Meta evaluation theory and methods were used to evaluate the review experts of science and technology projects. Dozens of meta evaluation criteria can be found within two categories: the objective data of the experts' review results, i.e. the coefficient of deviation, the Spearman rank correlation coefficient, the reliability coefficient of binary value; the subjective data of the experts, i.e., the degree of experts' participation, the degree of review punctuality, the qualification of experts, and the degree of review seriousness. How these criteria impact each other has few been examined, and how to integrate the objective and subjective criteria need a comprehensive model with considering the fuzzy characteristics of the subjective criteria. This study targeted these two questions. An empirical study was adopted with hundreds of experts taking part in reviewing hundreds of Sci-Tech projects. The impacting relationships among the criteria were analyzed based on the empirical study. In order to deal with the intuitionistic fuzzy data on the subjective criteria and improve the estimation efficiency, an IVIF-BWM (best worst method under interval-valued intuitionistic fuzzy environment) was proposed by combining IVIF and the classical BWM to generate the importance weight for each criterion. The MULTIMOORA (multi-objective optimization by ratio analysis plus the full multiplicative form) was used to determine how to combine these criteria. At last, the proposed meta-evaluation model based on IVIF-BWM and MULTIMOORA was applied in a real case. The case study results supported the accuracy and reliability of the proposed model.

运用元评价理论和方法对科技项目评审专家进行评价。元评估标准可分为两类：专家评审结果的客观数据，即偏差系数，斯皮尔曼等级相关系数，二进制值的可靠性系数。专家的主观数据，即专家的参与程度，准时的评审程度，专家的资格以及评审的认真程度。这些标准之间如何相互影响的问题鲜有研究，如何结合客观和主观标准需要考虑主观标准模糊特性的综合模型。这项研究针对这两个问题。一项由数百名专家参加的实证研究参加了对数百个科技项目的审查。根据实证研究分析了标准之间的影响关系。为了处理主观判断上的​​直觉模糊数据，提高估计效率，提出了IVIF-BWM（区间直觉模糊环境下的最佳最差方法）来产生重要性权重。每个标准。MULTIMOORA（通过比率分析进行多目标优化加上完整的乘法形式）用于确定如何组合这些标准。最后，在实际案例中采用了基于IVIF-BWM和MULTIMOORA的元评估模型。案例研究结果支持了该模型的准确性和可靠性。