chapter 5. 



outputs and impacts 



New knowledge resulting from 

 R&D is sometimes identifiable by 

 discrete events which can be used 

 as measures of the output of R&D 

 activity. Scientific literature and pa- 

 tents are two such output indicators. 

 Scientific publications reflect re- 

 search of significance to the schol- 

 arly community; patent registration 

 reflects inventive activity of poten- 

 tial commercial consequence. 



The impact of new knowledge re- 

 sulting from R&D is more difficult 

 to measure, especially with regard 

 to basic research. R&D would, how- 

 ever, be expected to have an impact 

 on such economic factors as pro- 

 ductivity, demand for technological 

 know-how, and international trade. 

 These factors are, in turn, quantifi- 

 able in the form of manufacturing 

 output, royalties and fees, and tech- 

 nology-intensive trade; these can be 

 used as rough measures of the im- 

 pact of a nation's R&D efforts. 



scientific literature 



The measurable outputs of S/T lit- 

 erature include journal articles, con- 

 ference papers, technical reports, and 



literature citations. Conventions for 

 measuring the volume and impor- 

 tance of these various outputs differ 

 among countries,^^ and Japanese- 

 developed indicators for most of the 

 outputs tend to assign higher levels 

 of absolute and relative productivity 

 to Japan than do similar U.S. -de- 

 veloped indicators. 



A recent Japanese study by the 

 Mitsubishi Research Institute (MSK) 



'^There are a number of reasons for national dif- 

 ferences in indicators. Regarding scientific publi- 

 cations, for example, U.S. and European researctiers 

 use the SCI for many of their analyses. Critics of 

 this data base argue that it (1) overemphasizes sci- 

 ence relative to technology, (2) relies on a limited 

 set of the world's influential scientific journals {over 

 3,500 since 1981), and (3) the majority of these pub- 

 lications are in English, fiou'ever, coming from the 

 preshgious, world-class journal set which publishes 

 articles from a variety of countries (countr\'-of-au- 

 thor is assigned at the individual article level), the 

 articles do represent a standard of quality and po- 

 tential impact. Another benefit of the SCI data base 

 is that it provides citation data with which to mea- 

 sure an article's subsequent usage or impact. 



A bibliometric data base has been developed by 

 MSK which includes more than 10,000 journals with 

 greater representation of technology and of non- 

 English journals. The MSK system reviews all ar- 

 Hcles from this broader base, including those articles 

 which present a minimal level of new data collection 

 and/or original analysis. For more information on 

 MSK study, see footnote 36 in this chapter. 



estimated the world volume of sci- 

 entific literature output for the years 

 1972, 1977, and 1982 and found that 

 the amount of such literature rap- 

 idly increased over this 10-year pe- 

 riod.'''' During 1972-82, the volume 

 of Japanese scientific literature grew 

 at a faster rate than the world total, 

 slightly increasing the Japanese share 

 of each type of literature (table B-25). 

 This growth was most substantial in 



"MSK surveyed SAT publications in the fields of 

 mathematics, physics, biology, geology, medicine, 

 engineering, and agriculture. Data were collected 

 to cover several newly industrializing countries in- 

 cluding Korea, Taiwan, Thailand, and Indonesia, as 

 well as the major industrialized nations. For sci- 

 enhfic publicarions, MSK drew from field-specific 

 data bases: Mathfile, INSPEC A for physics, CA 

 Search for chemistry, BIOSIS Preview for biology, 

 GEOREF for geology, Excerta Medica for medicine, 

 COMPENDEX for engineering, and CAB Abstracts 

 for agriculture. For technical reports, a data base 

 was created from reports listed by the U.S. National 

 Technical Information Service, Britain's System of 

 International Gray Literature and several Japanese 

 sources. The principal sources for the conference 

 papers included the Conference Paper Index, ISI's 

 Index to Scientific and Technical Proceedings, 

 Chemical Abstracts, the EI Engineering Meeting 

 Bullehn, and JlCST's International Conference file. 

 For more detail, see Mitsubishi Sogo Kenkyusko, 

 Kjigiiku Gijutsu joho no Kokusai-teki Ryutsu no Arikata 

 m Kiinsuni (Tokyo, Japan: MSK, March 1984), p. 7ff. 



31 



