2024 Regular and Panel Sessions

Thursday, May 23

Session One

Navigating the Ethical Landscape of ChatGPT and Fostering Critical Thinking and Digital Citizenship in Language Teaching and Learning
Colum Yip, Sunao Fukunaga

This action research examines the ethical use of ChatGPT in writing instruction at a Japanese university. We held an 11-week workshop on critical digital literacy (Bacalja et al., 2021), academic integrity, and the utilization of ChatGPT as a writing aid. Our quantitative and qualitative results indicate that students were satisfied with the grammar and audience awareness feedback provided by ChatGPT but found the content feedback to be inconsistent. Overall, students felt confident in their ability to use ChatGPT ethically but emphasized a critical need for more training in prompt engineering.

L2 Incidental Acquisition of Spanish Lexical Items through Engagement with Dynamic and Static Images and Captions on the Rockalingua Language Platform
Tetiana Tytko

This paper investigates the influence of captions on young learners’ L2 vocabulary acquisition in the music-based language learning platform Rockalingua. Specifically, it examines how Rockalingua utilizes bimodal multimedia input for better learning and retention of high-frequency target vocabulary structures in Spanish. Drawing on SLA theories of multimedia language learning along with the fuzzy lexical representations hypothesis, this paper observes the effect of two experimental conditions, dynamic captions and images in the music videos (DYNCON) and static captions in the worksheets with song lyrics and still images (STACAP), on immediate and delayed vocabulary gains measured by productive and receptive tests.

“But I’m a good employee!” A Social-deduction, Company-crisis Game in an L2 Chinese Online Synchronous Program
WeiHsuan Lo, Kevin Fedewa

This presentation describes the design, implementation, and evaluation of a social deduction game themed on a company crisis and used in an online-synchronous Chinese language program for novice and intermediate learners. The presenters describe the game design that adopted Fullerton’s (2018) framework: (a) conceptualization, (b) (digital) prototyping, (c) playtesting, (d) functionality, completeness, and balance, (e) fun and accessibility. Presenters will also report the results of student’s self-evaluation surveys and focus group interviews, with special attention to perceptions of 21st-century 4C skills learning and language skills development.

Looking into Learner Satisfaction in a Massive Open Online Language Course (LMOOC)
Ana Gimeno

Massive Open Online Courses (MOOCs) are still a growing phenomenon and rely on technology to reach out to potential learners in both academic and industrial sectors. One of the most prominent MOOC providers alone, edX.org, currently has over 4000 MOOCs on offer, with 993 results for the query “industry”. Higher education, in particular, has embraced this education “outlet” as a way to cater for an increasing demand for high quality online course materials to cover the needs of professionals who would like to engage in lifelong learning, and to satisfy the need to be at the forefront of educational developments and gain more international visibility. MOOCs, in this sense, are one of the areas of confluence between the industrial sector and academia. This presentation will report on the findings and conclusions deriving from over 21,500 learner responses from enrolees on four upper-intermediate English language MOOCs to a survey conducted longitudinally over a period of five years to shed light on some of the factors involved in learner motivation, expectations and learning styles.

Developing a Teacher Supporting AI System: 2xAI (eXplainable & eXchangeable)
Yohan Hwang, Sangmin-Michelle Lee, Sunjoo Hong, Yeonjeong Park

The objective of this study is to develop a data-driven teacher support AI system for use in educational settings, while also validating its usability and effectiveness through real-world classroom applications. AI has the potential to serve as a valuable tool in realizing personalized and tailored education, thereby supporting learners’ learning processes and assisting classroom teachers. However, data-driven teacher support AI systems remain scarce in the public education sector. Therefore, in this study, we aspire to create a teacher support AI system that aids teachers in designing lessons by performing the following tasks: 1) Analyzing learner data, 2) Predicting academic performance, 3) Grouping learners based on their characteristics and needs, and 4) Curating learning materials.

Defining Language Teachers’ Readiness for Online Language Teaching: Toward a Unified Framework
Luca Giupponi, Emily Heidrich Uebel

While the COVID-19 pandemic forced a large number of instructors to familiarize themselves with online learning technologies, it cannot be assumed that Emergency Remote Teaching (ERT) experience can translate to purposefully planned Online Language Teaching (OLT) proficiency (Gacs et al., 2020). This presentation reports on efforts to identify and define key competencies for OLT, conceptualize a set of OLT readiness can-do statements, draft a practitioner-facing self-assessment framework, and conduct an expert domain validation study conducted with 26 CALL experts based in the USA and Europe.

Session One Panels

Research Methods in Virtual Reality, Virtual Exchange and Gaming
Hiba Ibrahim, Tricia Thrasher, Carla Consolini, Mery Diez-Ortega

Conducting research in CALL environments (e.g., Digital Games, Virtual Reality) presents numerous challenges, particularly when it comes to data analysis. This panel will advance efforts in this area by presenting 4 examples of different methodological approaches that could be adopted by researchers: researcher as mentor, analysis of interactional data from digital games, recording and analysis of speech data in virtual reality, and assessment of vocabulary learning in virtual environments. Panelists will walk attendees through how these methods have been used in their own studies with the aim of providing step-by-step guidelines that graduate students and early researchers can follow in their own work.

CALL and the Learner Experience – Lessons from Multicultural Contexts
Hope Anderson, Deniz Gokcora, Farah Akbar, Liss Kerstin Sylvén

For continued language study and promoting interest in languages and cultures, a focus on the learner experience in language courses is paramount. CALL can contribute to this positive experience if used in helpful ways that correspond with what we know about second language acquisition, technological affordances, and learner preferences and needs. The panel includes perspectives from ESL virtual exchange, Dörnyei’s L2 motivational self to compare in-class with extracurricular experiences, gamification in online learning environments, bridging students’ personal lives with an English linguistics class, and language teacher/learner customer discovery in an entrepreneurship setting. Panelists’ findings demonstrate effectiveness of tools, genres, styles, and settings, providing needed lessons for future L2 educational opportunities.

Session Two

Implementing AI Chatbots in L2 University Courses: Building Pragmatic Competence and Critical Digital Literacy
Natalie Amgott, Devon Renfroe

Despite the “powerful learning potential” of AI chatbots for facilitating pragmatic competence (Godwin-Jones, 2023, p. 8), little literature has provided pedagogical approaches for integrating current iterations of AI chatbots with L2 students. In this pedagogical presentation, we review the affordances and constraints of three AI chatbots (ChatGPT, Google Bard, and Lingostar) before providing four illustrative lesson plans for intermediate L2 learners to navigate pragmatics and critical digital literacy in university French and Korean courses. The presentation concludes with a discussion of ethical use of AI chatbots for L2 conversational practice.

A Comparative Study of AI-Generated and Human-Written Structured Research Abstracts
Erik Voss

GenerativeAI (GenAI) tools have the potential to help humans write more efficiently and be more productive. However, we don’t know the extent to which current GenAI models are capable of producing texts that replicate authentic human academic writing. This study will explore the extent to which AI-generated research abstracts are comparable to abstracts written by humans following Swales & Feak’s (2009) communicative stages. AI-generated abstracts became more similar to human written abstracts with specific prompting and providing example abstracts; yet, structural and linguistic differences remained. Findings have implications for the use of AI-generated text to assist humans with academic research writing.

Competitive Collaborative Games: Project-Based Learning through Gamification
Mahmoud Amer

Gamified learning allows educators and learners alike to embrace and envision achievement beyond traditional tests and quizzes. Gamified learning is limitless as it is applicable to a variety of fields and learning contexts, and virtually has universal appeal and transcends cultures and languages. The next generation of learners are reward-centric (i.e., likes, upvotes, badges) within a context of multiple connected networks (i.e., sharing across a variety of social media sites, and remixing content into something new as in Tiktoks and YouTube Shorts). This presentation will demonstrate the application of this gamification model using a variety of open-source and experimental tools.

Supporting Multilingual Learners: Exploring the Impact of High-Immersive VR on College Students’ Foreign Language Anxiety
Eunkyoung Cha, Zheying Zhu

This study investigates the potential of high-immersive virtual reality (VR) to reduce the foreign language anxiety of international college students. Although existing research highlights the effectiveness of VR in language learning, little attention has been given to its impact on the anxiety levels of language learners while interacting with native speakers. Our quasi-experimental study involves college students in language learning courses, utilizing high-immersive VR. Participants’ self-reported FLA will be collected, with the hypothesis that high-immersive VR can effectively mitigate FLA. This research aims to contribute valuable insights into enhancing language learning experiences through immersive technology, with data analysis to follow.

Designing Game-based Augmented Reality Scenarios for Second Foreign Language Learning: Sharing Experiences from Four European Funded Projects
Stella Hadjistassou, Shaunna Joannidou, Pablo Molina

Within the framework of design-based research, this presentation draws on four European projects, ReDesign, DiMPAH, DIMPE, and ARIDLL funded in the last eight years to examine some of the underlying iterative processes involved in the design, development, and deployment of AR game-based experiences for second and foreign language learning, interaction, and collaboration. Taking an interdisciplinary transnational approach, this presentation addresses the following research questions: (1) what are some of the iterative processes involved in the design, development, and deployment of AR game-based applications for language learning? (2) what are some of the affordances, limitations, and contradictions emerging in this process?

Session Three

Automated Evaluation of Argumentative Writing Using Fine-Tuned GPT 3.5 and GPT 4 Models
John Gayed, Qiao Judy Wang

Automated writing evaluation (AWE) of argumentative essays is complex and traditional NLP tools which rely on lexico-grammatical features, often struggle to capture writing quality nuances in their assessments. However, the rise of large language models (LLMs) signals a shift towards more sophisticated qualitative text analysis. This research explores LLMs’ capabilities in AWE using the ETS TOEFL Independent Writing Task dataset. An optimized GPT 3.5 model achieved an overall accuracy of 72.5%, while an untrained GPT 4 model on the same prompts and samples yielded an overall accuracy of 65.8%. These results underscore LLMs’ potential in offering high-quality qualitative writing assessment.

CSU + MeTabi: A Model for Making Public-Private Partnerships Work for Higher Ed
Alyssia Miller de Rutte, Scott Payne, Nick Gossett

In this presentation, we explore the innovative partnership between Colorado State University’s Department of Languages and Literatures and MeTabi, Inc., developers of a cutting-edge language learning platform. This collaboration focuses on tailored language learning for specific purposes, enhancing educational opportunities. We will outline the partnership’s objectives, structure, and the mutual benefits for both entities. Key topics include funding strategies, intellectual property rights, and the expansion of research opportunities for faculty and students. Join us to discover how this partnership is shaping the future of language education.

Multimodal Grammar Learning through Animal Crossing, A Commercial Game in Japanese
Kayo Shintaku

This study explores learning Japanese grammar—giving and receiving (ageru, morau, kureru)— using Animal Crossing Pocket Camp, a commercial mobile game. For two 3-week phases, Japanese-as-a-foreign-language (JFL) students played the game, wrote reflections, and completed surveys during outside-of-class time while they were taught target grammar in class. The analysis showed participants’ appreciation for gameplay as they learned grammar in a different context, recognizing the link between in-class and outside-of-class learning. The result also illustrated that the game helped players understand grammar through character interactions and visual cues and encounter literacy realities, providing a sense of achievement for understanding authentic materials.

Using Immersive Virtual Reality to Practice Public Speaking Skills in a Second/Foreign Language (L2)
Naiyi Fincham, Leeseul Park

Immersive virtual reality (IVR) holds promising potential for facilitating the development of L2 speaking skills by engaging world language learners in simulated real-life scenarios. This presentation reports on an ongoing project that aims at identifying and leveraging the affordances of existing commercial IVR applications in fostering L2 speaking competence. Presenters will introduce how an IVR application is integrated into a pre-capstone Korean course to help students practice public speaking skills, discuss how app-generated data is used to monitor students’ speech performance and provide individualized feedback, and share effective strategies for teachers to support and guide students throughout this experience.

Creating Targeted Technology Standards: Supporting Teachers in the Canadian Settlement Language Sector
Phil Hubbard, Greg Kessler

We report on the process of developing a set of technology standards specifically for English language teachers in the Canadian settlement language sector and on lessons learned from that experience. The resulting seven standards and 41 performance indicators are written in direct, teacher-friendly language and include explanations and reflection questions for each performance indicator together with vignettes describing how the standards are realized in authentic teaching situations. We discuss reactions from teachers and teacher educators and progress on parallel standards for programs and learners. The standards are available to the audience and others under a CC-BY, non-commercial, share alike license.

