对我国高考英语命题中建构性题型答案可控性的探讨

Controllability of Answers for Constructed-Response Items in Matriculation English Test

随着我国高中英语课程标准对测试题型的建议和基础英语教学及测试界对英语语言产出能力要求的提高,全国版和大部分分省命题省份高考英语试卷引入了一些建构性题型,从而扭转传统高考英语试卷中选择性题型占绝对优势的现状,但是,建构性题型答案可控性和高考英语测试的效度和阅卷信度密切相关,最终影响高考公平。本文对全国各地高考英语试卷中的典型的建构性题型进行了研究和思考,借鉴国外实践和作者的思考,提出了系列加强我国高考英语命题中建构性题型答案可控性的措施,以供有关命题机构参考,以尽力确保我国高考的权威性和公平性。

The national matriculation English test papers and most provincial versions of MET have employed constructed-response（CR）test item formats for a less-MC-dominant testing style in response to testing-related recommendations of the National High School English Curriculum Standards, increasing expectations from ELT and English testing experts for English learners＇ productive English language skills. Controllability of answers for constructed-response items impacts dramatically on MET validity and score reliability and potentially threatens college entrance fairness. Based on research into some typical CR test item formats in MET test papers and CR item writing practice in foreign countries and the authors＇ reflections, the authors have recommended a series of suggestions for limiting potential answers for CR items to ensure test fairness of college entrance examinations.