The Speaking Assessment Problem
The Four Standard Assessment Dimensions
How smoothly does the speaker communicate? Not just speed — but absence of hesitation beyond natural pause, coherent flow of ideas, and ability to self-correct without losing communicative momentum.
Does the speaker use appropriately varied and precise vocabulary? Can they paraphrase when they lack a specific word? Do they make lexical errors that impede comprehension?
Do they use varied grammatical structures appropriately for their level? What is the density and impact of grammatical errors on communication?
Are they understood without undue effort? Not: are they accent-free. Assessment of pronunciation should be intelligibility-based, not native-likeness-based.
Assessment Dimensions
Fluency
Smooth, natural delivery — not speed, but communicative flow
Vocabulary
Range, precision, and ability to paraphrase when words are missing
Grammar Range
Variety and accuracy of structures — at appropriate level expectations
Teacher Tip
“Record speaking samples (with the student's agreement) at the start of a course and every 6 weeks. Play old recordings together: 'Do you notice what's different now?' Students often can't perceive their own progress day-to-day. Hearing the difference between month 1 and month 4 is profoundly motivating.”
Frequently Asked Questions
Should I correct grammar errors during speaking assessment?
During a formal speaking assessment, no. Note errors, assess performance, and provide feedback afterward. Correcting during assessment changes the task conditions and prevents you from observing natural performance.
How do I create a simple speaking rubric?
Three levels (emerging / developing / proficient) × four criteria (fluency, vocabulary, grammar, pronunciation) gives a 12-cell rubric that's simple enough to use in real-time and detailed enough to be meaningful. Describe each cell with one specific observable behaviour.
Do holistic or analytic rubrics work better?
Analytic rubrics (separate criteria) provide more diagnostic feedback — better for teaching. Holistic rubrics (single impression) are faster — better for high-volume assessment. For private tutoring, analytic rubrics deliver more value.