Learning L2 Vocabulary through Kahoot and Utilizing Multimodal Approaches
Olesia Pavlenko

This study addresses the underexplored effects of game-based learning via Kahoot on L2 vocabulary acquisition, with a focus on delayed effects and varied platform functionalities. Data from 63 English language learners (ELLs) were collected, categorized into three learning conditions (words+translations+gifs, words+translations+static pictures, and words+translations with no pictures). The study revealed statistically significant vocabulary gains across all conditions, with consistent retention over time. While Kahoot proved effective, no statistically significant differences were observed between the conditions. This research highlights the potential of Kahoot in L2 vocabulary learning, emphasizing the need for further exploration and scrutiny of different learning modalities.

Tales with Tails: AI Image Generation to Enhance for Storytelling
Seoah Yun, Eda Yildirimer

As technology rapidly evolves, one factor instructors must consider is the cost of classroom tools. This proposal focuses on how the free AI image generator Kandinsky 2.2 may enhance the creation of collaborative stories. The presenters will demonstrate the generator and its use in a 5-day writing and AI illustration lesson.

Panel Presentation
Bridging the Gap: Transferable Research Skills Between Industry and Academia
Lillian Jones, Natalie Amgott, Tricia Thrasher, Lisa Frumkes

This panel, hosted by the Graduate Student SIG delves into the crucial topic of transferable skills between industry and academia. Specifically, we aim to help graduate students gain awareness of research skills and strategies acquired in industry/academia, and how they can be applied to a variety of situations. Panelists will discuss strategies such as task/project management, goal setting, and performance assessment, emphasizing data-informed decision-making, effective communication of these skills in resumes and job applications, and application between academia and industry.

Session Four

“Am I teaching my students to cheat?” Teacher Reflection Incorporating ChatGPT in College Level Writing
Farhana Ahmed

This presentation emerged from the ethical questions and concerns that teachers, researchers, and policy makers today have about the integration of ChatGPT into educational settings. The research investigates teachers’ reflections on incorporating ChatGPT in lessons, aiming to understand how AI can offer unique opportunities to enhance students’ critical thinking and writing instruction. By exploring ChatGPT’s use in the writing curriculum, the talk provides valuable insights for educators, curriculum designers, and policymakers seeking to enrich the educational experience for learners. This study paves the way for innovative and effective approaches to a new era of AI-powered writing instruction.

Generative AI for Supporting Recently Immigrated English Learners’ L2 Lexical Development: Linguistic Alignment, Lexical Complexity, and Learner Perceptions
Daniel Dixon, YouJin Kim, Youngmeen Kim, Robin Cathey

In this presentation, we discuss the affordances and effects of AI-generated chatbot-based interaction on the language production of L2 English learners who recently immigrated to the US. Using Python and OpenAI’s (2023) GPT 3.5 API, our research team developed four interactive conversational tasks deemed important for this learner population through a needs analysis. Data collected during chatbot-learner interaction assesses learners’ written production in terms of lexical complexity and linguistic alignment. Learner perceptions towards carrying out chatbot-assisted collaborative tasks was examined through surveys and interviews. Implications on using generative AI for L2 learning and future research directions are discussed.

Large Scale Experimentation of a Serious Game Targeting Oral Comprehension in English as a Second Language for 6–8 Year-olds: Some Lessons Learned
Mathieu Loiseau, Émilie Magnat, Audrey Serna, Anne-Laure Guinet, Marie-Pierre Jouannaud, Coralie Payre-Ficout

Between 2018 and 2020, the Fluence project tested three learning apps with 700+ pupils for more than two years. The protocol neutralized the novelty effect by having concurrent apps targeting different skills administered to each group. In this symmetrical protocol, the test-group for app1 is the control group for app2 and vice-versa. One of the three application of the Fluence project is Luciole an EFL learning game for 6-8 year-olds. The pupils using Luciole significantly improved their English oral comprehension (Mandin et al. 2021) compared to the other groups. In this communication we delve deeper in the explanation of the results.

Potential of Immersive Social VR for Developing Interaction Abilities in L2 Spanish
Naoko Taguchi, Elizabeth Hanks

This study assessed the effectiveness of a two-week long instructional intervention using social VR. Twenty-six beginning-level Spanish learners completed four sessions of task-based interaction using the social VR application Altspace. They formed small groups and interacted as if they met each other for the first time. The learners’ interaction abilities were compared between the first and fourth (last) VR session. Results revealed significant gains in their interaction abilities with a large effect size (Cohen’s d of 1.81). Specifically, they became able to sustain a conversation for an extended time using interactional resources such as backchanneling, topic nominations, and follow-up questions.

Online Language Course Development and Teaching
Nevine Abraham, Sue-mei Wu

Developing and teaching an online language involves an array of processes, such as implementing innovative pedagogical approaches, developing digital content, considering the learners’ psychology, and adapting to available learning management systems, to name a few. This session invites proposals that explore online language teaching and learning, including less-commonly taught languages, and that may include: – Innovative pedagogical approaches – Empirical research and data-based experimental work – Practice-based reports – Cultural literacy – Digital and cultural audio and video components – Available tools and limitations of learning management systems – Challenges to less-commonly taught languages online – Assessment and evaluation.

Unleashing the Potential of Chat GPT in Curriculum Development and Lesson Planning
Valentina Lukin, Maryam Sahebfosul

This presentation delves into the transformative role of AI, notably ChatGPT 3.5, in language education. Exploring the benefits of AI integration, with a spotlight on lesson plan development, we showcase examples from two critical languages, Farsi and Russian, and less-commonly taught ones like Czech and Slovak. Unveiling the specific commands employed for educational material creation on ChatGPT 3.5, we address language and non-language-specific challenges and innovative solutions. Join us to unravel the potential of AI in meeting the dynamic needs of learners and educators.

Session Five

Igniting the Engine: Leveraging Tactical, Human-mediated Support to Drive Engagement and Outcomes with Technology-mediated Language Learning
Katie Brown, Mackenzie Campbell

Well-designed language learning programs will not deliver outcomes if learners do not use them. For decades, studies of self-paced language learning have grappled with both inconsistent learner engagement and inconsistent learning outcomes. While the personalization of learning experiences in online programs can help drive learner engagement, it is the blending of human support with this personalized instruction that leads to sustained learner persistence. This session will discuss the impact of a large-scale implementation (N = 20,000) of human intervention in a technology-mediated English learning program, including multilingual learner success coaching, targeted outreach, live online workshops, and on-demand small group classes.

Evaluating AI for Integration in Language Teaching across Contexts: A Conceptual Framework
Jeffrey Maloney, Austin Pack

In this presentation, we present a framework for use in evaluating AI tools across teaching contexts. This framework builds on previous work and other frameworks that have been developed for using technology in education (e.g., Hubbard, 1998; 2011; Chun et al., 2016; Rosell-Aguilar, 2017). However, with the rise of novel AI-powered technologies, additional approaches and considerations may be necessary. As has been highlighted in the research literature, differences in perspectives on the use of such tools may exist across stakeholders including students, teachers, and administrations. Next, we explore how this framework can be adapted to include frameworks of knowledge construction.

Integrating Behavior Analysis into the Design of a Serious Game for English Education?
Mathieu Loiseau, Anne-Laure Guinet, Audrey Serna, Émilie Magnat, Marie-Pierre Jouannaud

Luciole is a serious game designed with two main objectives: introducing 6-8 year-olds to oral comprehension in English as a foreign language and allowing them to work autonomously to support differentiation in the classroom. It was developed using a Design-Based Research method through several iterative cycles: design, development, expert & user tests, and log analysis. Language specialists, computer scientists, as well as teachers and elementary school pupils have been involved in a user-centered design process. In this communication, we present the hybridcooperative design framework set up to improve thea serious game based on the results of a first large-scale study.

Thinking about One’s Own Speaking: The Impact of Metacognitive Reflections on L2 Spanish Conversational Skills in Face-to-face and Online Settings
Luis Cerezo, Gorky Cruz, Cristina Hernandez

The CLEAR Chats project seeks to help L2 Spanish learners develop their oral proficiency by using metacognitive reflection. Learners meet individually with coaches to complete oral tasks, record their conversations, and compose metacognitive reports in which they reflect upon their linguistic self-competence and conversational strategies, getting assessed on the quality of their reflections rather than their oral proficiency. In a previous investigation we found that learners reported higher amounts of self-competence and conversational strategies when they conversed online than face-to-face. This study expands on our previous research by refining our coding scheme, increasing our sample size, and integrating survey-based data.

Improving Reading Comprehension using Cascade Reading: A Linguistically-driven Text Formatting Algorithm
Jack Dempsey, Kiel Christianson, Julie Van Dyke

Cascade Reading’s Linguistically-Driven Text Formatting (LDTF) algorithm produces a new text format that cues sentence structure via line breaks and indentations during reading. The format produced increased comprehension in elementary-age readers (Van Dyke & Dempsey, submitted); however, effects on experienced readers have not been investigated. The format’s emphasis on syntax could be particularly meaningful for ELLs and for producing ESL curricula. The current study investigates 90 adults (30 L1 English, 30 L1 Mandarin, 30 L1 Korean) reading in traditional or LDTF format. The Cascade Reading algorithm improved reading comprehension for all groups, especially for L2 English readers.

Panel Presentations

Building Bridges: Academic Insights into Educational Technology
Sebastien Dubreil, Bryan Smith, Drew Davidson, Nicole Mills

In the rapidly evolving landscape of computer-assisted language learning (CALL), the intersection of academia and industry should play a pivotal role in shaping the future of language education. This panel, “Building Bridges: Academic Insights into Educational Technology,” delves into the multifaceted relationship between academic research and practical applications within the realm of educational technology.

Our panelists – who draw from diverse academic backgrounds to play pivotal roles advising industry partners – will illuminate the critical role of academics in driving innovation and fostering collaboration in the development of effective educational technologies for language learning. Through a series of short, insightful presentations and engaging discussions, attendees will gain valuable insights into the latest research findings, emerging trends, and best practices in the field of CALL as it is implemented in industry..

Topics of discussion will include: starting conversations with industry partners, leveraging academic expertise to inform the design and implementation of technology-enhanced language learning tools, navigating the challenges of integrating research findings into real-world applications, and fostering meaningful partnerships between academia and industry to address the evolving needs of language learners in today’s digital age.

Join us as we explore the transformative potential of academic insights in advancing educational technology, and discover how collaboration between researchers and practitioners can bridge the gap between theory and practice in computer-assisted language learning.

Introducing a Paradigm Shift to Digital First Test Development
Alina von Davier, Geoff LaFlair, Andrew Runge, Yigal Attali

In this panel, we will discuss a new framework for the development and management of the item development stage for a high volume, high-stakes (English) language test. We refer to this framework as an item factory, a system where human expertise and AI prowess are combined for an efficient and high-quality test development process. We integrate item and test design and subsequent piloting activities into this framework and discuss how these activities could be made more efficient. The item factory framework relies on automation, AI, and engineering principles to make the item development process scalable for high-volume, high-stakes testing in the digital era. This framework can be applied to tests for other languages.  We are also presenting the human-centered AI framework that supports and regulates the item factory.  Human-Centered AI recognizes the importance of involving educators, experts, and students in the development process, ensuring that AI systems enhance, rather than replace, the role of teachers and developers. It also places a strong emphasis on transparency, fairness, and ethical considerations in educational AI. In this panel we will illustrate the application of these principles with specific examples of interactive items for reading, listening and writing from the Duolingo English Test. Each panelist will present the perspective of a different expertise involved in this new test development process: applied linguists, language testing scientists, AI engineers, psychometricians.

Session Six

Predictors of Attrition in a Longitudinal Study of App-based Language Learning
Ekaterina Sudina, Luke Plonsky, Yasser Teimouri

We present a longitudinal study to examine multiple individual differences as predictors of attrition among app-based language learners. A total of 601 beginners learning Spanish or French (314 attritors, 287 non-attritors) participated in a natural experiment investigating Duolingo app usage and proficiency gains over a 6-month period. Our results show L2 grit perseverance of effort and age as two meaningful predictors of attrition in app usage. Critically, we found that being older and having higher perseverance decreases the probability of language-app attrition. We conclude with implications for the conceptualization of grit in the context of app-based language learning and beyond.

Generative AI – Friend or Foe of Language Education?
Mathias Schulze

Intelligent CALL (ICALL) – the application of artificial-intelligence techniques and technologies to CALL – has its roots in the early phase of CALL. Early ICALL systems focused primarily on rule-based approaches, using expert systems and rule-based reasoning to provide corrective feedback to language learners (see Heift & Schulze, 2007). The most recent advances in AI have led to immense improvements in the quality of the output of machine translation and text generation systems, both of which work also for many less commonly taught languages. In this presentation, we will briefly review the AI technologies underlying three tools that are used frequently (not only) by language learners: Google Translate, DeepL, and ChatGPT, and their fruitful integration in language-learning processes will be discussed.

