A Simulational Research on the Types of Scholastic Ability and Selection Methods

采用计算机模拟技术，从1986年联合第一阶段成就测验(JFSAT)的考试成绩中提取了考生学业能力类型与高考选择方法的关系，确定了9个不同学业能力的正常人群。其中4个以文科和社会科学较好，其余4个以数学和自然科学较好。模拟是根据这些总体参数服从正态分布的随机数，即均值向量和协方差矩阵，从这9个总体中抽取考生进行的。对按正态随机数抽取的受试者进行典型测验，并进行评分。Fullmark被取为100。选择方法主要通过每个测试对象的权重来确定。也就是说，一种选择方法重在日语、数学和外语。另一种选拔方法是重自然科学，重社会科学等等。试点法，不分科目，每个考生取最高分。准确的结果出现在1988年3月发表的报告文件中。

The relationship between the types of the scholastic ability of the examinees and the methods of selection of the university entrance examination was investigated using the technique of computer simulation.The types of the scholastic ability of the examinees were extracted fromt he test scores of the Joint First Stage Achievement Test ( JFSAT ) conducted in 1986.Nine Normal populations with different scholastic ability were identified. Four of them were mainly good in literature and social sciences, the other four of them were mainly good in mathematics and natural sciences.The simulation was conducted by extracting examinees from these nine population according to the random numbers distributed as normal with these population parameters, i.e., mean vectors and covariance matrices.Typical tests were constructed using item parameters of the JFSAT conducted in 1987. Those examinees extracted according to the normal random numbers were subjected to take those typical tests and scored. The fullmark was taken to be 100.The selection methods were identified mainly by weights for each test subject. That is, one selection method put heavy weight on Japanese, Mathematics, and Foreign Language. Another selection method put heavy weights on natural sciences, another on social sciences and so on.A trial selection method, which takes the highest mark of each examinees without matter of the subject, was tried.Precise results appear in the report paper published in March 1988.