sign in sign up
<<
Home , Scientometrics

Comes of Age

Number46 November 12,1979 When I first conceived of the Beck of the department of physical Indexm (SCP), I was almost chemistry, Kossuth Lajos University, totally preoccupied with the problem of Debrecen, Hungary calls it “the quanti- information retrieval. I saw citation in- tative evaluation and intercomparison dexing primarily as a way to search the of scientific activity, productivity and —a way to achieve progress.”7 In more colloquial terms, better bibliographic control. 1 Although ISI”’s Morton Malin explains that “part I had at one time planned a career in the of scientometncs consists of applying of science, it was only later that I number crunching techniques to the began to recognize the tremendous po- study of the science of science, ”a that is, tential of in the history . and especially the of science. z The numbers used by researchers in Robert K. Merton of Columbla Univer- the science of science include but are sity and Allen Newell of Carnegie-Me- not limited to: the number of people re- llon University appreciated much better ceiving scientific degrees, the number than I how the SCI might affect not only of granted to scientists, the socio-hktoncal research but science it- number of scientific articles published, self.J (p. 52-53) There is little doubt that the number of scientists who publish pa- this has come to pass. As citation analy- pers, the number of references appear- sis, properly used, has provided new in- ing in papers, the number of to sigh m into historical and contemporary each paper, the amount of grant money scienceh there has emerged a new field awarded to scientists, and the amount called scientometrics. The SC1 is one of of money budgeted by research agen- the major tools of this field, but it is not cies for scientific activities. The various the only one. numbers may be used in scienc: policy The terms econometrics and socio- or program evaluation studies to mea- metrics have been with us at least since sme the scientific “strength” af varicxs the 30s. The term biometrics is even countries, regions, or a particular older. It was inevitable that the term university. One can follow the grow!h or would be coined. It is sur- decline of various fields or identify prising how long i~ took to come up with “where the action is.” In short, sciento- the term scientometrics. The term is ap- metncs is concerned with the demog- parently derived from the word “nauko- raphics of the worldwide scier, Lfic metria” which has been in use in the community. USSR for many years.s Much scientometnc research is bibh- Scientometncs is defined as “the ometric in nature. Alan Pritchard of the study of the measurement of scientific Northwestern-Polytechnic School of Li- and technological progress. ”b M.T. brarianship, London, defined biblio-

313 metrics as “the application of mathe- growth of science based on the evolving matics and statistical methods to picture of scientific publications over and other media of communication. ”9 the last 300 years. Price himself recalls When the media involve scientific publi- in a later edition of Babylon that this cations, you have bibliometrics of along with Little Science, Big Sci- science. ence, [7 which expanded this work, The science of science emerged very touched off a “continuing series of re- slowly as a distinct specialty after J.D. search papers exploring many different Bemal’s The Social Function of Science quantitative investigations based on the was published in 1939.1° There were counting of journals, papers, authors earlier. attempts to analyze science and and citations. In no time at all there science indicators,ll-ls but Bemal’s were bibliographies and conventions de- work is considered by VasiKl Nalimov of voted to bibliometncs and to sciento- Moscow State University and others to metrics ....”l~ (p. 193-4) Price continued be a turning point in science policy anal- to pioneer in thk area during the 1960s yses. 14 After World War II, the history and 70s. In 1971, H.W. Menard of the and sociology of science received much University of California, San Diego, ex- greater attention. As these specialties panded on Price’s work in a volume matured, their practitioners increasing- called Science: Growth and Change. IS ly began to measure science activity. Yakov M. Rabkin of the University of The increasing tempo of quantitative Montreal agrees that Price’s work set studies led inevitably to scientomet- the study of scientometrics in motion, ncs. even in the USSR.5 “Since the 196os Scientometnc research has wide po- quantitative studies of science in the tential applicability. Government and USSR have developed mainly in re- research institutions have expressed in- sponse to Derek de Solla Price’s re- terest in applying “quantitative knowl- search on patterns of scientific edge of scientific growth to the manage- growth ... . Price’s work attracted the at- ment of science. ”s Scientific indicators tention of Vasilii Nalimov of Moscow are being used in an effort to estimate University and Gennady Dobrov of the the relative “health” of science in vari- Ukrainian Academy of . ous countries. 15 Ultimately, scien- Through their efforts, quantitative tometrics may be used to help nations studies of science became a rapidly make decisions on what areas of re- growing field.” search need funding. Drexel University’s Belver Griffith The measurement and quantification adds, “Although the counting of scien- of science from the 1930s to the early tists or articles or something dealing 1960s generated occasional papers. with knowledge-mongering started at However, as Merton points out, two least as early as the 1920s, a key event events helped speed up the work per- was Derek Price’s 1965 article ‘Network formed in this areas (p. 51-57): 1S1 of Scientific Papers’ .. . . It has.. .a first began producing the SCI, which pro- meaningful meshing of an explicit mod- vided the raw material for many of these el for science with raw data, and the studies, and Derek de Solla Price of results were surprisingly elegant and Yale University published Science Since clever .“ 19 Babylon 16 ( 1961). In the called The growth of scientometrics has “Diseases of Science” Price developed been exponential. Space does not per- some mathematical models of the mit me to list the more than 600 articles