Using Data Analytics to Explore the Impact of a Task-based Digital Game on Learners’ Affect
Frederick Poole, Jason Kappes

In this presentation we build on calls to utilize data analytics to explore language learners’ experiences while playing digital games (Poole & Clarke-Midura, 2023). Past research on digital games has rarely leveraged such data to make connections between gameplay and outcomes. We present data from an indie game called Newcomer that illustrates how log files can be used to provide evidence of player behavioral patterns which subsequently can be used to demonstrate the effectiveness (or lack thereof) of game design features on learners affect (e.g., perceptions of learning, perceived difficulty, and interest).

Augmented Reality Design and Development for Cognitive Academic Language Proficiency (CALP) and Intercultural Communication Competence (ICC)
Shaunna Joannidou, Stella Hadjistassou, Pedro Molina

The presentation will discuss the design and development of Augmented Reality applications for the European Erasmus+ project Digitalisation of Multilingual Programs in EHEA (DiMPE) . This pan-European project involves 7 partners: 2 enterprises and 5 Higher Education Institutions (HEIs) from 6 different EU Member States, namely, Spain, Poland, Norway, Finland, Cyprus and Ireland. This presentation will discuss the bottom-up approach established for the design and development of 6 AR scenarios including the necessary technical and educational phases of this work in order to support Cognitive Academic Language Proficiency (CALP) and Intercultural Communication Competence (ICC).

The Potential of Plurilingual Audiovisual Input for Simultaneous Language Learning
Anastasia Pattemore

While audiovisual input has been proven beneficial for L2 language development, original version television has not been explored from the newcomer’s language learning perspective, when the viewer is exposed to L2 audio (mainly English) and L3 subtitles (the language of their new country), i.e., plurilingual subtitled input. This study explores the effects of prolonged viewing of plurilingual subtitled input (TV series) and measures the simultaneous acquisition of L2 English and L3 Dutch multiword units in newcomers to the Netherlands. The results will be discussed in light of the multilingual approach to language learning.

Mediating ESL Learners’ Use of Digital Reference Tools to Deepen Their Vocabulary Knowledge in Academic Writing
Yuwei Xia, Minjin Kim

Previous literature on digital reference tools has established the significance of training for the effective use of the tools. This study investigates the effect of three 30-minute training sessions of digital reference tools to deepen learners’ vocabulary knowledge. Four vocabulary items are included in every session with graduate level ESL learners for academic writing and mediation is directed to support learners to navigate through various aspects of the tools and build a more comprehensive understanding of vocabulary knowledge in terms of form, meaning, and use. Both quantitative and qualitative data are analyzed and interpreted to establish a possible training protocol.

Session Seven

Language in Digital Spaces: Contemporary Contexts for Cultivating Access
Stephanie Knight, Casey Upson

Eighteen US states have enacted divisive concepts legislation, or legislation limiting the teaching of certain topics pertaining to race and gender in public education. In light of this policy wave, language practitioners must rethink language pedagogies and the ways that those pedagogies are enacted in digital spaces. In service of this goal, this presentation provides data from an existing digital discourse analysis examining the framing of the policy and offers specific pedagogical practices (e.g., concept-based pragmatics instruction) that can be utilized by educators to ensure learner access to critical cultural content in areas where divisive concepts legislation is law.

Creative Writers’ Proactive Use of Generative AI and Its Implications in Language Education
Yu Jung Han

This presentation shares an exemplary use of Generative AI (GenAI) from a Reddit community of creative writers who have already used multiple GenAI applications even before the launch of ChatGPT. They have developed proactive ways to use GenAI as their co-writer/thinking partner while maintaining full control over their creativity and authorship. This presentation (1) provides exemplary usage and characteristics of using GenAI to augment and enhance writing, not to replace it, while maintaining a sense of control and ownership; and (2) discusses how to add autonomy/authorship/control to the current pedagogy to introduce GenAI into English writing.

Leveraging Internal and External Funding from Idea Conception to Commercialization
Stephanie Link, Robert Redmon

This presentation focuses on how our team obtained seed funding that turned into a major NSF grant, propelling our AI-based project in automated writing evaluation forward from conception to commercialization. The web-based tool called Dissemity for “disseminating research with clarity” leverages NLP and LLMs to enhance writing for publication. Development started with 133 potential user interviews and continued by marshaling transdisciplinary expertise to produce a scalable product. Dissemity has been integrated into programming at Oklahoma State University before commercial partners were formed to expand off campus, with the latest partner being the University of Houston.

Investigating the Impact of Group Video Projects on Investment in Language Learning: A Comparative Study
Francesca Marino

As digital genres have become increasingly integrated into learners’ lives, various digital multimodal composing (DMC) tasks (e.g., posters, blogs, video projects) have found their way into second language (L2) classrooms. While prior research has explored DMC’s impact on learner motivation, its influence on learner investment (Norton, 1995) remains under-investigated. This case study explores the impact of group video projects on investment among L2 Italian learners, comparing them to learners working on collaborative monomodal essays. The analysis of students’ artifacts, surveys, and interviews reveals differences and commonalities in learners’ investment, including commitment and identities, acquired skills, and challenges encountered.

American or British Accents: Students’ Attitudes in a Web-based Platform
Abdurrazzag Alghammas

This study employed the interaction hypothesis (Long, 1983) to investigate attitudes towards different English accents (i.e., American and British) of 40 male undergraduate EFL students majoring in English. It explored the reasons for such views, as well as identifying the accent the participants found most effective for communication. The study also examined students’ attitudes to online speaking by means of a Synchronous Computer-Mediated Communication (SCMC) website known as ‘Cambly’. Students were granted free access to the Cambly website for live interaction with Native English Speakers (NES). Each student talked for 15 minutes with an American and British interlocutor, enabling the researcher to recognize the common topics appearing within these conversations. Data were collected using a mixed-method approach, employing a web-based survey of closed and open-ended questions, alongside the recorded conversations. The key findings reveal that students enjoyed the SCMC conversation and also found it beneficial for improving their speaking skills. Furthermore, SCMC allowed students to choose the topic and negotiate meaning with native speakers during a lengthy conversation. This study establishes that students preferred American to British accents and felt more confident in understanding American speakers. The study concludes by highlighting the practical implications for teaching speaking skills, also suggesting new directions for future research.

Self-directed Learning of Language and Culture in Spanish Content-based MOOCs
Tanya Tercero

The presenter discusses her own self-directed language and culture learning in three Spanish content-based MOOCs (Massive Open Online Courses) on Flamenco, Federico Garcia Lorca, and the Sierra Nevada, through the University of Granada, Spain. An experienced online course designer and instructor, the presenter discusses affordances and critiques of the MOOCs’ course design, content, and delivery. She also offers suggestions on how language instructors might utilize content-based MOOCs as supplemental teaching resources, as well as how other self-directed lifelong learners might participate in content-based MOOCs for language, history, and culture.

Panel Presentations

AI Generated Content and Second Language Teaching
Feng Xiao, Ting Su, Gang Liu, Haixia Wang, Ding Wang Bramlett, Jonathan Becker

This panel discuss how AI generated content (AIGC) can be leveraged into second language (L2) classroom teaching, aiming to address two questions: What is the status quo of using AI in foreign language teaching, and what are the advantages and challenges of using AIGC tools such as ChatGPT in L2 classroom teaching? The panel will include four presentations. The first presentation will give a review of existing AI programs that have been used in foreign language teaching. The following three presentations will use L2 Chinese as an example to reveal how AIGC can be used in teaching linguistic and cultural knowledge and enhancing the quality of classroom activities through human-computer interactions.

Crossing Boundaries, Driving Innovation: Making Inroads in the EdTech Industry
Sebastien Dubreil, Lisa Frumkes, Joan Bajorek, J. Scott Payne

This panel brings together esteemed professionals who have successfully transitioned from academia to longstanding careers in the educational technology (EdTech) sector, with a particular focus on technology-mediated language learning. Through their diverse academic backgrounds and extensive industry experience, panelists will share their journeys, insights, and strategies for driving innovation in EdTech.

Through a series of short, insightful presentations and engaging discussions, attendees will gain invaluable perspectives on navigating the transition from academia to the EdTech industry, leveraging academic research in developing effective language learning solutions, and overcoming challenges in product development and implementation. The panel will highlight the importance of interdisciplinary collaboration, the integration of pedagogical principles with technological advancements, and the role of continuous learning and adaptation in staying ahead in the dynamic landscape of EdTech.

Session Eight

Using and Evaluating a Language Learning App: A Corpus-Driven, Autoethnographic Analysis
Lawrence Williams

This presentation offers a corpus-driven analysis of the content and task types available in Duolingo for learners of Czech as part of a language-learning autoethnography. Even though the close analysis focuses on Czech, insights and observations from other languages (Arabic, French, and Italian) are used for the purpose of comparison and contrast. The following three-part question guided this project: How can Duolingo be used by different types of learners in different educational contexts and for different purposes?

Global AI Language Learning at Scale: From Academia to Industry
Pamela Bogart

Generative AI has revolutionized the potential language learning opportunities available to individuals across the world, by providing exposure to and practice using target languages. This presentation describes the development of a global MOOC designed to equip any language learner of any language at any level to boost their learning capacity by strategically employing generative AI tools. This online course, created by academics specializing in how language learning works and emerging AI technologies, is designed for a platform that is predominantly used by global learners in careers in business, so a key focus is on multilingual communication in industry.

Decoding Student Writing: A Machine Learning Approach to Analyzing Student Texts
Larissa Goulart

It is not uncommon for writing instructors to ask students to write an informational essay and, in turn, receive an argumentative essay (or vice-versa). In this presentation, we will introduce an online interface that writing instructors can use to give students’ feedback. This tool draws on the results of a Machine Learning (ML) model to classify students’ assignments for communicative purpose and discipline. We envision this tool being used in the classroom to highlight the differences in communicative purpose across written assignments.

Building Bridges for Innovation
Julio Rodriguez, Naiyi Fincham, Suzanne Freynik, Molly Godwin-Jones, Rachel Hernandez, Leeseul Park

The Language Flagship Technology Innovation Center (Tech Center) collaborates with academic programs nationwide and various entrepreneurs to systematically integrate technology into every facet of the Flagship experience. Following Tech Center guidelines, the resulting resources and processes are open and shared widely with the intent to benefit the broader language education community. Join us to explore several Tech Center projects, such as a multi-institutional initiative to facilitate the acquisition of sociopragmatic knowledge; professional development modules geared toward increasing language teachers’ repertoire of strategies for teaching toward high levels of language proficiency; and collaborations with the private sector to discover promising language technology innovations.

Integrating Scenario-Based Design and Keystroke Analysis in Written Argumentation Assessment
Mahshad Davoodifard

This study examines the feasibility of a scenario-based assessment (SBA) and detailed keystroke log analysis to assess ESL/EFL learners\’ written argumentation ability. Utilizing machine learning techniques, the research proposes a method to identify patterns in test takers\’ responses and processes, enabling a task-specific assessment during SBA and a nuanced comprehension of factors influencing argumentative essay quality through process data analysis. The findings can be used to identify profiles of language learners, having significant implications for improving L2 written argumentation instruction and assessment.

The Efficacy of Technology-Mediated Task-based Pronunciation Instruction
Linh Phung

Learners’ substantial interest in pronunciation instruction (PI) has highlighted the need for effective PI techniques in instructed second language acquisition (ISLA) to inform materials, classroom activities, and technological applications. This study examines the effect of technology-mediated task-based instruction on the acquisition of the English /ɹ/ and /l/ sounds among Japanese students. Divided into app-based, pen and paper, and control conditions, participants complete in five picture-based tasks as well as forced-identification and control-response tasks in a pretest, posttest, and delayed posttest. With data collected from 150 students, this presentation will share the study’s key findings and implications for PI.

Session Nine

Objective and Perceived Insights into Digital Activity Difficulty
Matthew Pattemore

Many digital language learning tools, like games, employ a range of activities (fill-the-blanks, matching, multichoice, etc.), but there is surprisingly little evidence regarding the relative difficulties and effectiveness of the different activity forms. Using logdata and questionnaires from 1307 Spanish learners playing digital minigames to develop their L1 and L2 literacy, and using an item response theory analysis, we establish an objective measurement of relative activity difficulty. This presentation addresses which activity types are harder/easier for these learners both objectively and subjectively, as well as discussing which features (e.g. time pressure, answer visibility) may be influencing our results.

The Effect of Immersive Conversational AI on Immigrant Students’ L2 Oral Proficiency
Uliana Ovsiannikova

Following the call for research on immersive AI, this presentation will explore a mixed-methods study on integrating AI into ESL curriculum at an underserved high school in Illinois. Conducted over one academic quarter, the study involved 44 participants from immigrant youth population. The research aimed to assess students’ conversational strategies with AI and evaluate their improvements in willingness to communicate (WTC), fluency, and comprehensibility. Students’ interactions with AI, on desktop and in VR, were observed and analyzed. During this talk, the presenter will share the findings from the study and discuss the potential role of AI in supporting immigrant youth.

