Statistical Analysis On Academic Papers’ Editorial Delay Distribution And Correlation Between Editorial Delay And Citations

This thesis studied the human dynamic behaviors in the publication and the citation process of academic papers,and is mainly focused on the statistical distribution of the publication time(editorial delay time)and statistical analysis on the correlation between the editorial delay and the citation.Ideally,in a reviewing process,it is generally easier for referees to make faster and more reliable decisions for high quality papers,which ideally and on average will later attract more citations.The former statistical investigations revealed that there is week correlation between the editorial delay and the citation.We collected the data of the editorial delay(the duration from the received date to the accepted date or published date)and the number of the citations in the next 3 years after publication of 52,572 academic papers published in Nature,Science and Physical Review Letters(PRL)in last more than 10 years.In the first part,we made statistical analysis on the distribution of the editorial delay time and found that the distributions of editorial delay for the papers published in Nature,Science and Physical Review Letters are well fitted in log-normal distribution,and the data shows this log-normal distribution can be derived from the principle of maximum entropy.We also proposed a publication process model and the simulation gave this log-normal distribution.In the second part,we defined a ratio of above-median highly cited papers and performed analysis on correlation between editorial delay and this ratio.Using this alternative measure,we find that there are strong correlations between editorial delay and ratio of above-median highly cited papers for papers published in Nature,Science and Physical Review Letters,which reveal that on average papers with shorter editorial delay do have larger probabilities of becoming highly cited papers.