Dr Ying Zheng

• Psychometrics and test validation; Human scoring vs. machine scoring; Statistics in applied linguistics; Comparative Judgement in language testing
• Mandarin exams in the UK school system, including A-Level and GCSE exams; Mandarin Chinese teachers’ professional development
• Learner motivation and language teaching pedagogy; ESL/EFL learner characteristics and test performance

Project title:  From GSE to GSL: Developing a Global Scale of Language Competence

Funded by: AHRC IAA Secondment

The goal of this secondment is to contribute to a collaborative project on exploring the feasibility of expanding from Global Scale of English (GSE) to Global Scale of Language Competence, including Global Scale of Spanish (GSS) and Global Scale of Chinese (GSC).

Project title: Towards an Automated Scoring Model of Spoken English Language Proficiency

Funded by: Interdisciplinary Research Pump-priming Fund, University of Southampton 

The primary goal of this project is to explore an automated scoring model for the evaluation of student’s English language proficiency, focusing on spoken language proficiency.

Project title: Exploring AI application in essay-based assignments: affordances and risks

Funded by: CHEP Enhancement Fund, University of Southampton

This study aims to compare the human marking and feedback practices with AI marking and AI-enhanced feedback provision in university essay-based assignments.

Project title: Pearson Test of English Academic next phase: Perspectives from trait marking and item type efficiency

Funded by: Pearson Education

This project consists of three stages. The first stage is to review current language testing research to identify alternative task types which are suitable for the assessment of academic English. Focusing on assessment of spoken language in authentic contexts through single or integrated skills assessment. The second stage is to further elaborate on a study of single trait versus multi trait assessment of language ability. The third stage is to conduct a detailed analysis using recent data to evaluate item type efficiency for PTE Academic. The aim of the project is to gather literature information and empirical data from the perspectives of trait marking approaches and item type efficiency indices to inform the design of the next phase of PTE Academic.

Project title: Towards a new Psychometric model for a test of general English proficiency

Funded by: Pearson Education

This project will contribute to the wider field of language testing and assessment, particularly the field of English language assessment. The project involves detailed psychometric design work to build a matrix of data which will be used to calibrate test items and build a psychometric model to relate those items to an existing scale - Global Scale of English - and contribute to the new development of Global Scale of English General.