我国国家考试试题难度评估和预估研究动态  

Research Trends on the Evaluation and Estimation of the Difficultness of National Examination Questions in China

作　　者：陈赛君 刘飞跃 CHEN Saijun;LIU Feiyue(School of Economics and Management,Changsha University,Changsha Hunan 410022,China)

机构地区：[1]长沙学院经济与管理学院,湖南长沙410022

出　　处：《长沙大学学报》2024年第2期90-97,共8页Journal of Changsha University

基　　金：湖南省教育科学“十四五”规划2021年度国家教育考试研究专项项目“基于机器学习的国家教育考试试题难度预估研究”,编号:KS214568。

摘　　要：在将试题难度划分为先验难度和后验难度的基础上,梳理了我国国家考试难度研究存在的先验难度评估和后验难度预估两条主要进路。归纳发现:(1)先验难度评估研究的重点在于确定先验难度影响因素,对此应用了经验归纳、线性回归、方差分析、主成分分析等方法;(2)后验难度预估研究的重点在于确定后验难度与其条件变量间关系,对此应用了研究者规定、线性回归、人工神经网络、支持向量机等方法;(3)在两类研究共同的关键点之一即影响因素或条件变量的赋值方面,两者大都采用由研究者规定的主观标准;(4)迄今为止的研究实践中,先验难度评估的效果无法检验但可作相互比较,后验难度预估的效果存在多种检验指标但实际效果多不理想。总结提出已有研究主要存在以下局限:(1)先验难度评估研究表现出单向度趋同倾向;(2)后验难度预估研究多忽略了对被试因素及被试与试题交互过程因素的考察;(3)分析单位多限于单题,整卷研究不足。研究认为,缺乏共识性的统一研究框架和标准是造成已有研究中模糊性、矛盾性和局限性的根源之一,也是今后试题难度研究需重点突破和改进之处。On the basis of dividing the difficulty of examination questions into priori difficultness and posteriori difficultness,two main research approaches in the study of national examination difficultness in China are sorted out.The results show that:(1)the research on priori difficultness assessment focuses on determining the influencing factors of priori difficultness,and empirical induction,linear regression,analysis of variance,principal component analysis and other methods are applied to this study;(2)the focus of the research on posteriori difficultness estimation is to determine the relationship between posteriori difficultness and its conditional variables,and the methods of researcher’s prescription,linear regression,artificial neural network,and support vector machine are applied;(3)one of the key points common to the two types of research is the assignment of influencing factors or conditional variables,and both use subjective criteria set by the researchers;(4)in the research practice so far,the effect of priori difficultness assessment cannot be tested but can be compared with each other,and there are various test indicators for the effect of posteriori difficultness estimation,but the actual effect is mostly unsatisfactory.In conclusion,the existing studies have the following limitations:(1)the priori difficultness assessment research shows a tendency of unidirectional degree convergence,(2)the posteriori difficultness estimation research mostly ignores the inspection of the factors of the participants and the interaction process between the participants and the examination questions,and(3)the analysis unit is mostly limited to a single question,and the whole volume is not well studied.The study believes that the lack of a consensus unified research framework and standards is one of the root causes of ambiguity,contradiction and limitation in the existing research,and it is also the key breakthrough and improvement point in the future research on the difficultness of examination questions.

关 键 词：国家考试试题 试题难度 难度评估 难度预估 

分 类 号：G632[文化科学—教育学]