خود استنادی ژورنال، ضریب تاثیر، نفوذ ژورنال و تاثیر مقاله

This paper examines the issue of coercive journal self citations and the practical usefulness of two recent journal performance metrics, namely the Eigenfactor score, which may be interpreted as measuring “Journal Influence”, and the Article Influence score, using the Thomson Reuters ISI Web of Science (hereafter ISI) data for 2009 for the 200 most highly cited journals in each of the Sciences and Social Sciences. The paper also compares the two new bibliometric measures with two existing ISI metrics, namely Total Citations and the 5-Year Impact Factor (5YIF) (including journal self citations) of a journal. It is shown that the Sciences and Social Sciences are different in terms of the strength of the relationship of journal performance metrics, although the actual relationships are very similar. Moreover, the Journal Influence and Article Influence journal performance metrics are shown to be closely related empirically to the two existing ISI metrics, and hence add little in practical usefulness to what is already known, except for eliminating the pressure arising from coercive journal self citations. These empirical results are compared with existing results in the bibliometrics literature.

Evaluating research quality is fundamental to the Sciences and Social Sciences. Research assessment rankings are essential to evaluate the research performance of individuals and the quality of academic journals. The perceived research performance of individual researchers is crucial for hiring, firing, tenure and promotion decisions. In the absence of clear signals regarding the inherent, and frequently latent, quality of published research, the perceived quality of a journal is frequently used as a proxy, albeit inappropriately, for the quality of a research paper. This situation arises especially in the Social Sciences, where citations to published papers do not seem to be as widely used as they are in the Sciences. Most journal performance metrics are based on alternative transformations of citations and Journal Influence. The Thomson Reuters ISI Web of Science (hereafter ISI) database [16] is a leading high quality database for generating research assessment measures, especially citations, to evaluate the research performance of individual researchers and the quality of academic journals. Although there are caveats regarding the methodology and data collection methods underlying any database (see, for example, Seglen [18], Chang and McAleer [6] and [7], Chang et al. [8], [9], [10] and [11] for caveats regarding ISI), the ISI citations database is the oldest source of rankings criteria and the benchmark against which other databases are compared. This paper examines the issue of coercive journal self citations, and the practical usefulness of two new journal performance metrics, namely the Eigenfactor score, which may be interpreted as measuring “Journal Influence”, and Article Influence score, using ISI data for 2009 for the 200 most highly cited journals in each of the Sciences and Social Sciences. The paper also compares the two new bibliometric measures with two existing ISI metrics, namely Total Citations and the 5-Year Impact Factor (5YIF) (including journal self citations) of a journal. It is shown that the Sciences and Social Sciences are different in terms of the strength of the relationship of journal performance metrics, although the actual relationships are nevertheless very similar. Moreover, the Journal Influence and Article Influence metrics are shown to be closely related empirically to the two existing ISI metrics, so that they add little to what is already known about journal impact, except for eliminating the pressure arising from coercive journal self citations. These empirical results are compared with existing results in the literature. The plan of the remainder of the paper is as follows. Section 2 presents three key research assessment measures (RAM), namely the 5-Year Impact Factor (5YIF) (including journal self citations) of a journal, Eigenfactor (or Journal Influence) score, and Article Influence score, and discusses the incidence of coercive journal self citations in several disciplines. Section 3 reports some empirical analyses of these key RAM, as well as Total Citations, and compares the results with those that are available in the literature. Section 4 gives some concluding remarks.

Although the Sciences and Social Sciences are dramatically different in terms of the strength of the underlying relationship of the journal performance metrics considered in this paper, the actual empirical relationships are broadly similar. As both Article Influence and 5YIF are measured over a five-year period, although based on different definitions, it is perhaps not altogether surprising that the two scores are highly and positively correlated. Given the very high correlations between the Eigenfactor score (or Journal Influence) and Total Citations, and between the Article Influence score and 5YIF, and the corresponding high R2 values for the simple linear regressions, the Eigenfactor score and Article Influence score would not seem to be entirely necessary for the Social Sciences, and not at all necessary for the Sciences, relative to the leading journal performance measures that are already available, namely Total Citations and 5YIF, respectively. This statement holds in spite of the fact that neither the Eigenfactor score (or Journal Influence) or Article Influence score suffers from coercive journal self citations, whereas Total Citations and 5YIF are affected by coercive journal self citations, though to unknown degrees. Ideally, ISI [16] will in the future report a 5-Year Impact Factor excluding journal self citations, as ISI already reports a 2-year impact factor excluding journal self citations. However, as impact factors that exclude journal self citations are necessarily lower than impact factors that include journal self citations, whether coercive or not, journal editors and journal publishers are unlikely to advertise the former. As the journal performance measures captured in the Eigenfactor score and Article Influence score, which are said to measure “importance” and “prestige”, respectively, add little to what is already available in the ISI Total Citations and 5-Year Impact Factor (5YIF) of a journal, we have no hesitation in concurring with Indiana Jones, who made the following remark regarding his competitors who were searching for the Lost Ark of the Covenant: “They’re digging in the wrong place!”