科技期刊学术影响力综合评价模型与实证

An empirical study on the comprehensive evaluation model of scientific journals' academic influence

【目的】探讨单一评价模型和组合评价模型在科技期刊学术影响力综合评价中的适用性,以引起科技期刊对组合评价模型的重视。【方法】运用均方根法、熵值法、主成分分析法和因子分析法对18种科技期刊学术影响力进行评价,对评价结果进行Kendall检验,在此基础上建立模糊Borda组合评价模型对样本进行再评价。【结果】单一评价模型通过了Kendall检验,但从方法适用性上看,因子分析法最佳,主成分分析法次之,熵值法适用性较差,均方根法不太适合用来评价科技期刊学术影响力。模糊Borda组合评价模型具有较强的适用性,其评价结果与国际通用的复合影响因子评价指标较为接近。【结论】对科技期刊评价指标进行客观赋权,运用组合评价模型进行评价能够确保评价结果更为科学可信,提高复合影响因子、复合他引影响因子、复合即年指标、被引期刊数和Web即年下载率是扩大科技期刊学术影响力的有效手段。

[Purposes]In order to attract the attention of scientific journals in the evaluation model,w e discuss the problem of the applicability of the single and combination evaluation models in the comprehensive evaluation of the academic influence of scientific journals. [Methods] We use the square root method,the entropy method,the principal component analysis method and the factor analysis method to evaluate the academic influence of 18 kinds of scientific journals. We test the evaluation results by Kendall test,and on this basis,model the fuzzy Borda to evaluate the sample again. [Findings]The single evaluation models have passed the Kendall test,but from the perspective of the applicability of the method,the factor analysis method is the best; the principal component analysis method is second; the entropy method has poor applicability; the square root method is not suitable in the evaluation of the academic influence of scientific journals. The fuzzy Borda combination evaluation model has strong applicability,and its evaluation results are close to the impact factor evaluation index of the international general composite impact factors.[Conclusions] If we weight the evaluation index of scientific journals,using the combination evaluation model to evaluate the influence,it ensures the evaluation results are more scientific and credible. Improving the impact factor,the composite index,the number of cited journals and the Web annual dow nload rate is an effective means to expand the academic influence of scientific journals.