Page 133 - JOURNAL OF LIBRARY SCIENCE IN CHINA 2018 Vol. 43

ZHAO Xing / Exploring the measurement features of usage data for academic literature 133

citation data and related indexes themselves, which have drawn criticisms from many scholars.
In the application aspect, citation and related quantitative data have become a scapegoat for the
excessive performance management and “non-discretionary” system of academia, and have
raised doubts on the part of academic community. Although it is very difficult for the discipline of
Library and Information Science alone to solve the systemic problems in the application aspect,
the discipline itself is still responsible for guaranteeing the rigor of its methods and the data in the
theoretical aspect.
The approaches toward making a breakthrough include the supplementation for citation data and
related limitations. In September 2015, the Web of Science (WoS) platform launched a new type
of data targeting single literature and reflecting the degree of attention, that is, usage received by
the literature. This is the first time that the WoS platform has made an influence standard that was
not constructed under the citation analysis framework. Different from citation, usage reflects the
attentive behavior of academic users on a broader scale, and summarizes two types of user data,
which are: 1) the download count (usage) of literature, and 2) the export count (usage) of literature.
Thus, it is clear that the academic behavior characterized by usage is more grassroots and can be
described as the fundamental magnitude of data for academic literature.
Following a theoretical exploration of citation and usage, this study adopted the literature of four
disciplines (physics, computer science, economics, and Library and Information Science) as the
objects of empirical study and systematically explored the theoretical and application features of
the magnitude of data usage for the academic literature in bibliometric measurements, such as the
statistical features of usage, the distribution model of bibliometrics, the comparison with citation
measurements, etc. This paper attempts to provide a pilot and basic reference for the research and
application of the magnitude data for literature.

1 Theoretical discussion: Citation and usage

1.1 Limitations of citation and related evaluation data

The citation index is the most widely used data for evaluating academic influence. Although there
are still controversies over whether the citation index is related to or can represent the quality of
a research paper, the viewpoint that the index can reflect academic influence to some extent has
become a consensus in the bibliometrics circle (Liu & Li, 2013; Li & Zhang, 2014). In fact, the
basic evaluation indexes, such as total citation count, h-index, citation per paper and related impact
factor, are all constructed on the strength of this basic viewpoint. The total citation count measures
the overall influence, the h-index evaluates highly influential collected articles, the citation per
paper describes the average influence of collected articles, and the impact factor quantitatively
characterizes the average influence of articles published by a journal. It is clear that citation data
have laid the foundation of the current quantitative academic evaluation system.