A study into the effectiveness of pushed/non-pushed spoken output tasks focussing on upper intermediate students in the EFL classroom

Abstract

The study reported upon in this thesis investigated the effectiveness of pushed and non-pushed speaking tasks in a UK university setting with 21 upper-intermediate students of English. Specifically, the study addressed a) if a pushed speaking task produced more language related episodes (LREs) than a non-pushed speaking task b) in what ways did pushed or non-pushed tasks vary in the type of LREs that were produced by learners c) whether a pushed speaking task resulted in better performance in past narrative tenses and d) how student views regarding preference and effectiveness vary according to each type of task. The principal procedure used within this study comprised a pretest-speaking task treatment-posttest design with 21 students from an English as a Foreign Language (EFL) course. 11 of the students were given a pushed storytelling task whilst 10 students were given a non-pushed storytelling task; both tasks were completed with a native English speaker teacher with the only difference being that pushed students were supplied with feedback to inform them of any past narrative tense error. A stimulated recall activity was conducted with each student to ascertain thoughts during the storytelling task followed by a concluding interview to obtain perceptions of each task. Questionnaire data was also obtained from 66 students from the same EFL course to acquire more student views (this sample contained the 21 students from the treatment procedure). Results showed that the pushed storytelling task produced significantly more LREs than the non-pushed task and identified that the most common LRE type for both pushed and non-pushed learners was attributed to output correction. Furthermore, no significant gain in past narrative tense performance was found for either task and much variation was found in student perceptions of task preference and effectiveness with students suggesting merits and drawbacks of both. The study concluded that although significant performance gains were not achieved for pushed speaking tasks, creating a push during spoken output activities can increase instances in which linguistic processing, and subsequently interlanguage development, may occur.