancha_Inaugural lecture 21

The basis of the study The study stemmed from South Africa’s former president, Thabo Mbeki’s argument that “it is obvious that whatever lessons we have to, and may draw from, the West about the grave issue of HIV/AIDS, a simple superimposition of Western experience on African reality would be absurd and illogical” (as cited in Cohen, 2000) It has been observed that not only do the manifestations of the AIDS disease in Africa differ from those in the West but, as Cohen (2000) observes, AIDS-related diseases, and possibly disease progression itself, differ in the continent (i. e. from region to region) (Cohen, 2000) 11/7/2013 Onyancha_Inaugural lecture 22

Methods used • Content analysis and specifically co-word analysis – Co-Word analysis is a content analysis technique that “reveals patterns and trends in technical discourse by measuring the association strengths of terms representative of relevant publications or other texts produced in a technical field” (Coulter, Monarch & Konda, 1998: 1206) • Variables considered – Opportunistic infections, pre-disposing factors, risk factors, sexually transmitted diseases, and other tropical diseases most commonly associated with HIV/AIDS in Africa 11/7/2013 Onyancha_Inaugural lecture 23

Results 11/7/2013 Onyancha_Inaugural lecture 24

Opportunistic diseases AND HIV/AIDS was associated with 21 opportunistic infections, led by Tuberculosis, followed by Pneumonia, Mycobacterium Avium Complex, Cancer and Kaposi’s sarcoma Opportunistic infections/diseases that exhibited high strengths of association with HIV/AIDS 11/7/2013 Onyancha_Inaugural lecture 25

Pre-disposing factors AND HIV/AIDS The factors that could be influencing the spread of HIV/AIDS in the region include culture, substance or drug abuse, malnutrition, rural-related factors and activities, violence, rape, labor migration, ignorance, broken marriages, war, poverty, inequality, socioeconomic factors, refugees and war. Pre-disposing factors that exhibited high strengths of association with HIV/AIDS and may be playing a big role in the spread of HIV/AIDS in Africa 11/7/2013 Onyancha_Inaugural lecture 26

Risk factors AND HIV/AIDS Most common HIV/AIDS-associated risk factors constitute sexual intercourse, vertical transmission (mother to child during birth), blood transfusions and contaminated needles (intravenous drug use, needle stick injuries) Risk factors that exhibited high strengths of association with HIV/AIDS and may be playing a big role in HIV infections in Africa 11/7/2013 Onyancha_Inaugural lecture 27

Other sexually transmitted diseases AND HIV/AIDS Papillomavirus Infection was the most common in HIV/AIDS titles. There were other high co-occurrence frequencies from genital warts, hepatitis B, syphilis, bacterial vaginosis, and herpes zoster STDs/Is that exhibited high strengths of association with HIV/AIDS and may be increasing the risk of HIV transfer in Africa 11/7/2013 Onyancha_Inaugural lecture 28

Other tropical diseases AND HIV/AIDS Out of the total 24 diseases, slightly over one-half (½) co-occurred with HIV/AIDS The highest frequency of co-occurrence was recorded by tuberculosis (S=0. 17) Other terms that were linked to HIV/AIDS descriptors include cholera (S=0. 01), hepatitis (S=0. 04), hypertension (S=0. 01), leishmaniasis (S=0. 02), malaria (S=0. 08), malnutrition (S=0. 03), meningitis (S=0. 02), polio (S=0. 02), schistomiasis (S=0. 01), sickle cell (S=0. 05), and syphilis (S=0. 01) People living in the poorest countries are the most vulnerable to malaria. In 2010, 90% of all malaria deaths occurred in the WHO African Region, mostly among children under five years of age (WHO, 2013). 11/7/2013 Onyancha_Inaugural lecture 29

Can informetrics shape biomedical research? How? We believe that it can The studies have demonstrated that informetrics can assist in policy-making decisions on such matters as funding but also the findings can lead to the identification of possible research collaborators; predict or forecasting changes; identification of influential journals through which biomedical research can be disseminated as well as obtained; and the identification, recruitment, tenure, and/or promotion of researchers Various authors including this author have observed changes in informetrics structure can be correlated with changes in scientific knowledge and understanding (Onyancha & Ocholla, 2005; Onyancha & Ocholla, 2009; Small & Greenlee, 1989; and Russell & Rousseau, nd). BUT, as Debackere & Glanzel (2004: 261) argue, “bibliometric information may serve highly different objectives simultaneously” 11/7/2013 Onyancha_Inaugural lecture 30

Conclusion We conclude by reiterating Wormell’s (2000: 133) argument thus: Access to information itself today does not signify either competitive advantage or guarantee the feeling of being informed, neither in the research nor the business environment. The sophisticated value of online information provision is not to use the databases only for finding facts and accessing documents, but to tap the unique items of useful information, the nuggets of knowledge and (by synthesis and/or analysis) extract the ‘searched pattern’ in the raw data. 11/7/2013 Onyancha_Inaugural lecture 31

Thanks for your attention 11/7/2013 Onyancha_Inaugural lecture 32