314 on citation analysis in my fiIes. Zo And eludes many Soviet publications not citation analysis is only one part of sci- widely known in the West, entometrics. In 1972, the National Science Board Nevertheless, we can identify several issued its first Science Indicators, which important papers or that is now published biennially and presents appeared within the last ten years and masses of statistical information. zl A contributed to the development of the brief look at its contents page (shown in specialty. For example, the 1971 mono- Figure 1) will give you some idea of the graph by Nalimov and Mul’chenko dis- scope of thk publication. Its stated pur- cusses much of the literature on the pose is to “describe quantitatively the measurement of science up to that condition of science and research in the time. 14 The extensive bibliography in- U.S.”ZZ The bulk of the data presented

Ffgure 1: Contents page of Science Indicators-1976.

INTERNATIONAL INDICATORS OF SCIENCE AND 1 Resources for Research and Development 4 The International Character of Science 10 Technological Invention and . 20 The U.S. Role in International 30 Productivity and Balance of Trade 33 RESOURCES FOR RESEARCH AND DEVELOPMENT 43 National Resources for Research and Development 46 Federally Funded R&D in Functional Areas 51 Research Facilities. 59 Scientific and Technical Information, 59 RESOURCES FOR BASIC RESEARCH 65 National Resources for Basic Research 68 Basic Research in Universities and Colleges 75 Basic Research Expenditures in Federally Funded Research and Development Centers Administered by Universities 80 Basic Research in Intramural Federal Laboratories 82 Basic Research in Industry. 83 Basic Research in Nonprofit Institutions ., 83 Research Outputs and Applications 88 INDUSTRIAL R&D AND INNOVATION : 91 Resources for Industrial R&D. 94 Outputs from Industrial R&D. ~ ~ 108 Social and Economic Returns from R&D and Innovation 125 SCIENCE AND ENGINEERING PERSONNEL 129 Characteristics and Utifir.ation of Science and Engineering Personnel 132 Research and Development Personnel 140 Doctoral Scientists and Engineers 145 Women and Mhorities in Science and Engineering 152 Unemployment Among Scientists and Engineers 15-I Additions to the Supply of Scientists and Engineers 159 PUBLIC ATTITUDES TOWARD SCIENCE AND TECHNOLOGY 167 General Attitudes Toward Science and Technology and Toward Their Practitioners 169 Results of Science and Technology 173 Capabilities of Science and Technology. 177 Public Preferences Regardksg Science and Technology 179