Conceptualizing One’s Own Self and Cultural Identity(-ies) in Virtual Exchange
Maria Speggiorin

Virtual Exchange (VE) has long been a focus of research regarding language learning and intercultural competence. However, research on the latter has majorly focused on the interaction with and conceptualization of the “other,” rather than on students’ articulation of self and of their own cultural identity. Based on a set of videoconferencing data collected during a VE between German learners in the USA and English learners in Germany, the present study analyzes US students’ construction, conceptualization, and representation of their own “self,” in terms of their belonging to different cultural and social communities.

Please don’t hate my story!: Engagement in Harry Potter Fanfiction Author Notes
Maria Razcon

This presentation showcases a rhetorical moves analysis (Tardy 2015) on 30 Author Notes found in Harry Potter fanfiction texts in Wattpad. Patterns on these notes are analyzed and labeled based on their purpose or goal, in order to show the resources used by fanfiction authors to enable (or prevent) engagement among their readers and fan-communities (Herzog, 2012; Luzon, 2020). Findings from this analysis can inform digital and multimodal genre analysis as well as our understanding of language used by fans in the “Digital Wilds” to enable fan-affiliations and community engagement.

TBLT in Chinese and Korean Immersive Programs: A Computer-Mediated- Communication Task’s Design, Implementation, and Evaluation
Kevin Fedewa, WeiHsuan Lo, Jing Paul, Adelia Kim, Changu Kang, Nikeese Daniel

This presentation describes the design, implementation, and evaluation of a Computer-Mediated-Communication (CMC) task used across two immersion programs (one Chinese and the other Korean) whose curriculums both centered on the common tasks of US Foreign Service Officers. The presenters follow the six components of Long’s (2014) TBLT curriculum design to first present how (a) a needs analysis identified “answering Visa queries” as a target task, (b) how the syllabus was designed, (c) how the task prompts and materials were designed and modified for each language, (d) how the task was implemented, (e) tested, and (f) evaluated in each target language.

Friday, May 24

Session Ten

Understanding Drivers of L2 App User Persistence: An Academic-Industry Research Partnership
Shawn Loewen, Matt Coss, Hyun Bin Hwang, Kaitlyn Tagarelli

Mobile-assisted language learning (MALL) is near-ubiquitous globally, and research has regularly found links between MALL and L2 learning. Studies have also found that MALL often suffers from high attrition. The present study, a collaboration between university-based researchers and a language learning app company, examined survey responses and app usage data from 3,670 adult app users. A cluster analysis revealed three user-attitude clusters with varying levels of MALL acceptance. Survival analysis showed that the clusters differed significantly in their risk of MALL drop-out. This project offers implications for MALL engagement and demonstrates the benefits of academic-industry partnerships in L2 research.

Improved Automated Generation of Multiple-choice Cloze Questions for Vocabulary Assessment with VocaTT
Qiao Judy Wang, Ayaka Sugawara, Naho Orita, Ralph Rose

This study employed GPT-turbo 3.5 API, spaCy, and LemmInflect to improve an automated system for generating multiple-choice cloze (MCC) questions for English vocabulary assessment in the “VocaTT” project. Distractors were chosen from the same vocabulary unit to reduce guessing. Using 60 headwords from the 3rd sublist of Academic Word List, 105 question stems and 315 distractors were crafted. Expert evaluations revealed a 92.38% appropriateness rate for question stems and 89.21% for distractors, surpassing the previous version\\\’s rates. Challenges included stems not matching common usage, potential correct distractors, and occasional cultural inappropriateness. The study highlights the potential of automated MCCs but emphasizes the need for human oversight.

How Language Learning Differs Between a Narrative-based Learning Experience in Virtual Reality and Digital Picture Book
Jin Dong, Dongping Zheng

This project compares data from VR gameplay and a picturebook play in participants\’ home and additional languages. The VR game, Panda Reserve, employs HTC VIVE Cosmos gear, showcasing natural habitats and scenes of environmental degradation. Participants, including 18 Chinese and 10 English language learners, in groups of 3, engaged in 2 sessions of translingual interaction and reading. A subsequent focus-group interview analyzed feedback, revealing the VR group\’s strong immersion and teamwork, contrasting with the picturebook group\’s reliance on personal views. Both groups found motivation in the narrative\’s continuity and appreciated the collaborative idea generation feature, fostering increased self-awareness.

Learner’s Performance in Rehearsal Tasks: An Analysis of Discursive and Multimodal Features in Spanish Blogs
Andrea Olivares-Beltrán

This presentation examines linguistic and paralinguistic features produced by learners of Spanish in blogs as a rehearsal task. Participants were intermediate-level students of Spanish as a foreign language at a Canadian bilingual (English/French) university. The analysis of blogs was conducted at the macro-textual and micro-textual level using Gonçalves’ (2014) digital genre parameter analysis. Results suggest that, in the context of a rehearsal task, learners succeeded at imitating blog genre features, not only in terms of micro-textual and macro-textual organization of blogs, but also in some of the genre’s multimodal features. Conclusions contribute to the understanding of the impact of rehearsal tasks using Web 2.0 tools in the learning of Spanish as a second language (L2) at the university level.

From Script to 360° Simulation: Developing Research-Based High-Immersive VR Tasks for Chinese as a Foreign Language Request-Making Proficiency
Hanyu Jia

This presentation discusses the process of creating a high-immersive VR environment for Chinese as a Foreign Language (CFL) learners. The VR space uses 360-degree videos to teach request-making to CFL learners. The presentation covers the process of creating the VR space based on theory and principles of L2 pragmatics instruction, from material gathering, scriptwriting, filming, and editing for educational VR speaking tasks. The VR enables learners to practice Chinese request-making in realistic social and academic contexts. This presentation offers practical guidance for educators seeking to embrace VR technology while contributing to the scarce body of research on technology-assisted pragmatic instruction.

A Refugee’s Testimony: “I can be the teacher”
Alia Hadid

This study explored the experiences of five Syrian women refugees in the U.S. studying English online. Using narrative inquiry and through interviews, journal writing, online observations, and video recordings, we learn stories about the participants’ struggles with online learning and how they later changed their perspective about it. They reflected on how they grew as language learners throughout their learning journey. We see varying shifts in their identities; some became autonomous learners while others more confident and eager to learn. This research highlights how online learning can open doors for women refugees empowering them to become engaged learners.

Panel Presentations

Virtual Reality in Language and Intercultural Communication Education: Debating Its Potential Based on Real-life Examples
Melinda Dooly, Randall Sadler, Regina Kaplan-Rakowski, Kristi Jauregi Ondarra, Irina Golubeva

Despite increasing pressure for educational modernization, the adoption of VR-technology in language education faces significant scepticism from many stakeholders. This panel will demonstrate the potential of immersive Virtual Reality (VR) for language and intercultural communication education. First, we present a historical overview of VR in language learning (Paper 1) before exploring VR\’s potential in fostering empathy and intercultural sensitivity (Paper 2). Paper 3 discusses interculturality in VR-mediated virtual exchange. Paper 4 delves into VR interaction with young language learners and Paper 5 addresses unexplored VR-assisted language learning research areas. The session concludes with a discussion led by the panel organizers.

Participating in the Scholarly Debates: Journals in the Field
Senta Goertler, Bryan Smith, Ana Oskoz, Luke Plonsky, Jeffrey Samuels, Shawn Loewen, Jesse Gleason, Charlene Polio, Mathias Schulze

Have you wondered where or how to publish your work – be it your practitioner expertise or your research endeavors? Are you looking for inspiration about venues through which to share your scholarly contributions? Then, come on down to our panel, and learn some more about journals in our field from a team of journal editors and authors, just like you! Panelists will summarize the scope and aim of their respective journals, explain the review process, and highlight articles from past, current and/or forthcoming issues. Bring some questions and be prepared to play some journal-related trivia games!

Session Eleven

Synergizing Academia and Industry: Enhancing Language Pedagogy through Digital Tools and Mini-Projects
Lin Zhou

In recent language pedagogy, the fusion of digital literacy (Perera Muthupoltotage & Gardner, 2018) with activity-based principles (Engeström, 1987; Kaptelinin & Nardi, 2006) has emerged as an innovative approach. This study examines the impact of this approach on 40 students across three different classes. Findings indicate significant improvements in language proficiency, peer engagement, and critical thinking. Moreover, digital tools enriched classroom discussions and fostered curiosity. By leveraging modern technology, this research highlights the transformative potential of integrating digital literacy with activity-based learning, creating a bridge between technology and interactive language instruction.

Fan Practices for Learning Different Languages in the Digital Wilds: A Scoping Review
Shannon Sauro

Much of the literature on fan practices for language learning in the digital wilds published in English language CALL journals has focused mainly on extramural English and considerably less on the learning of other languages. Accordingly, this study reports on a scoping review of the peer reviewed and grey literature on fan practices for language learning published between 2000 and 2023 across multiple languages to explore the following questions: 1. What fan practices for language learning in the digital wilds have been researched? 2. What trends, if any, emerge regarding the relationship between fan practices and target languages?

Learner Engagement in Short- and Long-term Use of Duolingo in Two Class Modalities: In-person vs. Online Class
Daniel Castaneda

The researchers investigated learner engagement in short- and long-term mobile app use in two modalities: in-person vs. online. Elementary I and II Spanish students used Duolingo as supplementary to classroom instruction. Results indicate that the in-person group achieved higher app use than the online group. Activity decreased over time for the Elementary I group, whereas the Elementary II group exhibited lower initial app use, which increased midsemester and decreased by the end of the course; and a few participants used the app for a limited time after the course ended.

The Design of a Task-based Conversational Agent for L2 English in the German School Context: A Needs Analysis and First Learner Data
Elizabeth Bear

In this talk, we discuss the design and first learner data of a task-based conversational agent developed for use in L2 English classes in the German secondary school context. We present a needs analysis, insights from the interaction logs, and learner perceptions. Students from three 7th grade classes interacted with five tasks in our conversational agent and then completed a questionnaire assessing their needs and perceptions. Our findings provide some initial support for our approach and can be of interest to researchers and practitioners seeking to implement TBLT with emerging technologies and to balance learners’ needs with the curriculum.

Capturing Instances of Digital Literacy during the Writing Process: Micro-analysis of Learner-Task-Tool
Catherine Caws, Marie-Josee Hamel, Nicolas Guichon

This paper will present the results of a pilot study where FLS university B1-B2 level students (n = 20) completed a short writing task and were asked to capture their writing process and reflect on it post-task. The data collected consists of videoscreen captures as well as transcripts from semi-directed interviews conducted with a subset of the participants. We will highlight various types of strategies employed by students as well as specific instances of digital literacy development. Taking a stance from CALL ergonomics (Caws & Hamel 2016), we also analyse the quality of the ‘learner-task-tool’ interaction in that context and investigate the relation between students’ efforts and successes while soliciting various writing aids.

Quantifying ASR Pronunciation Gains with Large Learner Datasets
Dan Nickolai

CALL tools have become increasingly dependent on Automatic Speech Recognition (ASR) to provide learner feedback for L2 pronunciation. Studies to date on ASR have largely been conducted by research practitioners with small-to-medium sized subject populations at single institutions. The present study differs in that it examines a sizable 8-year dataset from iSpraak, an open source pronunciation tool. Quantitative analysis of anonymized learner interactions with this platform reveal significant gains in intelligibility measures across multiple languages. Results also suggest that the extent of ASR’s ability to improve learner pronunciation may be L2 dependent.

Session Twelve

Legislative Asks and Their (Potential) Impacts: JNCL, Multilingualism and Your Role in Advocacy
Denis Uebiyev, Betty Rose Facer

This presentation will introduce the role of the Joint National Committee for Languages – National Council for Languages and International Studies (JNCL-NCLIS), focusing on current legislation surrounding world languages. Learn how language professionals advocate to Congress on Capitol Hill to promote world language education and funding in the United States. These Official Delegates of JNCL-NCLIS are working with federal policymakers to take action nationwide. Join this presentation to understand the basics of this advocacy work, the role we all play in it, and how to make sure our voices are heard. Join our discussion on what that looks like.

Generation Z Goes Abroad: Assessing the Linguistic, Cultural, Social, and Psycho- Emotional Growth of Hyper-Connected Learners
Aurore Mroz

This study explored how smartphone usage in sixteen Generation Z U.S.-affiliated French learners related to multifaceted aspects of their learning while abroad. Framed by a language program evaluation approach and Complex Dynamic System Theory, reports on phone usage and affinity were triangulated with multiple measures of language, cultural, social, and psycho-emotional development characteristics of Generation Z. Pre- vs. post-study Kendall’s tau-b correlations were mapped out and significant changes assessed via Wilcoxon signed-rank test. Cluster analyses helped unveil different phone users. Results show that smartphones can foster meaningful connection, socialization, and exploration, but impede growth when used for audiovisual entertainment.

