基于马尔科夫链的校本成绩评价合理性的探索

Explore the Rationality of the School Evaluation Based on Markov Chain

本文利用马尔科夫链的预测模型来探索校本成绩评价的合理性,方法是将校本成绩和高考成绩均转换成Z分数并将Z分数划分成六个区间,统计各区间的人数,利用马尔科夫模型分别预测校本成绩及高考成绩在各区间的概率,并对其结果进行对比分析.结果表明马尔科夫预测模型能比较准确的预测校本成绩的状态,也能够在一定程度上预测高考成绩的状态,但是预测能力非常有限,并探讨了预测误差比较大的原因.

The rationality of the school evaluation scores is explored by the forecast model of Markoff chain in this paper.The school evaluation scores and college entrance examination scores are transformed into z-score which are divided into six intervals. The number of each interval are counted. The probabilities in each interval of the school evaluation z-score and the college entrance examination z-score are predicted by Markoff model which are compared with the actual condition. The results show that the school evaluation state can be predicted accurately but the college entrance examination. The reasons of the big prediction error are discussed.