315 in Science Indicafor.r is listed in tables, growth, ~“ “The first method starts with figures, and charts. The work presents the a~ailable data and then arbitrarily excellent raw materials from which pre- chooses a description (usually in the dictions about the future health of sci- form of a mathematical function) to fit encecan be made. Robert Wright, head the data. Predictions about future of the science indicators unit at the Na- growth can then be made by extrapola- tional Science Foundation, points out tion. The second method begins by that the first publication of Science Irr- making an hypothesis about a social dicators led to an increase of $50 million process in science, based on either being made available by the US govern- sociological investigation or on intu- ment in support of basic research in uni- ition. The implications of the hypothesis varsities. ~3 are then explored, a process which of- An in-depth critique of Science ln - ten results in the formulation of a math- dicatorsw asprovided in TowardaMet- ematical function.” ric of Science.. The Advent of Science Another review of scientometric liter- Indicator.~. ~~ Edited by Yehuda Elkana, ature was published in 1977, “A Prog- Joshua Lederberg, Robert Merton, Ar- ress Report on the Quantification of nold Thackray, and Harriet Zucker- Science, ” by Michael J. Moray csifi of man, the volume features the work of the Institute of Theoretical Science, contributors from many disciplines. University of Oregon, I< Moravcsik The book was based on a 1974 c~mfer- noted that scientometric researchers ence held at the Center for Advanced can quantify both scientific inputs and Studies in the Behavioral Sciences at scientific outputs. Primary inputs in- Stanford, California. It covers a wide clude manpower and money, Secondary range of topics relating to science indi- inputs are, for example, the number of cators. For example, models of scientif- laboratory buildings built or the number ic output were examined in one chapter, of computer hours clocked by scientists. while the political contexts of science As Moravcsik points out, little scien- indicators were discussed in another. 1 tometric work has concentrated on participated in this conference with these inputs, 1S1’s Henry Small. zf Outputs, on the other hand. have lent In Evaluative Bibliometrics: The U.re themselves more readily [o scientornet- of Publication and Citation Anal},.vi.~ in nc investigation. Some scientific out- the Eva[uatiorr of Scientific Activi~y, puts include: the number {>fpublishing Francis Narin and the staff of Computer authors, their geographical distribution, Horizons. Inc. summarized the state of the number of papers produced each the art of publication and citation-based year. the number of citations attributed evaluation for the National Science to each paper or to an author. Foundation. ~~ Pubfished in 197fI, this is In the span of time between the publi- a good review source for anyone inter- cation of Moravcsik’s review and now, ested in the subject. there has been no shortage of work in A few attempts have been made to scientometric areas. As he points out, summarize the scientometric research many aspects of scientific growth, c{ml- published thus far and to determine i[s munications. and activity are ripe for in- trends. For example, in their 1974 re- vestigateion. view in Science S[udies, G. Nigel For the researcher in this field, find- Gilbert and Steve Woolgar of the de- ing the right journal in which to publish partment of sociology [University of scientometric studies has been a prob- York, identify two distinct methods lem. Articles on scientometrics have ap- used in studies quantifying scientific peared in journals ranging from Scien

316 and Nature to the Journa[ of Documen- ple, the January 1979 issue covered such tation, Social Studies of Science, Amer- topics as research productivity and the ican Sociological Review, American visibility of the French scientific elite~9 Psychologist, and the Journal of the and the allocation of grants by the American Society for Information Sci- Swedish Social Science Research Coun- ence, just to name a few. Lherature in cil. ~ Contributors to the journal have this field is typical of a rapidly growing included many of the researchers who field. My own contributions to this field have been instrumental in creating and are not unknown to Current Contents@ defining the field. In addition to many readers, and my Essays contain dozens of the people cited above, H. Inhaber, of bibliometnc and scientometric stud- D. Crane, R. Rosen, and J. Vlachjf are ies.2a The newness of the latter term is but a few of the leading contributors illustrated by its absence from my new whose names will be recognized by any- book on citation analysis.q I do refer to one following scientometric research. bibliometncs. Scientomeln’cs is an international Considering this situation, many of us journal. Measuring science has become interested in scientometncs felt that a an important issue not only in industrial journal in this field was long overdue. In countries but also in less-developed September of 1978, the first issue of areas of the world. The public every- such a journal, Scientometn’cs, was where is demanding better use of scarce published. It is edited by Budapest research funds. Scientometrics can pro- chemist Tibor Braun, who is also editor vide the kind of quantitative data that of the Journal of Radioanalytical Chem- legislators can understand. The qualita- istry. It is published jointly by the tive conclusions one can draw from Akademiai Kiado in Hungary (the pub- such data will always depend upon the lishing house of the Hungarian Acad- wisdom used in their application. emy of Sciences) and Scientific As one of the editors of Scientomet- Publishhg Company. The chief editors n“cs, I am excited by the prospective include M.T. Beck, D.J.D. Price, G.M. developments that might be printed in Dobrov of the USSR, and your truly. its pages. As I stated in the introduction Scientometrr”cs is covered in several edi- I wrote to the first issue of the journal, z~ tions of Current Contents and the Social I often read with envy many studies Sciences Citation fndex ‘“. It is pub- which use data compiled at 1S1 a long lished six times a year. Subscriptions are time ago which we were unable, for one handled by Elsevier and cost $76 per reason or another, to analyze. But that year. is one of the prices of success in science. In its first five issues, Scientometn”cs It would have been far more painful had has covered a wide range of topics on our data been totally ignored. science and science policy. For exam- cJ,9~9(s,

REFERENCES

1. Garfield E. A unified index to science. Proceedings of the International Conference cm Scientific Information, 16-21 November 1958, Washington, DC, 2 VOIS.

