Library and Information Science

Studies on the document use in the bio-agricultural sciences in Japan have been carried out by various authors since about 1966. Most of the methods applied to these studies are counting of documents listed in secondary sources of scientific information or citation counting from primary sources. In contrast to these studies, many studies in the bio-medical sciences have been made in Japan based on circulation counting. The difference is probably due to the fact that, speaking generally, circulation records kept in bio-agricultural sciences libraries here are incomplete.

The writer selected the seven Japanese representative academic journals in the bio-agricultural sciences as source journals: Proc. Crop Sci. Soc. Japan, J. Sci. Soil Manure Japan, Ann. Phytopathol. Soc. Japan, Jap. J. Breed., Jap. J. Appl. Entomol. Zool., J. Jap. Soc. Grassland Sci., J. Agr. Meteorol. and since 1967, he has been conducting comparative studies based on data covering a five-year period of 1962–1966 using about 15,800 citations in these journals.

The writer's studies have revealed that citations per paper were eleven in average, of which 83% were journal articles, and that the use of domestic journal articles was 53% in average and foreign journal articles 47% in avearge. The half times of life of publications cited were 5.8 years in average, with variations of 10.6 years, the longest, in the field of breeding and of 5.0 years, the shortest, in the field of agro-meteorology.

The relationship between the rate of information available from journals and the number of key journals has been studied in accordance with Bradford's Law of Scattering. It has been revealed that it is necessary to have eleven titles of key journals to cover 50% of information in one specialized field, 22 titles in average for 60% and almost 400 titles for the total information to make all necessary information in a specific field available, as shown in Fig. 1.

The rate of information available from journals, which is called by the writer, “true in formation rate” is represented by multiplications of the information rate decided by the life of journals, of the rate of information available from journals, and of the ratio of the total information available in journals to the total information in all documents (83% in this case). The result is shown in Fig. 2, from which it is found that the “ture information rate” of ten key journals for a period of five years is approximately 20%.

The bio-agricultural sciences are composed of many specialized fields or subject areas in science and technology. The more fields or subject areas are to be required to be combined, the more journals are requested to be used for the common purpose. The result of an analysis of the rate of overlapped journals (D), drawn from various combinations of the seven Japanese source journals, shows that, setting the rate of information available from journals between 50% and 60% the rate of overlapped journals becomes 20% in average when four specialized fields are combined, about 30% for the combination of seven specialized fields, and about 35% for the combination of ten specialized fields. (See Fig. 3)

The rate of overlapped journals may be, however, dispersed widely according to the characteristics of special fields to be combined. The combination, of which the diviation rates are far below the standard deviation (<M-σ in Fig. 3), always include the field of entomology.

The journals ranked as the most frequently cited in each source journal were the ones same with the source journals, except in the case of the Ann. Phytopathol. Soc. Japan.

An analysis of the cited literature in the seven source journals brings out to the light that approximately 50% of journals information in a specialized field, with an exception of entomology, are to be acquired from the sixteen foreign journals, and the fourteen domestic journals listed in the top thirty journals, and the seven source journals.

Among the foreign journals listed in the top thirty, one can find the following journals as commonly used by agricultural research workers in different fields (See Table 2): As general scientific journals; Nature, Science. As general agricultural journals; Agron. J., J. Agr. Res., J. Agr. Sci. Others; Plant Physiol., Bot. Gag.

It is interesting to see that J. Agr. Res., which has been suspended considerably long ago, is still frequently used.

It may be noted that one of the characteristic elements to be a key journal is that a given journal should have a high frequency of publication. To speak in general, a key journal, if not a source journal, is published as frequently as or more frequently than bi-monthly.

The most defective of the citation counting method in its application to decide key journals lies in the fact that newer journals do never come up in the upper rank on the list of journals. Some journals with longer life times may often contain valuable articles. In this case, the value of a specific article is more highly evaluated than the value of the journal which published the article. It is hoped that examinations of the problems concerning key journals and key articles will be made in the near future.