Keywords

SLA, utterance fluency, integrated speaking task, ESL, topic, articulation rate, topic familiarity

Abstract

Fluency is a fundamental component of second language (L2) proficiency and plays a central role in both communicative competence and language assessment (Tavakoli & Wright, 2020). This study examined utterance fluency, specifically speed, breakdown, and repair fluency, within integrated speaking tasks, widely employed in high-stakes language assessments such as the Test of English as a Foreign Language Internet-based Test (TOEFL iBT). Although previous research has established strong associations between fluency and both communicative adequacy and overall proficiency ratings (De Jong et al., 2015; Révész et al., 2016), the interaction between task structure and topic remains insufficiently understood due to confounding task effects and variability in rater interpretation (Nakatsuhara, 2012).

Using a within-subjects experimental design, the study analyzed speech samples from upper-intermediate English for Academic Purposes (EAP) learners enrolled at a metropolitan university in the southeastern United States. Participants completed four integrated speaking tasks that systematically varied by task type (listen-and-speak vs. read, listen, and speak) and topic (academic vs. campus). Nine indices of utterance fluency, including articulation rate, pause patterns, filled pauses, and self-repairs, were examined using a two-factor MANOVA and follow-up nonparametric analyses.

The findings revealed that topic complexity had a stronger influence on fluency performance than input modality. Participants demonstrated higher articulation rates and shorter end-clause pauses when discussing campus topics than academic topics. Task type, by contrast, did not yield a significant main effect. However, an interaction emerged for self-repetition ratios, indicating that the addition of reading input produced differing effects depending on topic familiarity. These results suggested that different dimensions of utterance fluency were shaped by the interplay between conceptual accessibility and cognitive load.

The findings highlight the need to scaffold topic familiarity and lexical accessibility to support learner fluency. The study contributes to evidence that utterance fluency is a multidimensional construct, with distinct cognitive mechanisms underlying speed, breakdown, and repair fluency. Implications are discussed for both language assessment and instructional contexts.

Completion Date

2025

Semester

Fall

Committee Chair

Mihai, Florin

Degree

Doctor of Philosophy (Ph.D.)

College

College of Arts and Humanities

Department

School of Teacher Education

Format

PDF

Identifier

DP0029846

Document Type

Thesis

Campus Location

Orlando (Main) Campus

Download

Share

COinS