Poster 4 (in-person)
Poster 4 (in-person)
Embedded Files
Kat Belfort, University of Oxford
Kat Belfort, University of Oxford
PhD Student
PhD Student
Kat is a DPhil candidate and Researcher at the Department of Education, University of Oxford. Her research interests centre on second-language learning, cognitive and linguistic processing, and the use of artificial intelligence in language education.
In her doctoral research, she investigates how AI-mediated multimodal input influences vocabulary learning across languages with differing orthographic transparency. Her broader academic interests include the neuro-cognitive foundations of language learning and the interaction between linguistic, cognitive, and social factors in education, complemented by a strong interest in data analysis and statistical modelling.
Kat is currently assisting Dr Woore and Dr Ramezanzadeh on Phase 2 of the Learning and Teaching Arabic as an Additional Language (AL2) project by carrying out systematic reviews of experimental/quasi-experimental studies on AL2 instruction and learning of vocabulary, and grammar. In her previous role she was working as a Researcher for a project led by Dr Ariel Lindorff that examines methodological rigour in studies of educational equity, with a particular focus on socioeconomic status.
She is also a part of the Nation-Wonnacott Language Learning Lab in Oxford’s Department of Education, an interdisciplinary research group conducting experimental work on language learning mechanisms including statistical learning, second/foreign language acquisition (across speech, vocabulary, and grammar), language learning technologies, and literacy development in children and adults.
KJB-poster_RGB_flattened_new map.pdfTitle: "Brain Decodes. AI Comments. Equity... Depends: Examining Incidental Vocabulary Learning Through AI-Assisted Loud Reading in Multilingual Contexts"
Author: Kat Belfort
Abstract: This poster presents a proposed quantitative study on incidental vocabulary learning in multilingual contexts, focusing on whether reading mode and AI-mediated feedback influence vocabulary gains. The project compares three conditions: silent reading, reading aloud, and AI-assisted loud reading, with a particular focus on learners whose language backgrounds differ in orthographic transparency. The study is grounded in the view that learning is a multi-layered process shaped by cognitive mechanisms, social context, and technological mediation, and that AI feedback must be interpreted critically rather than treated as neutral or fully objective. The research asks whether reading aloud outperforms silent reading for incidental vocabulary learning, whether AI feedback improves form and meaning retention, and whether working memory and first/target language mapping consistency moderate these effects. The design is cross-sectional, quantitative, and between-participants with random condition assingment, using vocabulary post-tests, working-memory measures, acoustic-prosodic data, and human-verified evaluation of AI feedback accuracy. Analyses will use a distribution-appropriate regression model. The study aims to contribute evidence on how reading mode, AI support, and orthographic background shape vocabulary learning, while also addressing equity, bias, and learner autonomy in AI-mediated language education.
Page updated
Report abuse