摘要
对多级评分的测验题型,给出了其Bayes模型,把先验分布从无信息先验推广到了更一般的Dirichlet先验分布情形,求出了参数的Bayes估计,指出了其意义。
This paper deals with the Bayesian Model of multi-grade scored items. The prior distribution is generalized from non-information to what is more generally known as Dirichlet. The Bayesian estimates are provided with their significance. The estimates of super-parameters of prior distribution are also given under different sample conditions.
出处
《淮阴师范学院学报(自然科学版)》
CAS
2002年第4期1-6,共6页
Journal of Huaiyin Teachers College;Natural Science Edition
关键词
多级评分
BAYES估计
先验分布
超参数
multi-grade scoring model
Bayesian estimate
prior distribution
super-parameter