Since the 21st century, it has been difficult to solve various and complex problems of the modern society in the area of individual discipline, leading to the development of the interdisciplinary study to examine a problem through borrowing and integr...
Since the 21st century, it has been difficult to solve various and complex problems of the modern society in the area of individual discipline, leading to the development of the interdisciplinary study to examine a problem through borrowing and integrating knowledge and methodologies of many studies from a new perspective and developing a proper solution. In interdisciplinary research fields, a research paper is likely to cite a variety of theses, also with many possibilities to be cited. However, there's no clear examination about whether interdisciplinary studies encouraged on a national basis have made influential findings. So, it's required to examine a correlation between interdisciplinarity and research impact.
Therefore, this study intends to understand an interdisciplinary structure of Library and Information Science(LIS) recognized as a interdisciplinary field and analyze relationship between its interdisciplinarity and research impact.
As for scientific journals classified into the subject category, Information Science & Library Science of the subject field, JCR's LIS, this study analyzes 10 kinds of journals, at the top of 2011 JCR's 5-year Journal Impact Factor(5-JIF). On WoS from 2006 to 2010, bibliographic data were collected on 1,873 articles. Using Bradford’s Law, 64 kinds of journals at the top of citation frequency were selected for journal co-citation analysis. Cosine similarity matrix among 64 kinds of journals was framed and Pathfinder Network was generated. Generating a cluster by Parallel Nearest Neighbor Clustering(PNNC) algorithm, subject fields related to LIS were examined and journals having a great impact on network were identified and analyzed using the centrality measure of the weighted network. Citation Rate, Interdisciplinarity Borrowing Index, Brillouin Index and Betweenness Centrality(BTC) were selected as the interdisciplinarity indicator, and as the research impact indicator, Immediacy Index, 5-JIF, Article Influence Score(AIS) and Journal h-index were selected. For the interdisciplinarity indicator and the research impact indicator calculated by 10 kinds of journals, correlation analysis was conducted using Pearson's correlation coefficient.
The findings on interdisciplinary structure of LIS through the network analysis, are as follows.
First, 2 types of clusters were generated by PNNC algorithm. The first type of the cluster was divided into 2, the second into 9 ones. Business and management journals are mainly in clusters located at the central of the network, occupying a great part by LIS and Computer Science. Both subject fields were mainly given at the same journals together, located uniformly on the network. Accordingly, the interdisciplinary structure of LIS may be divided into Social Sciences including Economics, Business & Management, Psychology, Sociology, Communication and Sciences including General & Internal Medicine, Medical Informatics, Multidisciplinary Science.
Second, in order to identify journals having a great impact in the network clusters and decide a representative subject field of the cluster, Nearest Neighbor Centrality(NNC) measuring local centrality was inspected. It showed accord generally between the result analyzing the representative subject field of the cluster applying PNNC algorithm and the result analyzing the subject field of the journal having the greatest NNC by cluster.
Third, in order to identify the general impact on the network, Triangle Betweenness Centrality(TBC) measuring global centrality was examined. Journals of Business & Management, LIS, Psychology, Computer Science fields belonging to the greatest clusters on the network had a much higher impact while journals of General & Internal Medicine and Medical Informatics belonging to the opposite end clusters had a lower impact. Therefore, it may be said that LIS cites Social Science journals more than Science lines, having a higher thematic relation with the subject field of social sciences.
The analytic findings on relationship between the interdisciplinarity indicator and research impact indicator are as follows.
First, it shows that BTC of the interdisciplinarity indicator has a statistically significant higher correlation with 5-JIF and AIS of the research impact indicator. As both cases have a positive relation, it shows that there's a correlation in direct proportion to interdisciplinarity and the research impact. As one of centrality measures of network analysis, BTC has a characteristic that the value of centrality increases when certain node is located at the shortest path among other nodes in the network. Journals of higher BTC are likely to be located at the position connecting many clusters, taking a role of connecting various journals. 5-JIF measures the average impact of a journal, AIS for the average impact of an article recorded in the journal. So, it's regarded that journals frequently cited with other journals have higher average impact of the journal and article.
Second, it shows that there's no statistically significant correlation among the research impact indicator and Citation Rate, Interdisciplinarity Borrowing Index, Brillouin Index excepting BTC of interdisciplinarity indicator. Therefore, it was impossible to verify a clear relation of the research impact with the number, variety and distributional uniformity of other subject fields cited in the journal. In the research impact indicator, there was no statistically significant correlation among the Immediacy Index, Journal h-index and interdisciplinarity indicator. Accordingly, a clear relation among immediate and general impacts and interdisciplinarity of the journal couldn't be verified, either.
This study examined influential journals in the related disciplinary fields through analyzing the interdisciplinary structure of LIS by network analysis. At the various libraries and information centers, it'll be helpful for information service related to LIS and collection development. In addition, this study suggested a positivistic basis that there's a correlation between interdisciplinarity and research impact of LIS, and compared characteristics of every index by introducing and measuring diverse interdisciplinarity indicators and research impact indicators. It's expected to provide useful information in the future interdisciplinarity analysis studies.