Pre-task planning time and working memory as predictors of accuracy, fluency, and complexity

Abstract

Working memory, which accounts for the ability to process information in the face of interference, is critical to second language acquisition (SLA) and use. The interaction of working memory capacity (WMC) with specific pedagogical interventions is a logical place for empirical SLA research, both to examine the cognitive processes underpinning second language performance and to identify instructional treatments that may differentiate learners based on their WMC. A good candidate for such an examination is planning time, a pedagogical intervention that has been the subject of extensive empirical research, which has, thus far, been largely unrelated to WMC. The study undertaken here considers WMC along with two different types of pre-task planning time (guided and unguided) as predictors of the accuracy, fluency, and complexity of learners' discourse.

Ninety-two intermediate ESL students from seven classes at a community college participated in this study by completing two different working memory span tasks as well as two different "there-and-then" oral story-telling tasks. The treatment condition of the story-telling tasks was manipulated so that learners' performance could be considered in terms of provision of pre-task planning (± planning), type of planning (guided vs. unguided), and order of planning (planning first or planning second).

The results demonstrate that the relationship among type of planning time, order of planning time, and WMC is complex. Task order had a clear effect on learners' production, regardless of the provision of planning time. When learners began the series of story-telling tasks under the + planning condition, their output on the subsequent, unplanned task varied according to whether they had first received guided or unguided planning time. In addition, guided planning time and unguided planning time also have very different effects on learners' production, with guided planning time promoting a focus on accuracy at the expense of complexity and unguided planning time fostering fluency. Finally, this study indicates that task conditions can affect learners with high and low WMC in different ways. Learners with high WMC are more likely to comply with complex story-telling instructions, improving their focus on grammatical form at the expense of fluency, whereas learners with low WMC are more likely to improve their fluency as a result of task repetition, regardless of the task conditions.