Assessing Turn-Taking in Paired Speaking Tests: Differences in Turn Allocation Practices in a Collaborative Task across Cambridge English™ Exams

Abstract

In language assessment, paired speaking tests have been used to elicit evidence of examinees' interactional competence. A notable example is in the Cambridge English™ Qualifications. While research suggests that interactional competence is comprised of a range of features, the rubric for scoring Interactive Communication on these exams only include generic descriptors about initiating, responding, and maintaining interaction with no attention paid to finer grained features like turn taking. This conversation analytic study addresses that gap by examining turn allocation practices among examinees during a collaborative task in paired speaking tests from the C2 Proficiency, B2 First, and A2 Key exams. The study focuses on three candidates who received the same Interactive Communication score across each exam. Extracts were transcribed from video recordings and analyzed for turn allocation practices with a focus on gaze, gesture, intonation, speech overlap, and pauses. Findings show that the C2 candidate used precisely timed verbal and embodied resources for turn allocation, while the B2 candidate relied on more explicit prompting and formulaic rising intonation and pausing and the A2 candidate displayed basic, delayed turn allocation constrained by disfluency. The study concludes by proposing descriptors related to turn allocation for the Interactive Communication rubric for each exam.