The effect of manipulating sociopragmatic variables on task complexity and second language output

• Main Author: Clarke, Victoria
• Published: University of Reading 2012
• Subjects: 401.93
• Online Access: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.577788

Tasks used in language assessments vary in terms of their complexity and the language they elicit. L2 learners' performance varies from task to task. With clear evidence for task-induced variation, a better understanding of the relationship between test tasks and test performance is necessary. Studies conducted to investigate the effects of task features on L2 performance have mainly focused on the effects of cognitive features of tasks. However, the findings of these studies have been inconsistent and have not been replicated in testing contexts. To provide more insight into the concept of task complexity in test tasks, a study was designed in which sociopragmatic variables were used to determine task complexity and its effect on L2 oral production. Forty-two Egyptian L2 speakers of English completed a series of test tasks in English in which the sociopragmatic variables of social distance (D), relative power (P), and the absolute ranking of imposition (I) were manipulated to increase or decrease task complexity. Results from quantitative measures showed that candidates produced more lexically complex language in tasks in which they perceived the power and social status of their interlocutor to be higher than their own. In addition, improvements in syntactic complexity were made in tasks in which the degree of imposition was high. For fluency, the findings were mixed, and no significant effect was found for accuracy. In addition, raters judged learners' performance to be more accurate and complex but less fluent when the task demands were increased. The learners also perceived tasks that had been manipulated in terms of sociopragmatic variables to increase complexity as more difficult to perform. These findings are significant due to the implication of the context and interlocutor on language production and the difficulty with sequencing tasks in the assessment of language.