Assessing Pragmatic Competence for the Classroom: The Impact of Interaction in Computer-based Spoken Language Assessment

Assessing Pragmatic Competence for the Classroom: The Impact of Interaction in Computer-based Spoken Language Assessment

This study examines a central question in contemporary digital speaking assessment: whether computer-delivered speaking tasks can meaningfully simulate interaction and elicit features of interactional competence.

It focuses on digitally delivered speaking tasks that use pre-recorded prompts to simulate real-time interaction, comparing these with structured and freer live, human-delivered role-play formats. Using a mixed-methods design, the study analyses test-taker performance, rater behavior, and spoken interaction across task types, combining statistical analysis with close analysis of spoken discourse.

By bringing together psychometric, discourse-analytic, and stakeholder perspectives, the study contributes to ongoing discussions around digital speaking assessment and highlights the importance of assessment literacy in helping institutions understand how interaction is elicited and evaluated in online speaking tests.

Philip Horne, Head of Technical Content for English Language, Trinity College London