Search

Bibliometric findings obtained through the analyses of articles published in journals are used to create scientific and technological road maps and to develop collection management policies in libraries. Yet such analytical studies reviewing the bibliometric characteristics of Turkish articles are scarce. This paper tests if the distribution of some 518,000 Turkish articles to journals conforms to the Bradford Law. Data comes from the Bibliography of Articles of the Republican Era 1923-1999 (BARE) that was recently published on CD-ROM by the Turkish National Library. Although a significant percentage of all articles in a given subject appeared in a relatively few numbers of journals, the distribution of articles to journals did not quite fit the Bradford Law. Eighty "core" journals constituting only 3.2% of all journals contained one third of all published articles while 182 journals contained half, and 358 journals contained two thirds of all articles. A library collection containing the core Turkish journals will likely satisfy the significant percentage of the information needs of users in scientific subjects. Some 2,151 journals constituting 86% of all journals contained a mere one third of all articles. The number of the most productive core journals in each subject was higher than that predicted by the Bradford Law. Articles in linguistics fitted the law best while articles in technology, medicine, social and natural sciences and mathematics fitted the least. Bibliographies with homogeneous subjects seem to fit the classical Bradford Law better. Findings of bibliometric studies should be used to measure the performance of Turkish science and research, and to develop a science and technology "road map". Core Turkish journals based on bibliometric findings should be acquired by libraries. [ABSTRACT FROM AUTHOR]