Electronic Thesis and Dissertation Repository

Thesis Format



Master of Arts




Boers, Frank


This study aimed to investigate under what conditions multiple-choice exercises benefit second language learners’ acquisition of lexical phrases. Of particular interest was the question whether the distractors in the multiple-choice items create interference when learners’ later try to recall the lexical phrases. Twenty advanced ESL (English as a Second Language) learners were given 20 multiple-choice items on verb-noun collocations (e.g., run a business, take a toll, speak volumes) followed by feedback. They were then tested on the same collocations two weeks later by means of gap-fill items. The participants were invited to verbalize their thoughts during the exercise and the test. They were subsequently shown the multiple-choice exercise again and asked if they could remember how they had responded to each item and what the correct response option was.

This mixed-methods study revealed poor effectiveness of the multiple-choice exercises overall. When learners failed to produce the correct response in the post-test, this was either because they could not remember which of the response options in the multiple-choice item turned out to be the correct one or because they simply had not recollection of the exercise item. The likelihood of producing a correct response in the post-test increased when learners (a) chose the correct response option in the multiple-choice item, (b) remembered the multiple-choice item, and (c) accurately recalled the feedback received on the item. Individual learner differences and item characteristics also influenced the effectiveness of multiple-choice exercises for learning collocations. The findings suggest that, for multiple-choice exercises to be relatively beneficial for collocation learning, they need to be designed and implemented in a way that ensures a high accuracy rate at the exercise stage.

Summary for Lay Audience

Selected-response items such as multiple-choice questions are commonly used for learning and testing purposes. However, their effectiveness in introducing new language knowledge is debatable. Selected-response exercises require learners to identify the correct answer among distractors, and exposure to incorrect information during a learning phase can be detrimental to later recollection (Remmers & Remmers, 1926). Some studies in the domain of language learning (e.g., Boers et al., 2014; Boers et al., 2017) have reported that such exercises lead to only small learning gains according to tests administered a couple of weeks later, because learners often produce one of the incorrect response options that they were asked to consider in the exercises. However, it remains unclear whether these wrong test responses are due to interference from the distractors in the exercises. After all, it is also possible that learners have simply forgotten all about an exercise item when they are tested after a time lapse. Besides, other studies have shown that making mistakes can be beneficial for learning, provided one recalls the mistake (and that it was a mistake), because remembering the mistake can help one to recall what the correct response turned out to be (Metcalfe & Huelser, 2020). However, this effect of learning from errors has been demonstrated for mistakes learners generate themselves. This may be different in the case of multiple-choice exercises where learners are asked to make a choice between response options some of which might not have spontaneously occurred to them in the first place.

In this mixed-methods study, we examined the impact of multiple-choice distractors when introducing new verb-noun collocations (e.g., pose a threat, run a business, set the tone). Twenty advanced ESL learners participated in the study and were invited to learn 20 verb-noun collocations through multiple-choice exercises, after which they received feedback. Two weeks later, they were tested on the same items using a gap-fill format. They were also provided with the original multiple-choice exercises and asked to recall how they tackled the exercises two weeks previously. The learners were encouraged to verbalize their thoughts during the exercises and the two-week delayed post-test.

The results indicated that some of the failed responses in the post-test could be attributed to interference from the exercise items while in other cases the learners could not recall the exercise item. Unsurprisingly, learners who accurately remembered how they had tacked a given exercise item and what feedback they had received stood the best chance of providing the correct response in the post-test. Interestingly, some learners who vividly recalled tackling certain exercise items were either not certain which response option turned out to be correct or misremembered the feedback they received. The likelihood of correct post-test responses was the highest when learners had chosen correctly at the exercise stage and had accurate recollections of the feedback they had received. Additionally, individual learner differences and item characteristics influenced the effectiveness of multiple-choice exercises for learning collocations.

Should teachers or course designers wish to keep using such exercises, then they are advised to implement them in a way that ensures a high accuracy rate at the exercise stage so as to reduce the risk of competition in memory between correct and wrong responses.