خلاصه ماشینی:
"Task-Based Listening Assessment and the Influence of Construct-Irrelevant Variance Arshya Keyvanfar Mojgan Rashtchi Islamic Azad University, North Tehran Branch Task-based listening tests such as IELTS require testees to listen to some information on a CD and simultaneously answer the related items.
Thus, when designing and using a test, it is necessary to define these listening tasks in terms of their features as well as the language ability and topical knowledge needed to perform them (Bachman & Palmer, 1996).
Research Hypothesis The purpose of conducting the present research can be summarized in the following research hypothesis: There are differences among three types of questions (task-based simultaneous, task-based consecutive, and multiple-choice questions) in terms of their effect on the listening performance of Iranian EFL learners.
The above research hypothesis was translated into the following null hypothesis: There is no significant difference among the listening performance of the three groups of EFL learners evaluated through task-based simultaneous, task-based consecutive, and multiple-choice items.
Discussion The purpose of the present study has been to investigate the construct validity of IELTS listening module with respect to two possible sources of construct-irrelevant variance, ie question type and timing of answering (consecutive/simultaneous).
To this end, three matched groups of IELTS candidates sat for three different listening tests, the first being the regular IELTS listening test (where students were to simultaneously listen and answer the questions), the second being the same texts followed by pauses of approximately five minutes to give test takers the time needed to respond to the questions consecutively and finally the third being a regular 40-item multiple-choice test of listening."