The effect of task type on the initial learning and retention of collocations: Evidence for or against the involvement load hypothesis

This research aimed to explore Laufer and Hulstijn’s (2001) assertion that neither input nor output tasks are inherently better or more effective. Instead, they argue that the effectiveness of a task is solely determined by the level of involvement it generates. Ninety English translation students at Islamic Azad University participated in one of three collocation learning tasks, each designed to have the same theoretical level of task-induced involvement (all with an involvement index of 3). The tasks varied in type: the first was an input task involving reading comprehension and looking up unknown collocations in a dictionary, while the second and third were output tasks, which included completing sentences with the target collocations after looking up their meanings, and writing sentences using the target collocations. The initial learning and retention of target collocations were evaluated using two posttests: one conducted immediately after the treatment and another two weeks later. Contrary to the Involvement Load Hypothesis (Laufer & Hulstijn, 2001), the study's findings revealed that the type of task (whether input or output, receptive or productive) did influence both the initial learning and long-term retention of the target collocations. Based on Swain’s (1985) Output Hypothesis and the distinction between receptive and productive tasks, several potential explanations for these results have been suggested. The study's findings could contribute to the creation of more effective pedagogical tasks.