과업 복잡성이 성인 EFL 외국어 학습자의 이야기체 글쓰기에 미치는 영향

Abstract

This present study examined effects of task complexity on Korean university students’ L2 narrative writing performance. Robinson's Cognition Hypothesis (2001a, 2001b, 2005, 2007a, 2007b, 2011) argued that if tasks are sequenced on the basis of their relative complexity, task performance from language learners should be influenced by the complexity of the task, especially with regard to accuracy, complexity, and fluency. In this study, task complexity is manipulated by two elements; tasks with -/+ causal reasoning which was presented as one of resource-directing elements and the ones with +/- task structure as one of resource-dispersing elements in the task complexity features by Robinson (2007a). In this study, the participants were sixty-nine undergraduate students in the English language department at a local university in Busan. They performed four narrative writing tasks, each having with -/+ causal reasoning and +/- task structure based on four different picture stories. The learners' written performance was coded by T-unit, S-nodes, errors, and the total number of words to measure accuracy, fluency, and complexity. Regarding accuracy, a measure of error-free T-units per T-unit was adopted. However, violations of word order, appropriate ellipsis of components in a sentence level, subject-verb agreement, and appropriate use of prepositions were identified as critical errors considering the language proficiency level of the participants. As for structural complexity, a measure of S-nodes per T-unit was employed. The total number of words was calculated for fluency measure. The data was analysed by a series of two-way analyses of variance (ANOVAs) and chi-square tests. The statistical assessment of the data analyses showed that the most cognitively demanding task demonstrated negative effects on accuracy with no significant difference on fluency and complexity, still remaining some subtle and careful interpretations on complexity. Interestingly, the result of the survey conducted to the subjects immediately after task performance showed a large gap between task difficulty experienced by the participants and task complexity operationalized by the researcher. The findings of this study partially support Skehan’s limited attentional capacity hypothesis, completely denying Robinson’s multiple resource capacity model. They also have some pedagogical implications with respect to task design in task-based language learning. The present study suggests task complexity should be manipulated matching up with learners’ target language proficiency and task complexity. As a result, it could lead to building a taxonomy of task types. In doing so, it might give a potential to overcome the limitations of not providing consistent results in research area on task complexity basis.