The work of the Bibliometrics Research Group (City University) and associates

Citation

Lewison, G. (2005), "The work of the Bibliometrics Research Group (City University) and associates", Aslib Proceedings, Vol. 57 No. 3. https://doi.org/10.1108/ap.2005.27657caa.001

The work of the Bibliometrics Research Group (City University) and associates

This special issue of Aslib Proceedings: New Information Perspectives brings together six papers on bibliometrics, or the quantitative study of publications, that have been written by people who have been working with the Bibliometrics Research Group (BRG) at City University during the years 2001-2004. The BRG was set up by the Wellcome Trust in order to allow the research outputs database (ROD), which had been developed within the Policy Unit there (formerly the unit for Policy Research In Science and Medicine (PRISM)), to operate more freely and attract new clients to bibliometric methods. During the three years of the contract, the BRG not only maintained the ROD and provided consultancy services based on it to the trust and its UK members, but also developed a separate bibliometrics consultancy for foreign clients and carried out innovative research.

The BRG also welcomed a number of guests who spent time working with the group on their own projects. Some used the ROD, which carried the financial acknowledgements to the half million or so UK papers whose details had been downloaded from the science citation index (SCI) and the social sciences citation index (SSCI). Others used the bibliometric techniques and macros that had been developed by the group to apply to the SCI records. These included new methods of classifying the papers – by subject, by potential citation impact, and by research level. Taken together, these three systems allow the huge number of scientific papers – now well over 800,000 are recorded in the SCI each year – to be put into categories that are defensible and transparent. The focus has been primarily on biomedical research and clinical medicine, although some work has been done in other areas. The new methods allow subject-based analysis to take place in ways that have not previously been possible and have led to many reports for clients and publications in journals.

Bibliometrics is now very much an international activity and the biennial conferences of the International Society for Scientometrics and Informetrics (ISSI) bring together researchers from over 40 countries. It is not surprising, therefore, that the associates of the BRG included people from three other countries: Brazil, Colombia and India, and the results of their work with the group are presented in this special issue.

The group’s first visitor, and the one who stayed the longest, was Adriana Roa-Celis, now Adriana Atkinson, a young Colombian who was working on her PhD dissertation for submission to the State University of Campinas in Brazil under the supervision of Professor Léa Velho. This is one of the top three universities in that country. It is a superb monument to the vision and organisational skills of Professor Zeferino Vaz who created and preserved it on an autonomous basis back in 1962 when the military junta was making independent academic activity very difficult in Brazil. Dr Atkinson has been comparing the immunology research outputs from Colombia and Brazil; despite Colombia’s much smaller size, its research is clearly of at least as high a quality as that of Brazil and is more international.

Dr Jacqueline Leta, a graduate of the Federal University of Rio de Janeiro (UFRJ), although still young has an established reputation as one of the leading bibliometricians of Brazil and we worked together both in São Paulo and in London in 2002 when she paid the Group a visit. Dr Leta, who previously worked at the University of São Paulo, has now had her expertise officially recognised by her appointment to a faculty position back at UFRJ. During the three weeks that she was with the group, we used her data and our techniques of subject definition to look at the roles of Brazilian men and women in three fields – astronomy, immunology and oceanography. Her paper in this issue continues this work and examines the formal research qualifications and the publishing habits of these researchers.

During 2001, I also had the pleasure of making my second visit to India, and went back to Delhi for a conference at the National Institute for Science, Technology and Development Studies (NISTADS). Bibliometrics is very actively pursued in India, and most ISSI conferences have a large number of Indian papers and posters. Quite a lot of these concern Indian research outputs and whether India is keeping up with its even larger and more populous northern neighbour, China – mostly it is not. Dr Aparna Basu, whom I had met at NISTADS the previous year, trained as a plasma physicist and was keen to use her specialist knowledge to define a filter for astronomy and astrophysics using similar methods to those that we had been using in biomedicine. This filter would be based on both specialist journals and the title words of the papers; the latter are important as they bring in about one third more papers including the influential ones in Nature and Science. She presented our preliminary findings at the ISSI conference in Sydney, Australia, later that year, and we subsequently worked together by e-mail as she moved down to Bangalore and then back again to her home in Delhi to improve our coverage of astronomy to include ten years of data.

The other three papers in this special issue draw directly upon the ROD acknowledgements data to identify and then analyse particular groups of papers. Dr Dwijen Rangnekar, who was previously at the School of Public Policy at University College, London, was interested to explore the role of “patient groups”, or disease-specific collecting charities, in the support of research on their diseases. He worked closely with two of them, the Multiple Sclerosis Society and the Parkinson’s Disease Society, both of whom were ROD members, to identify the papers that acknowledged their financial support and the grants that had given rise to them. His paper in this issue covers just the first of these studies and shows that a dedicated charity can make a big difference to the research activity in a closely defined area. It can also ensure that the needs of the present generation of patients are not neglected in the search for a cure in the long term. We have, however, found that most such disease-specific charities do support quite a lot of basic research as well; typically one third of their research portfolio does not pass through the filter used to define research relevant to the disease.

Kartik Kumaramangalam is a PhD student at the London School of Economics and Political Science who is examining the British biotech industry and the factors that lead firms to success or failure. There have been some notable examples of both among the more than 100 firms whose support for research, either intramural or extramural, has been recorded in the ROD, which spans the 14 years 1988-2001. Could the quality of this research, and specifically, the amount of collaboration with academia, be a useful indicator to the firms’ prospects? He has looked at the almost 3,000 such papers in the ROD and tried to tease out the factors that lead them to be published in high impact journals. In some later work he has looked at the extent to which big pharmaceutical companies acknowledge the utility of such work through their citation of it; this may be a rather better indicator of whether the firms will form the alliances that are necessary for their success.

Finally, this issue turns its attention to the National Health Service (NHS), the biggest single provider of healthcare in the world, and one of the biggest supporters of research either directly or indirectly through the provision of clinical facilities for use by other sponsors. But does this research get published in the journals that practising clinicians actually read so that it can influence them in their daily work? Teresa Jones, Dr Stephen Hanney and Professor Martin Buxton of the Health Economics Research Group (HERG) at Brunel University in Uxbridge have been surveying the reading habits of doctors in three specialties – psychiatry, surgery and paediatrics. They have found that they do not correlate very well with the journals in which NHS research is published. Clearly there are many steps between the publication of research in a learned journal and its being put to practical use, but it does seem that the use of citation impact factors as a means to evaluate biomedical research is only giving a small part of the total picture. Other indicators will need developing and we have been engaged with some of them during recent years. They include the references that form the evidence base of clinical guidelines, those in textbooks, and even in newspapers that are an increasingly important means by which biomedical research is brought to the attention of both policy-makers and the general public.

All articles were referred by three people – drawn from 12 countries. Articles were referred during the period October-November 2004.

Grant LewisonDepartment of Information Science, City University, London, UK(g.lewison@soi.city.ac.uk)