Statistical Model of Scientific Research Evaluation

Yifan Li,Fang Li,Zhilan Li
DOI: https://doi.org/10.3969/j.issn.1009-2412.2017.01.007
2017-01-01
Abstract:There are numerous traditional indexes (such as hindex and g-index) to quantify an individual's scientific research output.However,most of them cannot characterize the scientific output of researchers in different disciplines reasonably due to the fact that they ignore the gaps between different disciplines.Besides,after statistical analysis,we found that citations are related to the published time and it is unreasonable to measure the articles which are published in different time by citations.As a result,in order to eliminate the impact of disciplines and published time,we build a statistical model to correct the h-index as well as g-index and put forward a new index:REL-index,as the supplement and development of h-index and g-index.
What problem does this paper attempt to address?