Learning Spanish “In the Wild” with Duolingo
Bryan Smith, Xiangying Jiang

This study investigated learning Spanish with Duolingo in an extramural setting. 200 students completed an eligibility questionnaire eliciting a pool of 48 participants below a CEFR B1 level. Results on all standardized measures (n=5) (STAMP 4S Avant test) showed that learners gained one ACTFL level after about 2 hours of use per week over 12 weeks. Learners also showed significant and substantial gains on all other in-house (n=4) linguistic measures. Number of lessons completed predicted gains in vocabulary, grammar error identification and elicited imitation scores with accuracy during app usage predicting grammar error identification, reading, speaking, and overall proficiency gains.

Development and Evaluation of a Flipped Corpus-aided Spoken English Platform for Chinese English Learners
Hsueh Chu Rebecca Chen

This paper aimed to construct a spontaneous spoken English corpus to facilitate the creation of an innovative teaching and learning program (https://corpus.eduhk.hk/english_speech_corpus/). A total of 78 datasets were collected, ranging from IELTS band 5.0 to 9.0. These were annotated for fluency, coherence, vocabulary, grammar, and pronunciation. The analysis highlighted distinct linguistic features based on proficiency levels. A ten-session flipped English program was implemented for 20 Chinese participants. The experimental group significantly improved their speaking skills, as confirmed by self-evaluation tasks and interviews. This research demonstrates the efficacy of flipped, corpus-aided instruction in enhancing English speaking, particularly for those with lower proficiency.

Memoji and Facial Cues to Lexical Stress Perception in Heritage Spanish
Sebastian Leal-Arenas, Amanda Huensch

Facial movements in animated avatars have been reported to enhance linguistic processing in L1 and L2 speakers. However, the impact of visual cues on heritage speakers’ perception remains underexplored. The present study employed a cost-effective, user-friendly technology to render audio and facial movement: Memoji. We investigated its effectiveness as a visual cue to lexical stress in pairs of conjugated Spanish verbs in two perception experiments. Results indicated that participants were more accurate and faster in perceiving lexical stress when eyebrow raising was present. Since Memojis mimic facial movements, promising avenues for pronunciation practice and teaching are discussed.

Play and Learn: Lived Experience in iVR (Immerse)
Sangmin Michelle Lee

An immersive VR (iVR) environment can facilitate learner engagement, a key construct in learning. In iVR, learners become active learners who can learn by doing, living, and playing. The current study used Immerse, a commercial iVR language learning platform and investigated how iVR helped EFL elementary school learners to engage students behaviorally, affectively, and cognitively in language learning. Based on the mixed methods research, the study found that the participants (N = 25) enjoyed language learning in Immerse, and they were deeply engaged in the language learning activities in iVR.

Session Thirteen

Robotics and Storytelling: Empowering Early Language Learners Through Coding
Natalia Hernández

This presentation delves into 21st-century storytelling by exploring the integration of coding, storytelling, and early language learning. Participants observe examples of the Cubetto robot\’s adaptation for a first-grade Spanish classroom and engage in compelling coding challenges aligned with language learning objectives. The presentation highlights how sequencing is the backbone of both coding and storytelling, and how this is leveraged in an interdisciplinary language learning approach. Student work showcases the power of robotics to inspire interpersonal communication and collaboration between students. Anecdotal examples illustrate how the curriculum engages learners with diverse learning styles and interests, ensuring broad participation and learning opportunities.

Implementation of Direct Corpus Applications to L2 Collaborative Writing Pedagogy: The Impact on Learner Interaction, Writing Quality, and Development
Ahmet Egemen Curuk

While collaborative writing (CW) research showed many benefits of peer interaction, corpus-based writing research has mainly focused on individual practices. Integrating DDL and CW has been practiced in grammar teaching with corpus before but not specifically in writing pedagogies. This study aims to analyze the impact of direct corpus guidance on focus-on-form and writing output. Forty Turkish undergraduate students in an online English writing workshop were divided into two groups, with one group participating in DDL activities using the MICUSP corpus, the other using traditional methods. Pretests, posttests, and delayed posttests were conducted, and essays were analyzed for multidimensional syntactic complexity, lexical diversity and accuracy.

Developing(Lesson Planning) and Teaching(Debugging) Communicative Language Activities in Virtual Reality
Peter Grzywacz

For five weeks straight, I designed, coded, and taught communicative English lessons to eight Japanese students in VR Chat. Worksheets, textbooks, and slides became 3D meshes, programming scripts, and colliders. In this class report presentation, attendees will see how the activities were built, live demos of the activities, and how students’ feedback and reactions affected the design process. Attendees will also gain access to the website containing links to all VR Chat Worlds and Github repositories containing all files and code.

From Mundane to Memorable: 5 Ways to Contextualize an ESL Topic Using Edvibe.com
Raihan Rahman, Anastasiia Stroganova

There is an increasing number of AI tools and Learning Management Systems (LMS) available to users nowadays. Although it has the potential to make the learning process more contextualized, the integration of AI with LMS is still uncommon in educational technology. This presentation will demonstrate how ESL educators can enhance contextualization using interactive AI-assisted tools. The presenters will demonstrate one such LMS, Edvibe, which aids in lesson planning and design. The presenters will additionally provide attendees a URL that they can use to access a guidebook containing samples of how teachers can use Edvibe tools to contextualize their topics.

Panel Presentations

Advances in CALL Research and Practice
Senta Goertler, Jesse Gleason, Angelika Kraemer, Emma Britton, Hengyi Liu, Linda Jones, Theresa Austin, Xinyue Zuo, Curtis Maughan, David Fredrick, Stephanie Link

Every year, CALICO publishes a book in the “Advances in CALL Research and Practice” series. Recent editorial teams from the book series will present their work. Learn more about the series; get an overview of the current issue on crisis-prompted language teaching which is available for free online for CALICO members; take a sneak peek of the 2025 issue on critical CALL across institutions and borders; become inspired to contribute to the 2026 issue on game-based approaches to CALL; or go all in and respond to the call for new editors for the 2027 issue.

Scaling Research and Navigating Challenges: Linguist Entrepreneurs
Linh Phung, Raia Lichen, Heidi Brumbaugh, Hope Anderson, Will Jordan-Cooley

This panel brings together four pioneering linguist-entrepreneurs at the forefront of innovation in computer-assisted language learning (CALL). Delving into the theme of \’Confluences and Connections: Bridging Industry and Academia in CALL\’, the panel will explore critical facets of leveraging research for entrepreneurial success. They will discuss strategies to transform academic insights into marketable solutions, address misconceptions in language learning, and navigate the intersection of emerging technologies and language education. Panelists will share their personal journeys transitioning from academia to edtech, highlighting the challenges and triumphs of carving a niche in the dynamic landscape of language learning and technology.

Session Fourteen

Virtual Exchanges for Beginner Language Learners: Do They Work?
Ana Oskoz, Marta Gonzalez-Lloret

This study explores the effects of a Virtual Exchange (VE) between learners and expert speakers of Spanish on learners’ communicative strategies, their willingness to communicate, and their perceptions of learning. The data were then compared with that of intermediate low and intermediate high learners in the same program to observe the impact that proficiency may have on the effect of a VE for language learning and communication. Results highlight VE\’s potential, while revealing limitations. The presentation offers pedagogical guidance for incorporating VEs in introductory language courses.

Perceptions of AI-based Technologies in Foreign Language Education
Michael Hofmeyr

The capabilities and practical applications of AI technologies have advanced rapidly over recent years. Foreign language education has already been dramatically impacted by these developments, as learners gain access to increasingly powerful AI tools capable not only of improving grammar and vocabulary choices in academic writing, but also of generating highly coherent and cohesive original written texts. This presentation will review the key findings on language learner and teacher perceptions of and attitudes towards AI technologies, including AI-augmented machine translation tools and generative AI technologies such as ChatGPT. A work-in-progress research project investigating actual AI-based academic practices among university students and teachers in Japan will also be introduced.

Age and First Language Bias in Automatic Speech Recognition: Ensuring Test Fairness
Carey Nelson, Carol Johnson

Automated scoring of pronunciation proficiency tests using automatic speech recognition (ASR) has become commonplace. However, ASR biases in terms of participants’ age and their first language have been documented, which could affect test fairness. This study examines scores of 1000 pronunciation proficiency tests generated by Google Voice Typing and Microsoft-Transcribe, the target ASR-based dictation systems. A multiple regression analysis of the scores from each system was carried out with the predictors of age and L1 to determine possible bias. Implications for test scores and methods for mitigating bias will be discussed.

Communicating, Connecting, and Creating: The Impact of Social Media on L2 Motivation
Hannah Hautala, Kimberly Morris

Considering the global prominence of social media and its evolution as a space where frequent multilingual and multicultural interactions occur, this study examines the relationship between L2 learners’ social media use in their target language and its influence on their intrinsic motivation to learn and use their L2. Results from a survey of 214 university L2 learners revealed a significant positive relationship between participants’ L2 social media usage and their intrinsic motivation, which was supported by salient trends found in 13 semi-structured interviews. Thus, it is crucial to harness the affordances offered by social media through integration into language classrooms.

Designing a VR Leveraged Distributed Language Learning Environment for Language and Translanguaging
Dongping Zheng, Michaela Nuesser, Jin Dong, Weiying Huang

This paper presents data from a project focused on providing languaging and translanguaging experiences to English and Chinese learners through a designed distributed language learning environment (DLLE). The DLLE aims to simulate an environment where individuals proficient in each other’s native languages collaborate to solve complex environmental issues. Utilizing playing characters, non-player characters (NPC), and a Virtual Reality design played with HTC VIVE Cosmos, learners assume roles and navigate the VR Panda Reservation Village to interact with NPC citizens through written dialogues. Our analysis reveals nuanced tensions, highlighting the importance of balancing individual and collective goals within the learning environment.

Immerse: An Authentic, Live VR Language Learning Platform
Tricia Thrasher, Carla Consolini

This session will introduce attendees to Immerse, a VR platform specifically created for live language teaching and learning. Immerse is a social VR language education platform that allows students to take language lessons with highly trained language teachers in English, Spanish, and French. Students can also participate in instructor-led conversation groups, converse with AI-powered avatars, and play learning games. Attendees will learn about the pedagogy behind Immerse, have a chance to try out the application, and learn how they can use it with their own students.

Session Fifteen

Self and Social Organization: Liminal Activations as Sites for Collective Learning, Community Organization, and Innovative Partnerships
Christopher Daradics

This presentation offers an analysis of several mixed-reality experiences and explores key features and affordances of such liminal activations (Turner 1998). Beginning with an account of intersubjectivity, our talk calls into question traditional hierarchies and outmoded social habits in favor of reflexive, cocreative, and social pedagogies (Dubriel and Thorne, 2019). Through these cases, we will explore practical examples, lessons, and foundations for community activations across diverse language landscapes and institutions. Specifically, we will highlight how these experiences leverage play, ambiguity, and collective curiosity to operationalize generative, reflexive conditions and cocreation in pursuit of enlanguaging preferred and adjacent possible futures.

Conversational AI for Learning English: A Systematic Review and Future Directions
Yilin Zhang, Hao Yu, Faceia Hou

Recent advances in conversational AIs offer new opportunities for learning English. While current conversational AIs have presented promising potential and positive influence on English learning, they are still limited to interaction channels and would occasionally output unwanted information. (Dills & Hall , 2021, Schmitt et al., 2022). Our systematic review analyzed over 20 empirical studies focused on conversational agent models proposed from 2019 to 2023 with precise criteria for aiding English learning. The latest models exhibited cutting-edge performance, suggesting possible solutions to current challenges. We strived to give future researchers an overview of the potential uses of different models, providing insights to steer the development of next-generation conversational AIs for English Language Learning.

The Connection of Design and Immersive Language Learning: A Close Look
Yalun Zhou

In recent years, iCALL has made strides in the field of second language (L2) teaching. Conversational agents have emerged as effective tools for language learning. However, the integration of AI chatbots for non-English language learning, particularly Chinese, remains rare. This presentation shares data from an AI-assisted immersive Chinese course to illustrate design features and learner engagement. Examples demonstrate processing input, using live transcriptions, and conversing with AI avatars. The focus is on illustrating immersive language learning design and how learners engage with AI avatars, encouraging discussions on AI-mediated learning design and future development and collaboration between language teachers and designers.

