CET4题型设计模式研究——基于任务特征理论的大学英语测试实证探索

任务型语言测试观为大学英语测试（CET）提供了新的理论基础。在Bachman与Palmer的测试任务五特征基础上,创设了新题型设计基本模式,并依据该模式设计新题型对CET4测试进行了初步探索。信度和效度实证检验的结果表明,新题型组的均值虽因主观性试题量的增加而略低于旧题型组,但两者间并无显著性差异。而且新题型还呈现出同期效度值比旧题型高的态势：同时其信度系数也较高且具有显著的统计学意义。

An Empirical Study on the Basic Pattern for Designing CET4 New Item Types

This study,beginning with the necessity of reforming CET item types in light of the language testing development,based on Bachman and Palmer's theory of task characteristics,sets up "a basic pattern" for designing new CET items,designs some item types for a CET testing paper according to the pattern,and then tests the reliability and validity of the new paper.The control group(n=30) and the experimental group(n=30) did the traditional and new CET papers respectively.The results reveal that although the mean of the experimental group seems lower than that of the control group for the reason that the new CET paper carries more subjective items,there is no significant difference between the two means.The new item types do not decrease the validity of CET4,on the contrary,the concurrent validity of the new paper is higher than that of the old paper.It is also found in this experiment that the reliability of the new paper proves to be statistically significant.Thus,the study empirically gives support to "the basic pattern" for designing CET item types.