Optimal scaling regression analysis for impact factors of SCI papers output / 中华医学科研管理杂志

ABSTRACT

Objective To analyze the SCI papers output and its impact factors of a hospital during 2013-2017,to provide supports and proofs for scientific research administrators to update incentive policies for stimulate growth of SCI Paper publication.Methods Dynamic sequence and Optimal Scaling Regression methods were used to conduct statistical analysis of the output and impact factors of hospital-published SCI papers.Results The average increased number per year of hospital-published SCI papers output was 30.3 with an annual increase rate of 37.5％,which was 9.5％ higher than the national average.In 2017,the growth rate of a fixed base of hospital-published SCI papers was increased to 257.4 ％,which was 89％ higher than the national rate during the same period.The model was statistically significant (F =10.513,P＜ 0.001,adjusted R2 =0.409) according to the Optimal Scaling Regression analysis of 180 first/corresponding authors out of 516 SCI papers.We found that authors' status as tutor of post-graduate student (P＜0.001),average project fundings (P=0.001),professional titles (P=0.048) and genders (P=0.005) have significant impacts on the output of hospital-published SCI papers (P＜0.05).The sum of the importance of the four factors above was 0.961,namely account for 96.1％ of the whole model.The quantitative score conversion chart shows that doctoral tutors,high average project funding,senior professional titles,and male authors of the SCI papers have higher quantitative scores.Conclusions Tutor status,average project fundings,professional titles and genders could impact the output of SCI papers.Scientific research administrators could adjust incentive policies of SCI papers,based on the science platform construction inclining to doctor tutors,adding matching funds for science and research,bonding professional title appraisals and rising incentives for female researchers on SCI papers,to promote SCI papers output.