An Analytical Study On The H-index Of International Core Journals In Natural Sciences

Based on statistical perspective, the author analyzed the h-index of international core journals in natural sciences. Choosing 120 high-impact journals of six basic natural fields from ISI and Scopus, the author made a penetrating judgement of the relationship among journal's h-index, impact factor and total cites. Correlation analysis, regression analysis, and other statistical methods were used to get conclusion. Meanwhile, in order to construct comparable h-index of different domains, the author combined qualitative and quantitative methods together to conceive formula normalization and data normalization. This paper could provide a feasible reference system for academic evaluation and journal analysis.In the paper, the author first summed up the current achievements of h-index research such as origination, advantages, disadvantages, and mathematical models. And then, within the same field, quantitative methods were applied to analyze h-index, impact factor and total cites, and two models were chose to be fitted by actual conditions. After that, with the purpose of comparison among different disciplines, the author proposed two normalizations-formula normalization and data normalization—for the conversion of h-index. Finally, the author explored the results' difference of the two algorithms. After removing the objective interference, the conclusion came to be that the two algorithms with the same nature would support each other.The innovation of this paper should be summarized as two aspects: 1) Comprehensively discussed the correlation between h-index and other statistic indicators. 2) Converted h-index into comparable level with formula normalization and data normalization and obtained field-compable indices—he and hf, which construct the reference indicator systen for academic journal evaluation.