The Effectiveness of Using Twitter in Developing EFL Saudi Female Students’ Grammatical Knowledge
Abdurrazzag Alghammas, Lima Alzamil

The increasing popularity of social networking sites has tempted computer-assisted language learning (CALL) researchers to investigate their potential uses for language learning. Accordingly, this mixed methods experimental study aimed to determine the extent to which Twitter is effective in developing EFL Saudi female students’ grammatical knowledge, specifically their receptive and productive grammatical knowledge. It also investigated students’ perceptions of using Twitter to learn grammar. Fifty-five Saudi female students majoring in the English language at a Saudi university were divided into an experimental group (n=26) and a control group (n=29). The analysis of the pre- and post-tests indicated that Twitter was generally effective in developing the students’ productive but not receptive grammatical knowledge and its effectiveness differed based on the grammatical structures being taught. The data from the interviews revealed that the students generally had positive perceptions of using Twitter to learn grammar although they had some concerns. The study concluded with some pedagogical implications, in light of the study’s findings, and some recommendations for future research.

Fostering Autonomy: Strategies and Challenges in Training In-Practice Teachers to Use Data-Driven Learning in Language Education
Larissa Goulart

Recent studies (e.g., Boulton & Cobb, 2017) have shown the benefits of introducing Data-Driven Learning (DDL) in the language classroom. However, researchers have also pointed out that most DDL studies were conducted by research-teachers in the context of EAP and university classes. That is, the use of DDL has yet to make an impact in primary and secondary school contexts. One way to close this research-practice gap is through training in-practice teachers in the DDL approach. This presentation describes the steps and outcomes of a teacher training course that sought to introduce DDL to in-practice language teachers.

Voxy as an LXP: How Does it Affect Language Learning in the Classroom? Is it a Myth or Reality?
Dilara Saygin, Kamil Kirkiç

The goal of this study was to show whether a connection exists between student engagement on the learner experience platform (Voxy) and their academic success in university-preparatory classes. The quantitative research model was used to process the data. The researchers used data entries from Voxy and end-of-year grade results from the preparatory school. The results show a minor positive correlation between student participation levels and their final grades. Even though different instructors alter the results of LXP engagement levels, when Voxy was implemented in the classroom gender and type of instructor did not have an impact on students’ academic success.

Panel Presentations

Games for Language Learning: Connecting Industry, Research, and Practice
Frederick Poole, James York, Jason Kappes, Daniel Dixon

This panel discusses games from three different perspectives: design, research, and pedagogy. From a design perspective we discuss principles for designing mechanics and content to achieve learning outcomes. From a research perspective we provide an overview of techniques for studying games and key findings about their effectiveness, limitations, and future directions. Finally, from an instructor’s view we consider pedagogical implications like the teacher\\\’s role, task design leveraging game affordances, and impact in classrooms. Through this discussion we aim to highlight opportunities for deeper collaboration across disciplines to advance the design, study, and use of games to meet language learning goals.

Reassessing VR for Language Learning: Has It Fulfilled Its Potential?
Robert Godwin-Jones, Naoko Taguchi, Randall Sadler, Justine Meyr

2024 may be VR’s break-out year, with a slew of forthcoming devices, including Apple’s VisionPro. Emerging hardware may finally address issues preventing widespread VR use: accessibility/affordability, physical comfort, real-world integration. Recent meta-analyses of VR for language learning, however, show that most implementations fail to leverage its power, targeting vocabulary learning and featuring scripted role-play scenarios. This panel will showcase projects that take advantage of the affordances of VR to create meaningful learning opportunities, including in virtual exchange to assess intercultural learning, integrating AI chatbots into multiplayer gaming, and combining VR apps to provide a personally enabling environment for inter-generational learning.

Session Sixteen

Fostering Language Learning, Wellness, and STEM Education Through High-Immersion Virtual Reality
Regina Kaplan-Rakowski, Sabina Nowak, Prerna Choubey

This presentation introduces an innovative approach combining language learning, wellness, and STEM education through VR. Language learners swam with fish in VR and learned about marine biology. Students could absorb STEM content in a FL while benefiting from the relaxing effects of the VR ocean, promoting wellness and a connection with the subject matter. Advanced EFL learners (N = 120) were randomly assigned into two groups: VR and video-based STEM content. The study assessed technical vocabulary and STEM learning gains, enjoyment, mindfulness, cognitive load, and presence. ANOVA and Wilcoxon tests identified the optimal conditions for language, STEM, and wellness gains.

Crisis Response and Crisis Preparedness: Moving Forward in the ‘New Normal’
Senta Goertler, Jesse Gleason

COVID-19 is here to stay and going back to a pre-pandemic normal impossible. The COVID-19 pandemic was a health crisis and a global traumatic period not just medically, but also socially, financially, and emotionally, that disrupted learning, working, and living in multifaceted ways. The crisis brought to light inequities and issues in infrastructure, forced innovation and creativity, and the need for a more equitable and just ‘new normal’ that puts the affordances and limitations of technology at the center. This session summarizes policies, practices, and procedures to help language programs productively and sustainably navigate a just and equitable ‘new normal.’

Integrating Cultural Competence and Technological Literacy in Medical Spanish
David Ortega

With the advent of new technologies, such as AI, redesigned language for specific purposes courses (LSPs) are imperative to better prepare learners for target-language usage in the real world. This presentation focuses on a Medical Spanish course targeting medical professionals for effective engagement with Spanish-speaking patients. The program weaves cultural competence with advanced linguistic instruction, tailored for needed practical medical exchanges. It harnesses authentic multimodal resources and introduces Artificial Intelligence (AI) in healthcare communication, aligning language proficiency with current technological trends. We seek to showcase pedagogical choices aimed at equipping healthcare providers with the skills for empathetic and culturally informed interactions in an ever-changing and megadiverse landscape.

It’s What You Teach Not How You Teach It: A Corpus Study of Real-World Language Use in Pimsleur’s Tagalog
Nicole De Los Reyes

Since 2008, the ubiquity of commercial mobile phone and tablet applications for foreign language learning has allowed for investigations of their efficacy by researchers in mobile-assisted language learning. This study explores the extent to which Pimsleur’s Tagalog course reflects real-world language use, combining constructionist (Goldberg, 2013) and corpus approaches. N-grams and phrase-frames of varying lengths were examined from a self-compiled corpus of 30 half-hour conversation-based audio lessons and Sketch Engine’s OpenSubtitles 2018 parallel – Tagalog corpus (96,219 words; Lison & Tiedemann, 2016). Findings show how insights derived from construction grammar and corpus linguistics can facilitate idiomaticity in app content development.

“I like that I can talk to Spanish speakers in Illinois”: Students’ Perceptions of AI-powered Role Plays in Virtual Reality
Uliana Ovsiannikova, Tricia Thrasher

In response to a lack of research on high-immersion AI learning tools, this presentation will discuss a study examining how Immerse, a virtual reality language learning platform that offers AI-powered avatar role plays, impacts high school students learning Spanish and French. Students completed pre-study questionnaires targeting their beliefs about AI and assessing their language abilities. Next, they engaged in AI-powered role plays during class time over the course of a semester. Post-task evaluations followed each role play, culminating in a final language assessment and feedback on AI usage. This talk will present students’ perceptions of how AI impacted their learning.

Session Seventeen

Integrating ImmerseMe into the Classroom: World Language and Intercultural Development
Denis Uebiyev, Betty Rose Facer

Although it is exciting to integrate emerging technologies into a world language classroom, it’s vital to also attempt to measure students’ learning outcomes and development through this integration. Virtual reality (VR) has had a strong push in the field of CALL and has even been fully integrated into various world language classrooms at the university level. This presentation will analyze the progress of a year-long, ongoing project integrating ImmerseMe at a Mid-Atlantic university. This is done by evaluating students\’ perceived language and intercultural development through an end-of-year survey and aggregated data on students’ use of ImmerseMe throughout the semesters.

Exploring the Influence of Extramural Virtual Guided Conversations on L2 Oral Skills and Motivation
Luis Cerezo

Offering videoconferencing sessions with native speaker coaches has gained some traction in second language (L2) education. Consequently, there has been some research (e.g., Sama & Wu, 2019; Marull & Kumar, 2020) investigating the benefits of these platforms in L2 curricula. However, their efficacy in enhancing L2 oral skills and sustaining learner motivation remains untapped. To address this gap, our study explores the impact of videoconferencing sessions with native speakers on L2 oral proficiency and motivation. It has three main goals: evaluating oral skill development, exploring motivation shifts, and examining learners\’ perceptions on LinguaMeeting\’s role in fostering L2 oral skills.

WhatsApp versus Zoom: Learner and Instructor Perceptions and Attitudes of Mobile Learning Tasks
Lillian Jones

Text messaging, common among undergraduates, facilitates interactive and collaborative exchanges crucial to a sociointeractionist framework for second language acquisition (Blake & Guillén, 2020). The rise in mobile-assisted language learning research has yielded varied methodologies and insights (Stockwell, 2022), examining learner attitudes towards mobile learning (Burston, 2013; Li, Cummins & Deng, 2017). This presentation reports on a 10-week study exploring the use of WhatsApp (or Zoom) to complete interactive tasks. Topics discussed will be learner and instructor perceptions of task design, language partner engagement, and perceived skill development, highlighting the potential of mobile devices for out-of-classroom learning, and implications for teaching and research.

Using Mobile Phones in Resource-Poor Areas for Learning Language Skills
Thomas Robb

A large percentage of language learners around the world now have mobile phones at their disposal. Many surely use them to acquire some aspects of the language, but restrictions on device availability and access cost might mitigate against use in classroom contexts. Many applications work best with autonomous learners, while most learners still expect direction from teachers. With this in mind, for three target languages, we will look at commonly used apps and which components of language they might be effectively used for, keeping in mind both cost and the degree of autonomy required for effective use.

The Effects of Machine Translation and Virtual Exchange on Learning Spanish through Humor
Cristina Pardo

Research studies on machine translation (MT) have not only impacted L2 learning (Ducar & Schocket, 2018), but they have also exposed limitations for translating humor (Niño, 2020). This study evaluates the effectiveness of 24 Spanish learners’ virtual learning experiences using Spanish native speakers and MT for understanding source texts with a humor focus. Using the functional approach to translation (Nord, 2019) and action research (McNiff & Whitehead, 2005), this study seeks to identify the effect of incorporating MT and virtual exchanges into language teaching to help learners solve translation problems and produce creative target texts with a humorous effect.

Saturday, May 25

Session Eighteen

Can ChatGPT Reliably Score L2 Writing Assessments? The Devil is in the Prompt
Frederick Poole, Matt Coss

Researchers have created tools for automated writing assessment. However, such tools are rarely used in classrooms due to technical complexity. This study explores approaches to prompting ChatGPT to reliably assess second language writing samples. We analyzed ten ChatGPT prompts using a 144-response dataset scored by human raters, finding reliability levels ranging from 0.17 to 0.57 depending on prompting strategy. Our findings provide implications for optimal prompting strategies for using ChatGPT as an assessment tool. We conclude by offering guidance for automated language assessments via ChatGPT in the classroom.

Leveraging AI to Support Language Learners
Rachel Hernandez, Stacy Amling

Universal Design for Learning (UDL) provides guidelines for designing flexible and engaging learning environments that proactively build in supports to meet the needs of all learners (Meyer, Rose, & Gordon, 2014). UDL can be used to strategically integrate digital tools to address learner variability in diverse classrooms. This presentation will introduce principles of UDL, showcase generative AI tools that support UDL, and demonstrate their use to support language learners across all modes of communication. For example, digital tools, such as text-to-speech or text-levelers, can support interpretive reading, by providing multiple means of representation to make input comprehensible and help sustain engagement.

Bringing Digital Games into the L2 Classroom: A Systematic Review of the Pedagogical Implications Presented in Research
Lincoln Bain IV

There has been a growing trend in recent years to use vernacular (commercial, non-educational) and serious (educational ) video games in formal and informal learning environments (Peterson & Jabbari, 2022). This systematic review will be centered on the pedagogical implications provided within the literature. A total of 85 studies from 2020 – 2023 will be systematically reviewed and coded using a grounded approach to the recommendations, strategies, and considerations for instructors. This grounded theory analysis hopes to create a framework to analyze pedagogical implications and help improve them to be more valuable to practitioners and potentially eliminate the stigma behind games for learning.

Implementing Large-Scale Virtual Reality-Assisted Language Learning Research
Tricia Thrasher, Randall Sadler, Uliana Ovsiannikova, Dorothy Chun, Regina Kaplan-Rakowski, Justine Meyr, Ye Yuan, Yongluan Ye

Research on virtual reality (VR) and language learning has been growing, however, most studies have relied on small sample sizes due to the difficulty of acquiring and implementing VR equipment into classrooms (Dhimolea et al., 2022). This presentation reports on the initial stages of a large-scale project that aims to fill this gap. With grant funding received from Meta and Immerse, we have been implementing language learning research using 500 VR headsets at 12 schools in Texas, Illinois, and California. Our presentation will discuss the necessary steps to conduct large-scale projects which encompass successful VR equipment management as well as teacher and student VR training.

