publications in each field provides an ap- 

 proximate profile of a country's research effort. 

 These 1973 profiles, based on the 278,894 S.C.J. 

 publications in seven fields, i° are shown in 

 figure 1-7.20 for purposes of comparison, the 

 profile of publications produced by all countries 

 combined is shown also. 



The 1973 profile of the United States was 

 most similar to that of West Germany and the 

 United Kingdom in the relative proportion of the 

 total literature in each field, although chemistry 

 was emphasized somewhat more by the latter 

 two countries. The profile of France's scientific 

 research also resembles the United States, 

 except for a smaller proportion of engineering 

 research on the part of France and a larger 

 fraction of literature in chemistry. 



The country with the profile which differs 

 most from that of the United States in the 

 literature studied appears to be the U.S.S.R. The 

 life sciences (biology, biomedical research, and 

 clinical medicine) represent nearly 55 percent of 

 the U.S. literature compared with just over 20 

 percent of the Soviet scientific and technical 

 literature; conversely, engineering and the 

 physical sciences (chemistry and physics) ac- 

 count for some 20 percent of the U.S. literature 

 whereas the U.S.S.R. published almost 60 

 percent of its literature in these fields. 



Literature citations. The significance of a 

 nation's scientific literature is more important 

 than mere counts of publications. One indicator 

 of quality is the recognition that the research 

 reported was dependent on published accounts 

 of earlier investigations. Such a "citation index" 

 is based on the belief that the most significant 

 literature will be more frequently cited. In 

 support of this assumption are a number of 

 studies which demonstrate high correlations 

 between citation counts and other measures of 



'" These data employ a somewhat different taxonomy of 

 fields of science than that used for the 492-journal set; see 

 Appendix table l-7a for the detailed taxonomy of the fields 

 described in Indicators of the Quantili/ ami Quality of the Scientific 

 Lileralure, Computer Horizons, Inc., 1975. 



^o Because of the way this broad sample was selected, some 

 fields may be understated, such as Russian mathematics 

 However, the scope of the Science Citation Index is determined 

 by a 20-member, international editorial board consisting of 

 two Soviet scientists; one is an expert in the information and 

 documentation sciences area, the other is a mathematician. 

 In recent years, the Science Citation Index has been expanded to 

 include "0 percent of the 1,000 journals most highly cited by 

 articles in some 2,100 journals and 100 percent of the 575 

 most highly cited. 



scientific importance, such as judgments of 

 researchers in the field. -^ 



The quality of scientific research is far more 

 difficult to measure than its quantity. The use of 

 citation indicators is one such approach, but one 

 which requires considerable caution. Some 

 articles may fail to be noticed because scientists 

 do not have access to them, although this 

 characteristic of the availability of a nation's 

 scientific literature is itself an important aspect 

 of the internationalism of science. Articles may 

 be heavily cited only for the criticisms they 

 provoke, or because they deal with minor 

 improvements in methodology. Authors in some 

 countries may cite only a few outstanding 

 references for reasons such as journal space 

 limitations, while similar scientists in other 

 countries may give more complete citations. The 

 particular choice of a sample of journals to be 

 examined can have an effect on international 

 comparisons if countries do not have appropriate 

 representation in the sample. Because some 

 nations concentrate more on applied research 

 than on basic research, they may be un- 

 derrepresented in the scientific literature. 



The data source for this indicator was the 

 Science Citation Index, as augmented for improved 

 coverage of certain fields and countries, com- 

 prising 2,121 journals — virtually all of those 

 included in the S.C.J, for 1973. The index was 

 created by comparing the actual fraction of the 

 world's total citations in a given field with the 

 expected proportion based on that nation's share 

 of the total publications in that field. 



The resulting citation indices are shown in the 

 table below for six fields. An index of 1.0 means 

 that there were as many citations to a country's 

 literature in the field as would be expected from 

 its share of the world's publications; a larger 

 index indicates a proportionally higher level of 

 citation to the literature produced by a country 

 than could be accounted for simply by the 

 volume of its publications. 



-> See "Citation Analysis: A New Tool for Science 

 Administrators", Science. Vol. 188 (1975), pp. 429-432; 

 Jonathan R. Cole and Stephen Cole, Social Stratification in 

 Science. (Chicago: University of Chicago Press, 1973); Eugene 

 Garfield, "Citation Analysis as a Tool in Journal Evaluation", 

 Science. Vol. 178 (1972), pp. 471-478; J. Margolis, "Citation 

 Indexing and Evaluation of Scientific Papers", Science. Vol. 

 155 (1967), pp. 1213-1219; and C. Roger Myers, "Journal 

 Citations and Scientific Eminence in Contemporary Psy- 

 chology", American Psychologiit. Vol. 25 (1970), pp. 1041-1048. 



12 



