Statistical Analysis On Evaluation Index Of Sci-tech Journals

In recent years, with the continuous development of science and technology, it is obviously become very important to reasonably evaluate Sci-tech journals. One of important indicator for evaluation of scientific journal is impact factor. Large amount of agencies consider impact factors as main index to evaluate personal research achievement and influence of research institutes. But impact factors have limitations, for example, it can only reflect influence of papers published two years ago by the journal. In fact, any scientific and technical journals can exist more than two years. Aim at the limitations, analyzes types of evaluation index then corrected the impact factor and tested the corrected index by some data. Main work included in this paper as follow.Firstly, this paper introduces a kind of improved TOPSIS model based on entropy right to evaluate journals--to determine the weights of each index through entropy-right method, then assess the comprehensive index of journals through the TOPSIS method and test index through the data of JCR database in 2009.Secondly, the article analyzes their advantages and disadvantages of impact factors and be cited half-life. Weighted average synthesizing these two indexes then gives a revision of the impact factor, which synthesizes the advantages of long-term and short-term evaluation indexes. This method can make up the age limit of impact factor and make it reflect more accurately the journals cited, according to the data of JCR database in 2006 the journal, make scientific classification and statistical analysis and compare the revised index to verify the rationality of the revised journal evaluation indexes.Finally, we also discussed the prospect of the development of impact factors.