大规模语言测试翻译任务内容效度研究——以CET为例

Examining the Content Validity of Translation Task in Large-Scale Language Tests:A Case of CET

翻译任务是我国大规模语言测试的重要组成部分,但相关效度研究匮乏。本文基于Bachman和Palmer所提出的任务特征模式,以2017-2021年大学英语四、六级翻译任务为样本,从语篇输入和预期回答两个维度探究该任务的内容效度。研究结果表明,四、六级翻译任务整体内容效度较高,但同时也存在题材分布不均、体裁单一、六级翻译难度偏低以及考试构念表征不充分等问题,以上发现对进一步提高翻译任务质量具有重要参考意义。

The translation task is an important part in the large-scale language tests in China,but there is a paucity of empirical research on its validity.Based on the framework of language task characteristics proposed by Bachman and Palmer,this paper aims to examine the content validity of CET-4 and CET-6 translation tasks from both characteristics of the input and characteristics of the expected response by using CET translation tasks from 2017 to 2021 as samples.The results show that the translation tasks of CET have a fairly high degree of content validity.However,there still exist some problems including the uneven distribution of subjects, the lack of genre diversity, the lowdegree of difficulty of CET-6, and the under-representativeness of the stated construct. The resultscould be useful for improving the quality of translation tasks in large-scale language tests.