Language Learning Beyond Borders: Applying Cognitive Behavioral Game Design to the Virtual Space Developed with Spatial.io
Ogulcan Durmaz

This presentation will help scholars understand whether Spatial.io, an immersive social platform, is fully equipped to provide effective language teaching practices based on Cognitive Behavioral Game Design Theory. The space designed for language teaching on Spatial.io will be technically introduced and pedagogically analyzed in this presentation. The audience will be stimulated to evaluate and critically think if virtual spaces designed on Spatial.io are good fits for language teaching depending on the Cognitive Behavioral Game Design Theory.

Using Augmented Reality (AR) to Promote Cross-Group Communication
Thanaphan Thapthimhin

With English as lingua franca, communication by learners of different groups is possible and essential for the language learning. The advent of the augmented reality (AR) can make the learning activity become more engaging and intriguing with its capacity to generate realistic 3-D models and environments, provided that the instructor has an interesting language learning activity design. This presentation demonstrates a learning activity focusing on communication of learners across groups, using the AR as a tool for more interaction and communication through listening, reading, writing, and speaking.

Panel Presentation

Artificial Intelligence and Second/Foreign Language Writing
Oksana Vorobel, J. Elliott Casal, Charlene Polio, Erik Voss

Recent emergence and increasing use of artificial intelligence (AI) tools has a growing influence on language learners’ writing practices in their classrooms and beyond. AI tools are also changing how language educators and researchers approach writing in various contexts. This CALICO LTLT SIG sponsored panel will explore these influences and will showcase recent studies on using AI for instruction and assessment in writing classrooms and for research. At the end of the session, all participants will be invited to share their ideas and/or questions about their research.

Session Nineteen

Using Pimsleur for Learning Spoken Phrases in Brazilian Portuguese
Francis Will, Walcir Cardoso

This case study examines the self-regulated use of Pimsleur, a language learning platform (LLP), as a tool to aid in the acquisition of spoken phrases in Brazilian Portuguese (BP) and their related pronunciation. Like many LLPs, research on Pimsleur is scant, as is the number of studies done on BP compared to other major languages. This study aims to address this gap in research. The participant-researcher completed the Pimsleur program through daily study over a 10-week period, after which quantitative data were collected through a post-test and delayed post-test. The results showed that Pimsleur contributed to the learning of the target phrases in the short term and that the participant produced speech that was highly intelligible, moderately comprehensible, but heavily accented. This shows that Pimsleur can be an effective tool for developing spoken BP and can offer a unique learning experience with its methodology and mobile capability that mitigates some of the issues around mobile-assisted language learning (e.g., app attrition).

The Good, the Bad, the Ugly: Teachers’ Exploration of Generative AI for Teaching Language and Literacy Skills to English Learners
Shelley Xu

Technology integration has positive effects on English learners’ language and literacy development. This study focused on K-12th teachers’ exploration of three GenAI tools (Bard, ChatGPT, and Claude) and their educational potentials for supporting English learners (ELs). The teachers were generally satisfied with these tools’ outputs, which served as a source of texts beyond classroom curriculum. These tools also offered ELs interactive opportunities for practicing high-level reading and writing skills. The teachers were concerned about some GenAI outputs with inaccurate and/or incomplete information and with a language level beyond ELs’ ability to read and comprehend.

Building a Serious L2 Learning Game: Perspectives on Collaborations between Academia and Industry
Jonathon Reinhardt, Lincoln Bain IV, Deanna Terzian, Christian Bruccoleri, Nicky Ramos-Beban

What does it take to build a commercially successful and sustainable L2 learning game? What is the potential role of research and academic consulting in its development? In this talk, the research team involved in the design and usability testing of Mage Duel, an L2 learning game being developed by a small game developer for the USAF will present the game and discuss the processes of collaborating on the game\\\’s design, consulting vis-a-vis SLA theory and conducting usability studies. The team will offer general insights on L2 learning game development and discuss the value, challenges, and potentials of industry-academia collaborations.

The Use of Immersive Virtual Environments to Enhance Language Learning, Improve Learners’ Experience, and Foster Willingness to Communicate
Irene Soto

Research suggests that several features of immersive virtual environments can help foster and enhance the language learning (LL) process, providing authentic simulations of real-life scenarios and contextualized learning. This can situate learners in a realistic setting to produce the target language. Additionally, virtual reality (VR) can positively impact learners’ anxiety, motivation, engagement in the task at hand, willingness to communicate (WTC), and attitudes towards LL. This study reports how practice with the platform ImmerseMe impacted the level of engagement and WTC in university learners of L2 Spanish, exploring attitudes towards this tool and which aspects were viewed as most beneficial.

Creation and Generation: ChatGPT in the Spanish Writing Classroom
Gillian Lord

We report on an informal classroom study exploring the potential benefits of ChatGPT on student learning. Twenty participants in an upper-division Spanish Creative Writing class first wrote a short story and then instructed ChatGPT to create as similar a story as possible. Doing so involved providing multiple iterations of directions and guidance and required the students to think critically about the nature of their own writing. We discuss both quantitative and qualitative findings from the study. The language and linguistic structures of the human- and machine-generated stories are analyzed and compared, and student reactions are also summarized and reported.

LanPIP: A Specialized LLM-empowered Pipeline for Customizable Language Learning
Faceia Hou, Yiqing Shen

Language acquisition is a complex cognitive process that can benefit from interactions with large language models (LLMs). Despite the few-shot learning capability of LLMs, current applications lack integration to fit learners\’ needs. Our proposed LanPIP utilizes LLMs for automated inference and even training for personalizing educational responses. Innovatively, it supports cross-lingual, multi-modal input and output for an adaptive language learning system that utilizes LLM and prompt engineering embedded with card instructions to provide customizable language processing capabilities spanning multiple languages and modalities. The system is evaluated through comparative performance validation and learner-subject experiments to show its usability in supporting various language-learning needs.

Session Twenty

The Ups and Downs of Learner – Chatbot Dialogues
Ulf Schuetze

Empirical data collected from 8 groups of ESL students who used a Virtual Environment (created based on research on brain connectivity, the notion of play, speech recognition software and intrinsic motivation) at the beginner’s levels of proficiency is presented. Each group had the same controlled access to the environment but used a different feature or features as well as one of two devices (mobile, VR headset). The analysis showed the challenges leaners faced talking to a chatbot as well as the benefits, if they stayed engaged. Recommendations to improve the design and functionality of these types of chatbots to provide the learner with more choices are being made.

The Future is Now: Connexions francophones
Patricia J. Kyle

Experience the future of language learning and teaching at this session showcasing unique pedagogical opportunities now accessible through virtual cultural immersion. Discover the first digital-primary French electronic textbook, Connexions francophones: Identités culturelles et linguistiques, developed at the University of Texas at Austin. Observe the power of state-of-the-art digital media to immerse beginning learners in language and culture learning as they explorethe francophone world and advance to intermediate proficiency. Available now from McGraw Hill.

Enhancing Vocabulary Learning in University Classes through a Gamified Application Connected to Games
Enzo Simonnet, Elise Lavoué, Mathieu Loiseau, Sebastien Dubreil

The Lex:gaMe project aims to provide a broad set of tools to support and enhance vocabulary learning. The pivotal tool in the kit, BaLex is a shared vocabulary notebook designed to enhance vocabulary building, motivation, and engagement in the L2 classroom. It was also designed to allow interaction with other vocabulary-learning applications, in particular two games: MagicWord [5] (written) and Prisms, an extended version of Game of Words [6] (oral and written). This paper describes the protocol for a study that aims to assess the added value of connecting games to BaLex and preliminary results.

Less Commonly Taught and Indigenous Languages and CALL: Models of Support
Emily Heidrich Uebel, Luca Giupponi

CALL is vital for supporting less commonly taught languages (LCTLs). Often characterized by low enrollment rates and faced with a shortage of high-quality pedagogical resources (Blyth, 2013), instructor isolation, and inconsistent training (Johnston & Janus, 2003), LCTL instructors and programs need support (Brown, 2009). This presentation leverages the experience through two large grant projects to demonstrate how CALL, and more specifically, the use of technology to share courses, enhance professional development, and develop resources across institutions, can positively affect LCTL and Indigenous language instruction.

The Design and Development of Introduction to Research Methods App (IRMA): Educational Technology for Supporting New Researchers in Applied Linguistics
Daniel Dixon, Tulay Dixon, MaryAnn Christison, Adrian Palmer

In the interest of supporting innovative methodological teaching and training in Applied Linguistics (AL), we discuss the design and development of an open-source supplemental digital teaching tool titled Introduction to Research Methods App (IRMA). IRMA was designed with the aim of providing beginning-level researchers with a strong foundation in the core characteristics of high-quality research in AL through discovery learning. It guides learners through the development and use of quasi-experimental research through an interactive and intuitively designed graphic user interface. IRMA will be made freely available to students and teachers interested in research methods in AL.

Khanmigo – The AI Storyteller Weaving Tales of Genre and Grammar
Shamini Shetye

ESL instructors often grapple with the time-consuming task of creating classroom content. The presentation aims to unravel the feasibility of an AI tool, Khanmigo, (Khan Academy, 2022) to minimize teachers’ efforts by automatically generating instructional content for an Intermediate ESL Grammar and Writing class. 16 Khanmigo-generated reading passages spanning five diverse writing genres and eleven grammar forms were analyzed quantitatively and qualitatively by experienced teachers/linguists. Findings revealed that Khanmigo-generated texts exhibited grammatical precision, genre relevance and appropriate level of lexical complexity, thus illuminating Khanmigo’s key potential to reduce human effort and amplify efficiency in the creation of instructional materials.

Panel Presentation

The Digital Confluence: Merging VR and AI in CALL Environments
Denis Uebiyev, Carla Consolini, Tricia Thrasher, Nicole Mills

This panel examines the real-world applications of Artificial Intelligence (AI) in Virtual Reality (VR)-based language learning. Featuring industry experts from ImmerseMe and Immerse, as well as an academic researcher in the field, the panel will examine how generative AI is being integrated into these platforms to enhance educational outcomes. Attendees will gain insights into this integration\’s practical challenges and opportunities. The session will include a Q&A segment, encouraging open dialogue between the diverse industry and academia perspectives so that this cross-section\’s continued work and development can be further explored.

Session Twenty-one

To Speak or Type: Does Task Difficulty and Cognitive Burden Influence Learner Behavior in a Language Production Task?
J. Scott Payne, Lillian Jones

Mastering a language involves fluent conversational skills, but real-time oral communication remains challenging. Speaking and writing have cognitive similarities but differ in speed and permanence. Speech, being ephemeral, places a greater burden on working memory than text. Our study explores how task difficulty and proficiency influence learners\\\’ choice between text and speech as language production options in an immersive, VR language learning environment. We\\\’ll discuss whether learners choose text as a mode of production to simplify tasks and its implications for teaching and course design.

Boosting Metalinguistic Awareness with a Plurilingual Escape Game
Coralie Payre-Ficout

Various research studies have highlighted the value of an approach to languages based on the comparative observation of their functioning (Hawkins, 1984, Candelier, 2003). The challenge of a comparative approach to languages is to improve language and language skills, in particular through the development of metalinguistic skills, which are essential both for access to writing in the language of schooling and for learning a first language, then other foreign languages, etc. (Ober, Garcia-Debanc & Sanz-Lecina, 2004). In this presentation, we will present the pedagogical material constructed and its integration into an educational escape game and we will describe the evaluation protocol implemented.

Multimodal Instructional Approaches in CALL — More Is Not Always Better
Goretti Prieto Botana

Cognitive Load Theory (CLT) strives to identify forms of instruction that present information in ways that maximize learning outcomes. In second language learning, previous CLT research suggests that explicit instructional designs are superior as compared with designs requiring learners to find the solution themselves. However, not all explicit instruction facilitates learning, with some resulting in split-attention. Two studies were designed to gauge the differential effects of four different explicit information treatments, in recognition and production. Results suggested that more multimodal forms of explicit information may cease to be facilitative when structure and task complexity are higher.

A Universal, Application-Independent Vocabulary Learner Model: From Islands to Archipelago
Heidi Brumbaugh

Central to Intelligent Computer Assisted Language Learning lies the learner model, serving as a framework to track individual learners’ knowledge and adapt software to their specific needs. The goal of the current project is to marry this theoretical framework to the realm of big data.  I will present a new framework that allows various language learning applications to access a central model, which monitors the learner’s vocabulary acquisition across multiple modalities, including elicitation, task-based activities, reading, translation, and more. This comprehensive vocabulary model will identify the next, best word for which the learner needs each app’s particular type of engagement.

