Learning collocations through interaction: The effects of the quality and quantity of encounters

This study examined how short-term and long-term retention of two types of collocations (verb-noun and adjective-noun) was affected by the learning context. The experimental research design of the study involved two major experiments. In Experiment 1 (EX1), 109 male Emirati college students were randomly assigned to an experimental group (task-based activities) or a control group (mainstream exercises). EX1 involved 20 verb-noun collocations and consisted of two sub-experiments. In experiment 1a both the control and experimental groups were exposed to 20 verb-noun collocations four times. To clarify the effects of the instructional context, a second experiment (EX1b) was conducted where participants encountered the same collocations four times for the experimental group and eight times for the control group. As for Experiment 2 (EX2), it involved 108 male Emirati college students and targeted 20 adjective-noun collocations, and similarly, in Experiment 2a, both the control and experimental groups encountered the adjective-noun collocations 4 times, whereas Experiment 2b offered the experimental and control groups four and eight collocation encounters, respectively. The treatment consisted of exposing participants in both experiments to the target collocations using two different teaching methods. The experimental groups used four task-based activities that presented collocations as whole units (Ellis’s, 2003 chunking principle) while the control groups used mainstream textbook exercises to learn these sequences, breaking them down into their two constituents (verb + noun and adjective + noun). The experiment was carried out over a two-hour period during students’ regular English classes.The results showed that the experimental group learners in both EX1 and EX2 who used task-based activities to learn the collocations, and were exposed to these sequences four times only as whole units, further outscored their control group peers in all collocation measurements. Statistical analysis of participants’ test responses also showed that the long-term receptive knowledge category of the target verb-noun and adjective-noun collocations in both experiments was higher than the productive knowledge for all experimental groups. This study fills a gap in the research about the importance of the quality of encounter vs. the quantity of encounter in collocation learning and identifies an instructional method that is optimal for learning. The overall results suggest that task-based activities were superior to mainstream exercises and that the quality of encounter appears to be more important than the number of encounter in collocation learning; four highly interactive tasks presenting collocations as whole units, with only four encounters, could be more effective to retain unknown collocations than mainstream exercises (e.g., matching and fill-in) that offered learners eight encounters to the collocations broken down into their constituents. The implications for teachers may be that task-based activities, exposing learners to collocations as whole units, should be part of their language instructional pedagogy if they want learners to retain collocations in their long-term memory. For material designers, a well-balanced course would be one that prioritises collocations as chinks through interactive task-based activities