Ways To Measure Teaching | #2 Categorise teacher talk into functional types

10 Practice Ways To Measure Teaching

#2 Categorise teacher talk into functional types

1️⃣ The data surprised me again

After converting three one‑to‑one lessons (114.5 minutes) into transcripts, I calculated my actual Teacher Talking Time:

TTT = 62.6 minutes
STT = 51.9 minutes

So yes — I spoke slightly more than my learners.

But the transcripts showed something more interesting than the percentage.

2️⃣ TTT isn’t one thing

When I coded my teacher turns, I realised they weren’t all the same.

My talk fell into four clear functions:

1. Modelling
Pronunciation, grammar forms, example sentences, discourse patterns.

2. Clarifying
Explaining rules, unpacking vocabulary, resolving confusion.

3. Diagnosing
Asking questions that reveal what the learner understands.

4. Recasting
Rebuilding the learner’s output into a clearer, more accurate version.

This is the part that changed my mind:

Most of my TTT wasn’t “talking at the learner”. It was instructional work.

3️⃣ The distribution was revealing

Across the three lessons, my teacher talk roughly broke down like this:

Modelling: ~30%
Clarifying: ~25%
Recasting: ~20%
Diagnosing: ~15%
Other: ~10%

So although I spoke for 62.6 minutes, over 90% of that talk was functional — not filler, not anecdotes, not teacher monologue.

4️⃣ Why this matters

If we only measure TTT as a percentage, we miss the real question:

👉 What is the teacher actually doing while talking?

A 10‑second recast can reshape a learner’s grammar.
A 20‑second clarification can unlock a concept.
A 5‑second model can fix a fossilised pronunciation error.

Time is not the metric.

Impact is.

5️⃣ A better way to think about TTT

Instead of “How much do I talk?”, a more useful question is:

👉 What kinds of talk does my teaching rely on — and why?

That’s where the insight lives.

6️⃣ Question

If you analysed your own lessons, what functions would your teacher talk fall into?