Analysis Of Disciplinary Difference Of Journal Impact Factor And Its Causes Of Formation

Citation is an important element in the dissemination of knowledge and the development of science,and hence citation indicators have become an important means to evaluate the level of scientists and the influence of journals,which has been widely used in science and technology evaluation.However,the citation indicators of different disciplines often vary greatly.Taking the journal impact factor as an example,the impact factor of science and technology disciplines such as energy and biology is above 3 on average,while that of social science disciplines such as law is only around 1.The research of this dissertation on the causes of formation of this difference is helpful to correctly understand the disciplinary difference of citation indicators,and is of great significance for the rational use of citation indicators in different disciplines and the promotion of disciplinary fairness in science and technology evaluation.This paper focuses on the journal impact factor,which is a citation indicator in the Journal Citation Reports,and selects nine disciplines,mathematics,physics,biology,civil engineering,energy,computer,management,law and psychology,to carry out the research.Firstly,we measured the disciplinary difference of the impact factor from two dimensions: the overall level(median)and the imbalance(Gini coefficient).Then,we conducted an in-depth analysis of the causes of formation of the disciplinary difference of impact factor from three aspects: the growth rate of literature,the number of citations and the proportion of valid citations.Finally,the influence of these three aspects of causes on journal impact factor was explored by regression analysis method.The results show that there are great differences in the overall level and imbalance of the impact factor of different disciplines.For the selected nine disciplines,the median of impact factor of energy(with the highest overall level)is three times that of mathematics(with the lowest overall level),and the Gini coefficient of impact factor of physics(with the highest imbalance)can be regarded as "great disparity" according to economic convention,while that of civil engineering(with the lowest imbalance)can be regarded as "relatively reasonable".By analyzing the three causes of the disciplinary difference of impact factor,it can be inferred that in theory,a discipline with higher growth rate of literature,bigger number of citations and larger proportion of valid citations will have higher journal impact factor.It is found that in terms of the growth rate of literature,the natural science disciplines generally grow slowly and the engineering technology disciplines generally grow fast,which shows the same trend as the journal impact factor.In terms of the number of citations,the number of citations in social science disciplines is more than that in science and technology disciplines,but social science disciplines have a lower journal impact factor.Further analysis shows that the number of citations has a weak positive relationship with journal impact factor,which indicates that the number of citations has a small effect on journal impact factor.In terms of the proportion of valid citation,overall,the proportion of valid citation of science and technology disciplines is higher than that of social science disciplines,which also shows the same trend as the journal impact factor.The analysis result of the regression model shows that the growth rate of literature,the number of citations and the proportion of valid citations are all have positive effect on the median of the impact factor,and the order of the influence degree is the proportion of valid citations,the growth rate of literature and the number of citations.At the same time,the proportion of valid citations also has a positive effect on the Gini coefficient of impact factor.