Fluent but Off-Task
In a B1 paired speaking task, candidates must discuss options for a class trip and reach a decision in three minutes. A candidate, Dmitri, speaks fluently and at length. He gives a two-minute monologue about a trip his family took, uses good past tenses, and never asks his partner anything. The partner says six words in total. The assessor must score Dmitri on interactive communication and on task achievement. Dmitri's grammar and vocabulary are strong for B1.
1. Separate what Dmitri did well from what he failed, and say which scoring criteria are affected.
Read the worked answer
Well: fluency (sustained speech without long pauses), grammar range (narrative past tenses used accurately), vocabulary. These are real strengths and should be credited on those scales.
Failed: interactive communication - no initiations to the partner, no questions, no turn-taking, no response to the partner's contributions. Task achievement - the task was to discuss options and reach a decision; he narrated instead, and no decision was reached. His monologue is relevant in topic but not in function.
Key point: fluency is not interaction. A long turn with no exchange scores high on fluency and low on interactive communication - these are separate scales and must not contaminate each other. Do not let strong grammar pull up the interaction score.
2. The partner's near-silence might tempt an assessor to blame the partner. How should the assessor decide, and what would change the judgement?
Read the worked answer
The assessor should ask: did Dmitri create openings? Interaction is co-constructed, but the candidate who dominates bears responsibility for the missing exchange. Evidence Dmitri did not: no questions, no pauses for response, no acknowledgement of the partner's six words. That points to Dmitri.
The judgement would change if the recording showed Dmitri offering turns - pausing, asking 'what do you think?', inviting agreement - and the partner declining or failing to take them. Then the partner's silence is the partner's issue, and Dmitri's interactive score should reflect his (unanswered) initiations, since he demonstrated the behaviour being assessed.
Practical rule: score the behaviour the candidate produced, not the outcome that depended on both. Dmitri produced monologue behaviour, so the low interaction score stands.