Exploring task design, implementation procedure and second language oral performance : the effects of topic familiarity, task repetition and task types

Abstract

The development of communicative competence has become an important goal for second language (L2) learners. The use of oral tasks, which is proposed in the task-based language teaching (TBLT) approach, has been claimed to be effective to develop L2 learners’ communicative skills. In task design procedure, selecting a topic, task type, and implementation procedure are indispensable steps for language teachers. Previous empirical studies have offered pedagogical suggestions for task design, but some design variables (e.g. topic familiarity) remain relatively under-explored. Teachers may still feel uncertain about whether and how familiar/unfamiliar topics, different types of tasks, and task repetition affect L2 learners’ speaking performance. More studies about task design factors (topic familiarity and task type) and implementation variable (task repetition) are needed to provide more empirical data for the research field and L2 teaching.

This experimental study aims to explore and compare how topic familiarity, task repetition, and task types affect the oral performance of English as a foreign language (EFL) learners. Sixty undergraduate students from a university in China, divided into lower- and higher-proficiency groups, performed four oral narrative tasks, consisting of two picture-based storytelling tasks and two short speech tasks. One familiar topic and one unfamiliar topic were designed for each type of task. The participants then repeated the same tasks to the same audience. To capture their inner thoughts during speech production, stimulated recalls were collected from twenty-one participants, after their task performances. 480 pieces of task performance were quantitatively measured in terms of complexity, accuracy, and fluency (CAF), as well as discourse organization. Statistical analyses (repeated measures MANOVA) were conducted. Stimulated recalls were qualitatively analysed based on Levelt’s speaking model (1999), which includes conceptualisation, formulation, articulation, and self-perception.

Statistical analyses of oral discourses showed that familiar topics encouraged more complex, accurate and fluent oral production than unfamiliar topics. The participants had more structurally complex, accurate, and fluent speech production in their second performances than in the first performances. Higher accuracy and complexity were spotted in the short speech task performances than in the picture-based storytelling task performances. For interaction effects, participants produced the most grammatical mistakes for the picture-based storytelling task with an unfamiliar topic, and task repetition was more effective in improving complexity, accuracy, and fluency when combined with unfamiliar topics. Moreover, task repetition was more useful for enhancing lower-proficiency learners’ accuracy and fluency. Stimulated recalls revealed that topic familiarity may facilitate the conceptualisation and formulation stages. Task repetition may ease the cognitive load of the conceptualisation and formulation stages. Picture prompts might exert more cognitive load on the conceptualising of content and lexical choices than non-picture situations. Both theoretical and pedagogical implications are drawn. The findings support the claims underlying TBLT theories. Empirical data is provided to link task design and implementation factors to the speech production process, especially its conceptualisation and formulation stages. A two-dimensional topic familiarity construct was also proposed. In L2 teaching, repeating unfamiliar topics may be more effective for improving CAF, and task repetition may be more useful for lower-proficiency learners.