Developing fluency with multi-word expressions

<p>This three-part study was motivated by the need for empirically tested methods for teaching and learning multi-word expressions to develop fluency in language learning classrooms. Using an action research paradigm in an EFL university learning context in rural Japan, the study draws on earlier work by Boers, Eyckmans, Kappel, Stengers, & Demecheleer (2006) who linked speaking fluency with the use of multi-word expressions, and Wood (2009) who found increases in fluency after multi-word expression focussed teaching and practice with one learner in an ESL context. This study also draws on Nation's (2007) Four Strands framework for fluency building. In the first of the three studies, a conceptual replication of the fluency workshop (including phrase instruction, shadowing, dictogloss, and role-play for example) by Wood (2009) was carried out. In contrast to Wood’s approach, this study contained more than one participant (n = 52) and a control group (n = 35). The control group also followed a fluency building program but without a focus on learning target expressions. Learning effects were tested using pre- and post-test measures, including a cloze test of 30 target multi-word expressions to measure form and meaning knowledge, a dialogue role-play recording between participants to measure speaking fluency and use of multi-word expressions, and the first three levels of the Listening Vocabulary Levels Test (McLean et al., 2015) to measure general vocabulary knowledge. Feedback was collected from participants and teacher/researcher observations were recorded to evaluate the contextual appropriateness of experimental classroom activities. Results showed development of meaning and form knowledge of target expressions, but no discernible development in spoken use of the expressions or fluency for the experimental group when compared with the control group. There was also no difference in general vocabulary knowledge between the groups. The second study (n = 25) incorporated improvements to the class activities, such as adding time limits to activities, and data collection methods, including expanding the role-play scenario, based on feedback from the first study. A qualitative analysis of two speakers’ use of a target expression in conversation suggested that fluent use of target expressions could be achieved within nine class hours if participants had some prior knowledge of the expressions. However, a replication was necessary with more participants and a control group to be able to generalise target expression use results to a wider population. Therefore, the third study replicated the second with more participants (n = 65) and a control group (n = 51). Results confirmed that the adapted experimental teaching activities were effective for developing form and meaning knowledge as well as the use of the target expressions in conversation. However, while fluency improved within the experimental group, the improvement was not statistically significant when compared with the control group. The results from these studies suggest that a focus on multi-word expressions with speaking practice is helpful for developing knowledge of meaning, form and use. Therefore, language teachers are encouraged to layer up opportunities through a range of activities such as shadowing and role-play for learners to encounter and re-use frequent and useful multi-word expressions. Teachers are also encouraged to engage in action research so that they can discover learner preferences applicable to their contexts and adapt their activities to be more effective for learning and more enjoyable for learners. Developing fluency in an EFL context may indeed require more than nine intervention hours, therefore a conceptual replication of the third study with a longer intervention period would help to expand our understanding for how long it takes to attain fluency benefits from multi-word expression focused interventions in an EFL context.</p>