As of September 1999, only two percent(51 libraries)of the 2,560 public libraries in Japan were providing access to their OPACs via their WWW home pages. First, we investigated and evaluated 46 OPACs, which were accessible from September 17 to October 13, 1999, using standardized queries of 58 items. Then, on the basis of the survey results, we selected three OPACs--one to serve as a typical example of a sophisticated system, one for an average system, and one for a simplified system. Then we asked Internet users to access and use all or at least one of those three OPACs, and to fill in a questionnaire, consisting of seven items, about the OPACs. These evaluations took place during 49 days between October 14 and December 1, 1999. Among 119 respondents, 85 users were already experienced in using OPACs. Issues revealed from our two surveys are:(1)users' accessibility to the computer facilities, (2)quality of the OPAC manual, (3)browsing function of indexing terms, and(4)error messages generated from the system to the users in response to users' queries. The most notable point from the OPAC user survey is that users prefer a system that is friendly and easy to use over one that is sophisticated and offers full functionality.

View full abstract

In this paper, we examine some important consequences, which are to be imposed on the interpretative framework of bibliometric phenomena, of the fact that almost all statistical measures or parameters change systematically according to changes in the sample size of Lotka-type bibliometric data. Firstly, we empirically reconfirm the fact that most statistical measures change systematically according to changes in the size of bibliometric data, and then, on the basis of this recognition, we examine possible arguments relating to the interpretation of bibliometric phenomena by means of statistical measures derived from bibliometric data. We argue that it is more interesting and promising for the characterisations of bibliometric phenomena to take an standpoint in which, recognising the distinction between sample and 'abstract' population, the sample size dependency is explicitly taken into account.

View full abstract