Validity argument for the use of summative task-based language assessment in a language teaching program for adult immigrants

Abstract

This study investigates how a collaborative, argument-based validation process can support the development of valid, context-sensitive task-based language assessments (TBLAs) while simultaneously fostering teachers’ assessment literacy and competency. Conducted within Quebec’s adult francization programs, the study involved francization teachers in the co-design, piloting, and iterative refinement of 19 summative assessment tasks targeting real-world scenarios such as job interviews and healthcare interactions. Guided by Kane’s (2013) argument-based validation framework, the process emphasized classroom-based inquiry, reflective practice, and local pedagogical relevance. Drawing on logbooks, focus groups, and observation grids, the findings reveal how engaging with validation concepts — such as domain definition, scoring, and generalization — enabled teachers to make principled decisions about task design and implementation. The collaborative model promoted a reconceptualization of assessment as an integrated component of teaching, enhancing teachers’ confidence, agency, and understanding of communicative competence. This study contributes to both the TBLA and teacher education literature by demonstrating how structured, practice-embedded collaboration can serve as a powerful vehicle for professional learning and sustainable assessment design.