Font Size: a A A

Impact Factor Correction Method Based On Median Construction Of New Indicators

Posted on:2020-03-11Degree:MasterType:Thesis
Country:ChinaCandidate:W J TangFull Text:PDF
GTID:2518306314495804Subject:Books intelligence
Abstract/Summary:PDF Full Text Request
The article introduces the nature of the impact factor is designed to characterize and horizontally compare the influence of the journal.The journals should be evenly distributed or normally distributed between the cited and the articles.When the distribution conditions are not fully satisfied,the influence factor will be distorted to varying degrees.On this basis,new indicators are designed to correct the impact factors,to correct the distortion phenomenon of the impact factors,and further propose the strategy of interdisciplinary journal evaluation.First,the article uses the median of the indexed data of the journal article to be standardized as the correction factor to correct the impact factor.The corrected impact factor I_IFM can be expressed as IIFM=(Median-Min)/(Max-Min)×IIF,IIFM is the influence factor,Median is the median of the cited data,Max is the maximum number of times the cited data is cited,and Min is the minimum number of times the cited data is cited.Then,the article analyzes the effect of the revised index by analyzing the data of the journals in the eight directions of the largest and smallest number of times of the number of cited times in the four disciplines of computer science,biology,mathematics,and physics.By comparison,it is found that the standard deviation of the journals after re-ranking the revised impact factor is lower than the pre-correction ranking.Therefore,it can be considered that the indexed data distribution of the journals with a large value after the correction of the influence factor is more balanced.And in each direction,two specific journals were selected to analyze the distribution of their respective papers to verify the impact factor correction effect.Based on this,the data of Computer Science,Interdisciplinary Applications and Physics,Condensed Matter are further analyzed.In the direction of Computer Science,Interdisciplinary Applications,we analyze the distribution of cited frequencies of the top three influential factors and the top three journals in the revised index by the number of citations and the corresponding number of articles,and study the low citations.The relationship between the ratio and the impact factor distortion phenomenon.In the direction of Physics,Condensed Matter,we extend the concept of the Lorentz curve and the Gini coefficient in the field of economics to the field of journal evaluation.The Gini coefficient is used to quantify the distribution equilibrium of the cited data of the paper.After in-depth analysis of the two angles in these two directions,it is verified that the impact factor after the median correction is helpful to reflect the academic influence of the journal more realistically.Finally,based on the revised impact factor,the paper proposes two interdisciplinary journal evaluation strategies,and empirically analyzes the journals in eight disciplines,and compares the advantages and disadvantages of the two interdisciplinary journal evaluation strategies.
Keywords/Search Tags:Impact factor, Median, Standard deviation, Interdisciplinary journal evaluation
PDF Full Text Request
Related items