Washington, DC: National Academy of Sciences, 1959. Vol. 1, p. 461-74, ● 2. ------Citation indexes in sociological and historical research. Amer. Dec. 14:289-91, 1963. ” 3. Merton R K. The sociology of science: an episodic memoir. Carbondale, IL: Southern 111.Univ. Press, 1979. 151 p. 4. Garfield E. Citation indexing—its theory and application in science. technol~g.v and humanities. New York: , 1979.274 p.

317 5. Rabkfn Y M. Review of: Naukometricheskie issledovania ~,khimii (Scientometric studies in chemistry). Social Stud. Sri. 6:128-32, 1976. 6. Brusifovsky B Y A. Partial and system forecasts in scientometrics. Techrrol. Forecast. Sot. Change 12:193-200, 1978. 7. Beck M T. Editorial statement. Scienfomefrics 1:3-4, 1978. 8. MaMn M V. Personal communication. 22 August 1979. 9. Prftclmrd A. Statistical bibliography or bibliometrics? J. Dec. 25:348-9, 1969. 10, Berrsal J D. The sociaf function of science. Cambridge, MA: MIT Press, 196’7.482 p. 11. Cole F J. The history of comparative anatomy. Part I. A statistical analysis of the literature, Sri. Progr. 11:578-96, 1917. 12. Hrdmre E W. Statistical bibliography in relation to the gro wth of modern ci~,iliza[ion: t w,o lectures delivered in the Uni~ersify of Cambridge in May, 1922 by E. Wyndham Hulme. London: Butler & Tanner, 1923.44 p. 13. AUers E S. Periodicals for mathematicians. Science 70:529-94, 1929. 14. NaUmov V V & Mul’cbenko Z M. Measurement of science. Study of the development of science as an information process. Washington, DC: Translation Division, Foreign Technology Division, United States Air Force Systems Command, 13 October 1971. 196 p. 1S. Moravcsfk M 1. A progress report on the quantification of science. J. Sri. lnd. Res. 36:195-203, 1977. 16. Prfce D J D. Science since Bcrh.vkrn. New Haven, CT: Yale [Jrsiversity Press, 1975. 215p. 17. ------Little science, , New York: Columbia Lfniversity Press. 1963, 118 p. 18, Memard H W. Science: growth and change. Cambridge, MA: Harvard University Press, 1971.215 p. 19. Grfffitb B C. Science literature—how faulty a mirror of science? ASLIB Proc. 31:381-91,1979. 20, Garfield E. Editorial statement. Scientometrics 1:5-7, 1978. 21. McGfnnis R. Science indicators/ 1976: a critique. Sot. [ndic. Res. 6:163-80, 1979. 22. National Science Board. Science Indicators 1976. Report of (he National Science Board, Washington, DC: National Science Board, National Science Foundation, 1977.304 p. 23. Wrfght R. Personal communication. 2 October 1979. 24. Elkana Y, Lederberg 1, Merton R K, Thackray A & Zuckennan H, eds. Toward a metric of science: the advent of science indicators. New York: Wiley. 1978. 354 p. 25. Garfield E, MaUn M V & Smaff H. Citatiorr data as science indicators, (Elkana Y, Lederberg J, Mertcrn R K, Thackray A & Zuckerman H. eds, ) Toward a metric of science: the advent of science indicators. New York: Wiley, 1978. p. 179-207. 26. Narfn F, Evaluative bibliometrics: the use of publication and citation anal.vsi.r in the evaluation of scientific activity. Cherry Hill, NJ: Computer Horizons, 1976. 338 p. 27 GUbert G N & Woolgar S. The quantitative study of science: an examination of the literature, Sri, Stud, 4:279-94, 1974. 28, Garfield E. Jkrays of an information scientist. Phdadelphia: 1S1 Press, 1977.2 vols, 29. Beaver D D B & Rosen R. Studies in scientific collaboration: Part 11. Scientific co-authorship. research productivity and visibility in the French scientific elite 1799-1830. Scientometric.s 1(2): 133-49, January 1979. 30. Boalt G & Bergryd U. Differences in research orientation reflected in the allocation of grants. A methodological study. Scientometrics 1(21:151-9, January 1979,

4Reprinted in: Garfield E. Essays of an information scientist Philadelphia: 1S1 Press, 1977, 2 vols.

318