Coaching conversation models — OSCAR, CLEAR, GREAT and TEB

GROW — Goal, Reality, Options, Will — is the model most ADI training starts with, and for most lessons it is all you need. (You will see the W written as either Will or What's next; they mean the same thing — the commitment to a specific action.) It is worth knowing three close relatives, though, because each one puts its weight in a slightly different place, and because Part 3 rewards an instructor who can visibly choose an approach rather than run the same script at everyone.

OSCAR — Outcome, Situation, Choices, Actions, Review

• Outcome — what does the learner want to be true at the end?
• Situation — where are they now, honestly?
• Choices — what could they do? Generate several before judging any.
• Actions — which one are they actually going to do, and when?
• Review — how and when will you both check it worked?

OSCAR is GROW with a stronger finish. The separate Review step makes it a good fit for a goal that spans several lessons — a subject the learner is building over time rather than something you will settle before the end of today's drive.

CLEAR — Contract, Listen, Explore, Action, Review

• Contract — agree up front what this conversation is for and what each of you is responsible for.
• Listen — genuinely, and before you steer anything.
• Explore — open up the learner's own understanding of the issue.
• Action — agree what changes.
• Review — check back.

The distinctive step is Contract. Setting out at the start who is doing what — "I'll handle the mirrors on this first one, you take the positioning" — removes an enormous amount of the confusion that gets marked down as poor risk management. CLEAR is the model to reach for at the very start of a relationship with a new learner, or when a lesson has been drifting and you need to reset expectations.

GREAT — Goal, Reality, Explore, Action, Timing

• Goal — what are we aiming at?
• Reality — where are we now?
• Explore — what are the options?
• Action — what will you do?
• Timing — by when, and for how long?

GREAT is GROW with the clock added. Timing is useful in a driving lesson in a very practical way: agreeing "let's give this six attempts" or "we'll try it your way for the next ten minutes" bounds an experiment so that neither of you has to decide mid-attempt whether it is working. It also gives you a natural, non-critical moment to stop and reflect.

TEB — Thoughts, Emotions, Behaviours

TEB is not a conversation structure; it is a diagnostic, and it belongs underneath all of the above. The idea is that the behaviour you can see from the passenger seat is the last of three things, and usually the least useful one to talk about:

• Thoughts — what will you be thinking?
• Emotions — what will you be feeling?
• Behaviours — what will you be doing?

A learner who freezes at a busy roundabout is not exhibiting a steering problem. Correct the behaviour and it comes back next week; reach the thought ("everyone behind me is getting annoyed") and the emotion (pressure, then panic) and you are working on the thing that actually produced it. This is sometimes described as the learner's unconscious co-pilot: the running commentary and feelings that shape decisions before any conscious choice is made.

Two ways to use it:

• Backwards, after the event. "When you stopped at the line there — what was going through your head? How did that feel?" Then link it to what the car did.
• Forwards, before the event. "We're about to head onto the dual carriageway. What do you think you'll be thinking? What might you feel?" Naming an emotion in advance takes a surprising amount of its force away, and gives you both a shared word for it when it arrives.

TEB pairs naturally with the top level of the GDE Matrix — the personal factors and attitudes that sit above vehicle control — and with the emotional-state triggers in adapting the lesson.

Choosing between them

You do not need to announce which model you are using, and you should not try to run all of them. A reasonable default:

What the examiner is looking for is not a named model. It is that the goal was the learner's, that you found out where they really were before you taught anything, and that you agreed what happens next together — which every model on this page is a way of remembering to do.