Yiboon Chang

DOI: Vol.22(No.1) 9-40, 2019

This study investigated 34 multilingual MBA students’ attitudes and engagement in teacher and peer-group assessment for learning on academic presentations in an intact classroom. In this teacher research, the assessment rubric was designed and implemented by integrating multimodality into academic presentations. Language, visual, auditory, gestural, and spatial design elements were introduced to the students as a means of meaning representation, followed by in-class practices of peer-group assessment on sample multimodal presentations. The students found both teacher and peer-group assessment strongly helpful to improve their presentation skills. They had strong belief in teacher expertise in response to the teacher assessment, and highly valued the opportunity of evaluating peer-group performance to reflect and improve their own presentation skills. Despite some concerns about inconsistent assessment results from peer groups and difficulties in understanding the new concept of multimodality, the students appreciated various audience perspectives and opportunities of learning multimodal literacy practices through the assessment practice. The teacher and peer-group assessment moderately correlated, showing the strongest correlation in the assessment of multimodality. This study suggests potentials of utilizing assessment for learning to enhance student awareness of designing multimodal texts, and meaningful pedagogical implications to develop and implement multimodal assessment.

Assessment for learning, Academic presentations, EAP, Multimodality

Namhee Kang

DOI: Vol.22(No.1) 41-84, 2019

This study aims to explore an effective learning and teaching collocation model for lower-level EFL learners in a general English course integrating two approaches: deductive and inductive data-driven learning. Twenty-four Korean university students were exposed to deductive collocation learning in class and then submitted inductive collocation assignments for one semester. A treatment group did the assignments through Corpus of Contemporary American English (COCA) and a control group through their own resources. Both groups' five writings were illustrated that lexical collocations outnumbered grammatical collocations and the treatment group wrote more collocations that increased gradually in a series of the writings. Also, a pre- and a post-TOEIC scores implicated the collocation learning model was statistically meaningful toward general English proficiency. However, assignment tools didn't tell statistically significant differences presumably because of pre-exposed collocation learning in class and collocation-based online dictionaries that most control group referred to. Student logs, a survey, and interviews showed the model motivated self-regulated learning, exposure to diverse contexts, full comprehension of contexts based on part of speech. Especially, the treatment group took an advantage of authentic and colloquial contexts, but lack of them seemed to cause the control group not to keep motivating their learning in the end.

collocation, data-driven learning (DDL), self-regulated learning, lexical collocation, grammati cal collocation, Corpus of Contemporary American English (COCA)

Eunjou Oh

DOI: Vol.22(No.1) 85-111, 2019

This study investigated learners’ experiences in a scaffolded task of reading online media articles in college English classes offered for the students of A2 level of CEFR. Forty-seven freshmen enrolled in College English1 participated in the study. The adopted scaffolding model was analyzed based on theoretical consideration. Students’ perceptions of controllability and value of the target task along with their perceptions of self-regulated learning and self-efficacy were measured at the beginning and the end of the semester, during which the task was implemented twice. Students’ achievement emotions at three phases (before, during, and after the target task) were also tracked over the course of the semester. The analyses of the data showed that the effects of the scaffolding model proposed were mixed. Students’ perceptions increased in the value of the task and self-efficacy in four language skills after the implementation of the task, whereas their perceptions of controllability and self-regulated learning decreased slightly. In prospective outcome emotions, there was an increase in a positive emotion but a decrease in negative emotions after completing the task. However, anxiety stayed as the predominant prospective outcome emotion after the completion of the task. Possible ways to polish the model were discussed.

scaffolding. authentic texts, media articles, translanguaging, collaborative translating activities, lifelong learners

Junhee Park

DOI: Vol.22(No.1) 112-131, 2019

Computer-assisted language learning has been popular among EFL learners and teachers. In particular, automated English grammar checkers have made a significant contribution to improving learner’s writing ability and reducing the workload of teachers as well. However, studies investigating the validation of the AI-based grammar checker are scarce. This study collected 40 writing samples collected from Korean EFL high school second graders. The grammatical errors from collected samples were analyzed by the AI-based grammar checker and human raters. The study also identified the error types that an AI-based English grammar checker produced by comparing the feedback messages to human-raters’ corresponding corrections. The grammar checker disclosed the wrong feedback, same and different feedback compared to the human raters. The results indicated that the program failed to detect the prevalent grammatical errors and sometimes provided incorrect feedbacks, while it showed several successful ones. Nevertheless, this study shed light on the possibility of application of AI-based computer programs in EFL classroom. In addition, this study will be of interest to the English language teachers who deal with a large amount of students’ writings.

AI-based grammar checker, human rater, Korean EFL learners, English writing, error analysis, automatic error correction, corrective feedback

이장호 Lee Jang Ho , 김혜영 Kim Heyoung , 신동광 Shin Dong Kwang , 양혜진 Yang Hyejin

DOI: Vol.22(No.1) 132-153, 2019

The present study investigated the potential of voice-based conversation AI chatbot for foreign language learning. Among commercially available AI chatbot products, Naver’s Clova was selected, and its English speaking mode was utilized. The dialogs between this chatbot and one of the authors were recorded, and used as the primary data in this study. In analyzing these dialogs, the widely used frameworks in foreign language learning research were adopted, including “Initiate-response-feedback,” “teacher’s question type,” and “negotiation of meaning strategies” frameworks. The results showed the selected chatbot appeared to take the role of a teacher (tutor) in conversation with the user, in that on most occasions it started a conversation (new topic) by asking questions to the user. Furthermore, the chatbot primarily used relatively simple types of questions to guide the user conversations. In terms of the uses of negotiation of meaning strategies, which are considered one of the most important aspects in conversation-driven foreign language learning, the chatbot was found to be rarely using this type of strategies, and the absence of such strategies resulted in communication breakdown with the user. The present study also proposes considerations for the future development of the chatbot for foreign language learning.

Artificial Intelligence, Chatbot, Clova, Conversation analysis, Foreign language learning