A Comparative Study Of The Cognitive Validity Of Two Types Of Listening Tasks

The task type of listening comprehension part in College English Test(CET)has recently been reformed both in Dec.2013 and in Jun.2016.Listening passages have always been a compulsory test content.Two task types are involved in the listening passages,one of which is the multiple-choice questions(MCQ)and the other is the blank-filling task.The former task type stays unchanged over the years,but the latter has undergone constant changes,from compound dictation to words and phrases dictation,and finally it will be deleted from Jun.2016.What makes MCQ in listening passages so special that it has always been adopted in such a large-scale high-stakes test?Why does the blank-filling task suffer so much and will be deleted in the end?Are there any differences in the cognitive validity?This thesis focuses on these two task types,examining their cognitive validity in detail.In examining the cognitive validity,cognitive process questionnaires were needed in the present study to provide data for further statistical analysis.The questionnaires were developed according to the listening cognitive process models proposed beforehand with reference to Weir's writing cognitive process questionnaire in 2014.Then quantitative research method was employed.An experiment was done,in which 99 sophomores from a university in Shanxi who were going to take part in CET 4 were asked to finish the tasks of listening passages in an original test paper(Section B and Section C)in Jun.2015 and two related questionnaires within 50 minutes.Primarily by means of the factor analysis conducted with Statistical Packages for Social Science(SPSS),several factors were respectively extracted from students'answers to the two questionnaires.Through the qualitative analysis of these factors and a comparison between these factors with the cognitive process models proposed before,the author found that 1)the two types of listening tasks can elicit most of the processes employed by an elite listener in a real-world listening event,namely,goal-setting,decoding,discourse construction and monitoring.The processes of parsing and meaning construction proposed in the previous listening cognitive process model are not significantly reflected by the two task types according to SPSS.2)The cognitive validity of Section B:multiple-choice question(MCQ)is higher than Section C:words and phrases dictation(blank-filling)task.The findings suggest that listening passages in CET 4 value higher level cognitive processes,especially discourse construction.In daily English studies,teachers should pay attention to cultivating students' higher level cognitive processes and students should actively expose themselves to the English language environment and learn to employ presently available information,such as context and background knowledge to complete the tasks.