ORAL FLUENCY DEVELOPMENT ACTIVITIES: A ONE-SEMESTER STUDY OF EFL STUDENTS

Abstract

The development of speaking fluency is a major goal for many EFL language learners and several researchers have proposed frameworks for fluency instruction based on theories of cognitive science. It is unclear however, whether EFL students with restricted opportunities to use English outside of language classrooms can benefit from fluency development activities. This main purpose of this study was to determine whether EFL students’ speaking fluency improved whilst participating in theoretically grounded fluency development activities. In addition, student use of formulaic language, participation in classroom activities, and repetition of previously used lexical items was examined in order to determine its relation to development in speaking fluency. 32 first-year Japanese university students from four intact discussion skills classes took part in the study, which was conducted over a period of 12 weeks. Data were collected by recording student performances in a variety of fluency development activities and interactive communicative tasks. There were four main questions investigated in this study. The first research question was focused on long-term fluency development by using multi-level modeling to determine whether gains were made in fluency measures in four 2-minute speaking monologue tests that were conducted at regular periods throughout the semester. Three in-class performance variables - the amount of repeated words, the amount of tokens spoken, and the amount of formulaic language spoken, and three individual difference variables—willingness to communicate, extraversion, and first language fluency were also examined to determine if they were related to any growth observed. The second research question looked at short-term fluency development across a time-pressured speaking activity and similarly used multi-level modeling with the same predictor variables. The third research question concerned the relationship of complexity and accuracy to the fluency measures derived from the monologue speaking tests. Finally, the fourth research question was an investigation of the relationship between the objective fluency measurements and subjective expert ratings. The results indicated that the participants made very small, but significant gains on their mean length of pause across the monologue speaking tests. While none of the predictor variables had a strong relationship with this development, post-hoc analyses suggested that other fluency measures and oral proficiency level could have covaried with this growth. There was also a small but significant improvement in the phonation/time ratio, however, none of the predictor variables appeared to covary with this development. No other significant relationships were found in the long-term fluency measures. For the short-term fluency measures, significant growth was seen across the three deliveries of the speaking activity. While several of the predictor variables, most prominently repetition and the number of tokens spoken, had a significant relationship with this growth, a closer examination revealed that the degree of covariance was extremely slight. The relationship among complexity, accuracy, and fluency also became more significantly correlated over time, suggesting that learners produced higher quality samples of language as the study progressed. Finally, the expert ratings had significant correlations with three of the five fluency measures, indicating that human raters were able to detect small differences in spoken fluency. The findings of this study show that fluency can develop in instructed foreign language settings, however, the role of practice and repetition might be more complex than has been suggested in the research literature. This study provides some insight into that complexity and suggests a number of directions that can be followed to understand more about fluency development.