题库优化设计的回顾与展望

题库优化设计是指在题库建立前确定一份最佳的题库开发蓝图，使题库更好地服务于组卷。目前，在线性计算机化测验的情景下，题库优化设计方法主要包括：线性规划法、成本函数法；在计算机适应性测验（CAT）情境下，题库设计方法主要包括：影子测验设计法、p-优化设计法。文中首先介绍这些优化设计法的基本思路、具体步骤与相关的应用研究，然后评述当前题库设计研究的局限，指出未来的可能研究方向。

Optimal Item Bank Design: Review and Prospect

In term of the optimal item bank design problems,an optimal item bank blueprint is generated before setting to construct an item bank,so that it will better satisfy the test assembly demands.Item bank design methods for linear computer-based tests mainly include basic linear programming method and cost function method,and the item bank design methods for computerized adaptive tests include cost function method with shadow test,p-optimality method.This present study first reviewed basic ideas of each method,procedures for designing the item bank,and related application studies,then the limits and further development of research on item bank design were discussed.