The Effects Of Tasks With Different Involvement Loads On The Retention Of Formulaic Sequences

In recent years, studies investing effective ways of learning formulaic sequences have received increasing attention, for formulaic sequences are widely used in the English language and they help learners to be more natural and fluent in both speaking and writing. The involvement load hypothesis (Laufer & Hulstijn,2001) states that vocabulary retention is closely related to the involvement loads (a combination of need, search and evaluation) of tasks which involve the use of target vocabulary. Empirical studies so far have generally confirmed that the greater the involvement loads, the better the vocabulary retention. On the basis of the involvement load hypothesis and previous empirical studies, this study investigated the following three questions:whether tasks with different involvement loads led to differences in the retention of formulaic sequences in both the immediate and delayed post-tests; whether tasks with different involvement loads led to differences in the receptive retention as well as the productive retention of formulaic sequences, and whether different tasks led to different degrees of retention in culturally familiar and unfamiliar formulaic sequences.Four classes in Nanjing Agriculture University participated in this study. One class participated in the pilot study for the selection of target formulaic sequences, while the other three classes were the three groups assigned to different tasks in this study. The three tasks (true/false, fill-in, sentence-making) differed in the involvement loads they induced. All participants were given a post-test immediately after they finished the tasks and another post-test one week later. The scores of the two tests were submitted to SPSS for analysis. To better interpret the results, two students from each group were selected for an interview. The analysis of the test scores generated the following results. First, tasks with different involvement loads have a significant effect on the retention of formulaic sequences in the immediate post-test. Retention scores were highest in the sentence-making task, lower in the fill-in task, and lowest in the true/false task. However, tasks do not have a significant influence on retention in the delayed post-test. Second, the three groups of participants revealed significant differences in the receptive retention of target sequences, but did not show much difference in their productive retention. Third, tasks with different involvement loads have a significant effect on the retention of both culturally familiar and unfamiliar sequences, but their retention patters are different. Scores were higher for the culturally familiar sequences than for the culturally unfamiliar sequences when the receptive knowledge was tested. Scores were higher for culturally unfamiliar sequences than for culturally familiar sequences when productive knowledge was tested.The research findings have both theoretical and practical values. Theoretically, this study provides evidence that the involvement load hypothesis (Laufer & Hulstijn, 2001) applies to the retention of formulaic sequences. This study also revealed different retention patterns of two types of formulaic sequences, culturally familiar sequences and culturally unfamiliar sequences. Practically, this study helps us find which task is more effective for the retention of formulaic sequences, and it also helps language teachers design appropriate tasks for their students'learning of formulaic sequences.