Promoting discovery learning of formulaic language with the aid of online resources: A classroom-based study with intermediate EFL learners

<p>This classroom-based study was motivated by the call for effective pedagogical paths to help learners of English as a Foreign Language (EFL) acquire multiword expressions (MWEs) in the target language. The central question addressed was how training EFL learners to identify MWEs in authentic texts (text-chunking) with the aid of online dictionaries and the online interface of the Corpus of Contemporary American English (COCA) (http://corpus.byu.edu) affected their uptake and awareness of formulaic language. Learners’ discovery strategies during text-chunking activities were also investigated. The participants were 54 Vietnamese intermediate-level university students, randomly assigned to an experimental or comparison group. Over ten weeks, the experimental group (n=26) was asked to look for MWEs in texts, while the comparison group (n=28) used the same reading materials for content-related activities. Utilizing a mixed methods approach, this project analysed data gathered from a pre-test, an immediate post-test, a five-month delayed post-test, students’ classroom notes and audio-recorded discussions, and immediate and delayed questionnaires. Four main findings emerged from this study. Firstly, the practice of text-chunking benefited the students’ retention of the MWEs they identified in the texts. Whereas test results indicated that learning occurred in both groups, the experimental group demonstrated the greater gains in MWE knowledge. Secondly, experimental learners decided to investigate whether given word strings might be MWEs mostly if they were phrases (a) that they had previously seen, (b) that included known words combined in novel ways, and c) that included unknown content words within familiar structures such as verb + preposition, preposition + noun, or binomials (i.e. noun ‘and’ noun). Also, learners relied mostly on constituent words and language contexts to guess the meaning of newly-found MWEs. Thirdly, text-chunking practice with the aid of online dictionaries and COCA heightened the learners’ awareness of the importance of expanding their knowledge of formulaic language. This heightened awareness endured over the five-month gap between the end of the classroom experiment and the final questionnaire. The intervention also concentrated experimental learners’ MWE learning strategies on those practised during the treatment including underlining new MWEs, taking notes, looking them up in dictionaries and COCA, and discussing with peers. Fourthly, while the experimental group reported increased autonomy in strategy use to support their MWE acquisition, the use of online dictionaries and corpora still posed challenges for learners’ self-study. Stumbling blocks included how to choose the resources to consult and how to extract the necessary information from not-so-user-friendly interfaces. This thesis also discusses pedagogical and methodological implications of the findings so that teachers and learners in EFL contexts can make judicious use of text-chunking activities and online resources for the ultimate purpose of promoting learners’ autonomous MWE acquisition.</p>

Promoting discovery learning of formulaic language with the aid of online resources: A classroom-based study with intermediate EFL learners

<p>This classroom-based study was motivated by the call for effective pedagogical paths to help learners of English as a Foreign Language (EFL) acquire multiword expressions (MWEs) in the target language. The central question addressed was how training EFL learners to identify MWEs in authentic texts (text-chunking) with the aid of online dictionaries and the online interface of the Corpus of Contemporary American English (COCA) (http://corpus.byu.edu) affected their uptake and awareness of formulaic language. Learners’ discovery strategies during text-chunking activities were also investigated. The participants were 54 Vietnamese intermediate-level university students, randomly assigned to an experimental or comparison group. Over ten weeks, the experimental group (n=26) was asked to look for MWEs in texts, while the comparison group (n=28) used the same reading materials for content-related activities. Utilizing a mixed methods approach, this project analysed data gathered from a pre-test, an immediate post-test, a five-month delayed post-test, students’ classroom notes and audio-recorded discussions, and immediate and delayed questionnaires. Four main findings emerged from this study. Firstly, the practice of text-chunking benefited the students’ retention of the MWEs they identified in the texts. Whereas test results indicated that learning occurred in both groups, the experimental group demonstrated the greater gains in MWE knowledge. Secondly, experimental learners decided to investigate whether given word strings might be MWEs mostly if they were phrases (a) that they had previously seen, (b) that included known words combined in novel ways, and c) that included unknown content words within familiar structures such as verb + preposition, preposition + noun, or binomials (i.e. noun ‘and’ noun). Also, learners relied mostly on constituent words and language contexts to guess the meaning of newly-found MWEs. Thirdly, text-chunking practice with the aid of online dictionaries and COCA heightened the learners’ awareness of the importance of expanding their knowledge of formulaic language. This heightened awareness endured over the five-month gap between the end of the classroom experiment and the final questionnaire. The intervention also concentrated experimental learners’ MWE learning strategies on those practised during the treatment including underlining new MWEs, taking notes, looking them up in dictionaries and COCA, and discussing with peers. Fourthly, while the experimental group reported increased autonomy in strategy use to support their MWE acquisition, the use of online dictionaries and corpora still posed challenges for learners’ self-study. Stumbling blocks included how to choose the resources to consult and how to extract the necessary information from not-so-user-friendly interfaces. This thesis also discusses pedagogical and methodological implications of the findings so that teachers and learners in EFL contexts can make judicious use of text-chunking activities and online resources for the ultimate purpose of promoting learners’ autonomous MWE acquisition.</p>

Evaluating Relaxations of Logic for Neural Networks: A Comprehensive Study

Symbolic knowledge can provide crucial inductive bias for training neural models, especially in low data regimes. A successful strategy for incorporating such knowledge involves relaxing logical statements into sub-differentiable losses for optimization. In this paper, we study the question of how best to relax logical expressions that represent labeled examples and knowledge about a problem; we focus on sub-differentiable t-norm relaxations of logic. We present theoretical and empirical criteria for characterizing which relaxation would perform best in various scenarios. In our theoretical study driven by the goal of preserving tautologies, the Lukasiewicz t-norm performs best. However, in our empirical analysis on the text chunking and digit recognition tasks, the product t-norm achieves best predictive performance. We analyze this apparent discrepancy, and conclude with a list of best practices for defining loss functions via logic.

Usability and Accessibility of E-Health Websites

This chapter describes a multi-group research study of the usability evaluation and consequent results from participants’ experiences with the MyPryamidTracker.gov Website application. The authors report on a study of a sample consisting of 25 low-income participants with varied levels of vision (i.e., sighted, low vision, and blind Internet users). Usability was assessed via both objective and subjective measures. Overall, participants had significant difficulty understanding how to use the MyPyramidTracker.gov Website. The chapter concludes with major recommendations pertaining to the implementation of Website design elements including pathway/navigation, search, links, text chunking, and frames layout. An extensive set of actionable Website design recommendations and a usability questionnaire are also provided that can be used by researchers in their future evaluations of Websites and Web services.