Language Gourmet: Food Culture in Experiential Learning
Sue-mei Wu, Nevine Abraham

Are you interested in incorporating food culture learning into your language and culture teaching? Would you like to learn how to create digital materials to enhance your food culture curriculum? This panel “Language Gourmet: Food culture in experiential learning” is organized to share classroom pedagogical experiences related to Chinese, Israeli and Palestinian food cultures. The presenters will share innovative pedagogical designs and practices that help students immerse themselves in cultural learning activities through discussions, group collaborations, and hands-on practice. The presenters will introduce some digital food culture learning materials and demonstrate how to effectively incorporate them into learning activities.

Session Twenty-two/Asynchronous Sessions

Teaching Language Through Place: Exploring Teachers’ Attitudes Towards Place-Based Education
Lara Lomicka Anderson, Liudmila Klimanova

https://youtu.be/qp3ibJt5zMU

Place-based language teaching is the process of using place and space as a nexus of cultural and social practices in a community and a starting point to teach concepts related to language use and cultural practices, such as geolocality, cultural space, spatial storytelling, and mapping. This presentation will propose a new framework of PLACE in language education, outline the key principles of TEPBLL (Technology-Enhanced Place-Based Language Learning), and report on a study of 152 K-16 teachers’ experiences with place-based teaching in language classes. Implications for future TEPBLL research, activities, and training will be discussed.

Overcoming Barriers and Co-construction During Online Learning: Cognitive Presence for English Language Learners through Flip
Ellen Yeh

https://youtu.be/IC59BRGrBDI

This study explores the cognitive presence of English language learners within an online learning environment. The main purpose of this study is to examine the use of Flip using the Community of Inquiry and the Practical Inquiry Model. ELLs participated in a semester-long weekly Flip, a video-based asynchronous computer-mediated communication discussion. The data for this study includes main reflection videos (N = 137) and peer response videos (N = 153) and the content analysis provided an understanding to develop deeper learning and a critical reflection on using Flip for (co-)constructing knowledge and enhancing social interaction.

Maximizing Student Engagement: The Flipped Learning Approach
Ayano Kawasaki

https://youtu.be/VskbO7ygVII

The presentation demonstrates what a flipped classroom is like and its potential benefits for language learning based on my teaching experience in the EFL classroom. It also includes step-by-step instructions about how to design a flipped classroom and create technology-mediated materials, as well as practical tips to maximize the effectiveness of flipped learning by incorporating online games.

AI-Ready or Not: Reflective Inquiry into Pre-Service Teachers’ Preparedness for Digital Classrooms
Dini Arini, Jihee Im, Hadir Alderaan

https://youtu.be/MdPyBpmCWxE

This study examines the readiness of pre-service teachers to incorporate Artificial Intelligence (AI) in classrooms, based on reflective inquiries conducted in an interactive workshop. It captures the pre-service teachers\’ exploration, understanding, and preparation for integrating AI technologies in their future classrooms, as seen in their workshop responses. The inquiry covered four key questions about the use of AI in classrooms, including its application, challenges, student support, and personal utilization, using Diffusion of Innovation Theory (Rogers, 1962) as theoretical framework. The study reveals insights into educators\’ readiness, ethical considerations, and pedagogical strategies for integrating AI. The findings highlight the importance of reflective workshops for pre-service teachers in developing their digital competence and ability to utilize AI in education.

Activity Theory: Using Activity System Analysis to Investigate Technology-mediated Interactions
Robin Couture-Matte

https://youtu.be/47n4P8L9a1I

This presentation aims to describe how activity theory can be a useful empirical lens for the investigation of technology-mediated interaction. In this regard, activity system analysis will be presented as a tool to study how learners interact within their sociocultural contexts with the aim of reaching their learning goals. This empirical lens will be exemplified with a study in high-immersive virtual reality with young learners (aged 11-12) who carried out interactive tasks with head-mounted displays. The outcomes and challenges of the qualitative analysis that was carried out using activity system analysis will be presented.

Impact of H5P Interactive Videos on Student Learning in a Japanese Language Course: “You no longer have to imagine them yourself!”
Naoko Takei, Pauline Tiong, Chunhong Liu

https://youtu.be/rkWDoDpsDlM

Using H5P technology, a series of interactive videos was designed as pre-class self-learning materials in an introductory Japanese language course. The videos were designed to enact real-life Japanese conversations through the use of animations; they also assessed students’ language learning with embedded quizzes and immediate feedback. Our study showed students generally engaged positively with the interactive videos and shifted from being passive viewers to active language learners due to the interactive nature. The videos were also deemed useful in helping students visualise and learn Japanese language in real-life situations, thus bearing pedagogical implications of these videos as pre-class learning materials.

AI in Education: Crafting Engaging Language Lessons through a Workshop for Educators
Jihee Im, Dini Arini, Hadir Alderaan

https://youtu.be/xSuXc1cCI0M

Discover the intricacies of crafting an AI integration workshop for secondary education pre-service teachers. We unveil our tailored approach to engage participants, delving into the AI landscape. Learn how we leveraged their background knowledge to explore AI\’s potential and challenges in education. Gain insights into our hands-on methodology, introducing AI tools and innovative apps like HeyGenAI and AiVooV, empowering pre-service teachers for effective classroom AI integration. Join us to explore our approach in fostering engaging, interactive, and effective AI integration for education.

The Affordances and Restrictions of Chatbots as Conversation Partners for Novice Learners
Michael DeSalvo

https://youtu.be/imyiSYNbVEM

Large Language Models (LLMs) such as ChatGPT or other \”AI\” Chatbots can produce coherent and grammatical text, but does this make them good proxy conversational partners for students in language courses? In this presentation, I will discuss some of the problems and possibilities my elementary French students and I experienced when using ChatGPT to practice written conversational French. I will share chat logs from the chatbot conversations, as well as student commentary. Participants will contribute to the discussion on how to best incorporate, if at all, LLMs to provide learners with additional, automatically-generated converational practice in language courses.

“Are we teaching language or technology?” Uncovering the Digital Disparity at Indonesian High Schools from the Theory of Practice Approach
Eugenie Mainake

https://youtu.be/Y2KkJ5glrnk

Indonesian EFL teachers are ready to enact technology-enhanced language learning yet they face challenges in their practices (Mainake & McCrocklin, 2021). Grounded on the theory of practice (Bourdieu, 1977), I seek to unpack the digital disparity of EFL teachers’ TELL enactment. With a qualitative approach, I interviewed EFL high school teachers (N=6) from a province in Indonesia. I discovered that various technologies were used in classes, while students’ limited digital skills and infrastructures in schools were the inhibiting factors to the digital disparity practices. My findings suggested integrating technology into pedagogy training for teachers and changes in EFL class policies.

English Language Learning in the Age of AI: Challenges, Opportunities, and Tools
Hadir Alderaan, Dini Arini, Jihee Im

https://youtu.be/fm36LtkIDuw

As artificial intelligence (AI) continues to evolve across numerous areas, including education, its role in education becomes increasingly visible and significant. This poster delves into the effect of the rapidly shifting landscape of AI in the field of English language learning (ELL). It specifically aims to shed light on challenges that both teachers and students may face while using AI, along with the possible benefits it can bring to the ELL field. Moreover, the poster provides practical examples of some available platforms and tools designed to support teachers and help learners with their AI-driven language teaching and learning goals.

Using Pangea Chat to Analyze Vocabulary and Grammar Trends in Spanish for Specific Purposes Courses to Foster more Learner Autonomy
Timothy Ashe, Jr.

https://youtu.be/vHpuuY4e2LM

Computer mediated communication (CMC) tools like Mobile Assisted Language Learning (MALL) are a common reality for most instructors to use in their courses. Additionally, Language for Specific Purposes (LSP) courses like Spanish for Business and Health are becoming more popular, especially at the lower level in large universities. Thus, language teaching and learning with technology should be driven by students’ linguistic needs (Rodgers & Dhonnchadha, 2018). This study analyzed how 30 students completed various chat assignments to assess how their learner autonomy and experiences were affected through AI interactions to perform SSP tasks.

Explore AI-generated Feedback for ESL Writing Based on Human Scoring
Biyin Xu

https://youtu.be/FNk1eG1IHO0

GenAI has great potential assisting teachers with providing feedback. While most studies focused on using GenAI to score writing assignments, few studies used it to provide feedback based on human scoring. This study explored whether GenAI can offer helpful feedback aligned with scores assigned by the teacher for intermediate ESL writing assignments. Ten essays with scores and a rubric were presented to ChatGPT-3.5 to generate feedback based on which students wrote revised drafts. The results showed the extent to which ChatGPT-3.5 is accurate and helpful in delivering feedback. Findings have implications for reducing teachers\’ workload with AI-generated feedback.

Specific-Purpose Language Learning through AI-generated Characters in a Learning-Oriented Assessment
Soo Hyoung Joo

https://youtu.be/JfJJHES0FrQ

To closely emulate classroom practices, a growing number of researchers (e.g., Banerjee, 2019; Purpura, 2021) have leveraged technology to design a type of learning-oriented language assessment where examinees work with simulated characters toward real-world goals. With the advent of AI-generative text-to-video technologies, these characters, which previously took the form of static images, can now take the form of synthesized human-like media (Pataranutaporn et al., 2022). Therefore, the current research investigated how AI-generated characters promote specific-purpose language learning. Preliminary findings suggest that the AI characters situated the examinees in an authentic context and provided affective support, promoting domain-specific language learning.

ROAR, An Augmented Reality App Applied to Reading Comprehension
Karen Villalba, Heidy Robles, Gustavo Chacon Contreras

https://youtu.be/FWtk5DdABGk

Colombian university students face great difficulties in understanding texts, especially those related to their discipline. Augmented reality has emerged as an innovative tool for teaching purposes in higher education contexts as it provides interactive visual experiences. This ethnographic study explores the design and implementation of an application of augmented reality called ROAR at a Colombian university with 50 medical students. This technological device aims to examine the benefits and limitations associated with its implementation based on the approach called Pedagogy of Textual Genres. The results revealed improvements in reading comprehension, the analysis of complex concepts, and the use of authentic examples.

Automatically Distinguishing between Written Output Produced by Heritage and Non-Heritage Learners of Polish as a Foreign Language
Simon Zuberek

https://youtu.be/00oeNrRMZVs

The presentation will showcase a study examining written work by students learning Polish as a foreign language, focusing on errors in Polish nominal morphology. Specifically, the study identifies three case error types common among heritage learners. Using these types along with per-word entropy values, a supervised machine learning classifier is developed to distinguish between compositions by heritage and non-heritage learners. Results demonstrate the effectiveness of these features in differentiating the two groups, highlighting that certain features are more influential than others in this task.

Harnessing the Potential of Speech Recognition Software (SRS): A Pathway to Assess and to Practice Sec- ond Language Pronunciation
Angel Añorga

https://youtu.be/ftGWInyxFHg

This session explains using Speech Recognition Software (SRS) to assess second language pronunciation in student voice recordings. It examines the role of SRS as an inter-rated reliability coder, elucidating its additional potential as a pedagogical tool to foster pronunciation development. From a cognitive phonology approach, this poster session underscores SRS\’s role in aiding learners to grasp phonological intricacies of the target language, such as perception and mental representation of sounds. The presenter will discuss the role of SRS in research and as a pedagogical tool.

Rate L2 Argumentative Essay Using GPT4: The Effect of Human-centric Materials
Deng Yicun

https://youtu.be/jionzbXMvyY

Large Language Models (LLMs) like GPT-4 are revolutionizing second language assessment. Researchers, aiming for effective L2 Automated Essay Scoring (AES), are integrating human-centric materials for calibration. Yancey et al. (2023) found GPT-4, when calibrated with human raters’ annotations, accurately scores short L2 essays. This study extended this method to longer L2 essays from a corpus, specifically those for TOEFL iBT. Results indicate GPT-4, using calibration examples and official rubrics, produced ratings comparable to traditional machine learning models’ performance. This underscores GPT-4’s potential in AES when guided by human-centric methods, warranting further exploration.

Enhancing Language Proficiency Assessment: Investigating the Impact of AI Assistance with ChatGPT on ACTFL Writing Proficiency Test Prompts
Alexander Tang

https://youtu.be/NKlUSVswyeg

This conference presentation explores the integration of AI assistance, with a focus on ChatGPT, within the context of Computer-Assisted Language Learning Drawing from Technology-Mediated Task-Based Language Teaching, our study investigates the impact of AI on the scoring consistency of ACTFL Writing Proficiency Test prompts and its effectiveness in improving language learners\’ writing skills. We also delve into the attitudes and perceptions of test-takers and evaluators regarding AI-assisted assessments within CALL. This research contributes to the evolving field of CALL by shedding light on the potential benefits and challenges of AI integration in language proficiency assessment and task-based language teaching.