The Impact of Cognitive Task Complexity and Task Sequence on L2 Speaking Performance: A Technology-Mediated TBLT Study

Abstract

This study investigates how varying cognitive complexity levels of tasks and three predetermined task-sequence orders influence second language (L2) learners’ spoken performance within a technology-mediated task-based language teaching (TMTBLT) environment. Participants completed three monologic crime-reporting tasks, systematically designed to represent Simple, Middle, and Complex cognitive demands, presented in ascending, descending, or interleaved mixed sequences. Analysis focusing on syntactic complexity, lexical diversity, accuracy, and fluency demonstrated a partial inverted-U trend. Specifically, the Middle-level task frequently produced the most balanced complexity, accuracy, lexis, and fluency (CALF) profile across several linguistic dimensions, although this pattern was not consistently observed across all measures. In contrast, the Complex task typically enhanced accuracy but simultaneously constrained lexical diversity. Task sequencing exerted a noticeable effect primarily on fluency outcomes, with ascending sequences facilitating progressive fluency improvements over time. Conversely, beginning with the most challenging task initially diminished speaking speed but ultimately triggered notable recovery in fluency performance towards task completion. These findings lend empirical support to Robinson’s Cognition Hypothesis and Skehan’s attentional trade-off model, underscoring the potential for moderately challenging tasks, coupled with thoughtfully structured sequencing, to optimize oral proficiency development in technology-enhanced classroom contexts. Future research should incorporate a neutral baseline task and functional adequacy assessments to further elucidate these observed patterns and extend their generalizability.