Identifying and fixing faults, without narrating every one
The pupil brakes late for the roundabout. Again. Third time this lesson, and the easy thing — the thing your mouth wants to do before your brain has caught up — is to say “brake earlier”, tick it off, and move on. You have identified a fault. You have named a remedy. And you have almost certainly wasted your breath, because next lesson the late braking will be back, wearing a slightly different hat.
Spotting the fault is the part everyone can do. Identifying faults accurately, finding what actually caused each one, and choosing a remedy that holds — that is the analytical core of instructing, and where good instructors quietly separate themselves from ones who just talk a lot. This guide is about that work: the difference between what you saw and why it happened, fixing the cause instead of the symptom, and feeding it back without turning the passenger seat into a running commentary. A companion to the craft of a good driving lesson and to the kerbside debrief.
The fault is not the cause
Here is the trap. Late braking is not a fault you can fix. It is a symptom — the visible end of a chain that started several seconds earlier, somewhere else in the car.
Ask why the braking was late and you get real answers, all different. The pupil was fixated on the roundabout and never scanned the approach, so the hazard arrived before their eyes did. Or they were in the wrong gear, unconsciously buying time to sort it out. Or their forward observation is short — they read ten metres ahead, not fifty, on every road, and the roundabout is just where it bites. Or simple speed misjudgement: they did not know how much road it takes to lose 20mph.
Four pupils, one fault, four causes. “Brake earlier” is a sensible instruction for none of them, because none chose to brake late — it was the only move left once the real thing had already gone wrong. Fix the braking and you have moved a bump under the carpet: the pupil who target-fixates will now brake earlier and still not see the cyclist, because you treated the timing and left the eyes alone.
“You braked quite late there — what were you looking at coming up to it?” “The roundabout, mainly. Watching for a gap.” “And the parked car on the approach?” ”…I didn’t really see it until I was on it.”
That is the cause surfacing. It was never about the brake.
Identify, analyse, remedy
The loop that gets you from symptom to cause has been ADI training bread-and-butter for years, and it is worth saying plainly because under pressure everyone skips the middle step. Identify, analyse, remedy.
- Identify. Name what you actually saw, concretely and without a verdict attached. “Late and hard on the brakes at the second roundabout”, not “poor hazard perception”. The first is a fact you both witnessed; the second is a conclusion you may be wrong about.
- Analyse. The step that earns your fee. Why did it happen? Resist the first explanation — it is usually the symptom restated. “Late braking because they braked late” is a circle. “Late braking because their scan finishes ten metres past the bonnet” is somewhere you can work.
- Remedy. Pick a fix that lands on the cause. For the short scanner it is not “brake earlier”; it is a lift-and-scan exercise that stretches their eyes further up the road. Different cause, different remedy, even though the fault looked identical.
Hold that middle step firmly. Most disappointing lessons are ones where the instructor ran identify → remedy and dropped analyse from the middle. It feels faster. It is slower, because you end up remedying the same fault six weeks running.
One-off, or pattern?
Before you spend a remedy on something, decide what kind of thing it is. The two demand opposite responses, and treating one as the other wastes a lesson.
A one-off is a slip. The pupil can do the thing, has done it all lesson, and dropped it once because a lorry surprised them. The right response is often to note it and say almost nothing. Make a production of a single slip and you teach the pupil that every wobble is a catastrophe, which breeds exactly the tension that produces more wobbles.
A pattern is a habit that has not formed yet — the same fault, in the same conditions, three or four times. Not bad luck; a gap in the skill, and worth stopping for. The DVSA’s own marking runs on this distinction: a repeated driving fault in one competency can be assessed as serious precisely because repetition reveals what a scattered handful of slips does not. Three mirror faults in half an hour is not three slips; it is one missing routine, and one thing to teach.
So hold your fire on a single occurrence and watch whether it recurs. Instructors who correct the first instance of everything spend the lesson chasing noise and never see the shape of the real problem.
Chase the primary fault, not all of them
Watch a nervous new instructor and you will see them try to fix everything they notice, in the order they notice it. It feels conscientious. It is the fastest route to a pupil who feels picked apart and learns nothing, because faults are not a flat list — they are a tree, most branches growing from one root.
The pupil who is a gear behind all lesson also shows up as hesitant at junctions, slow on progress, harsh on the accelerator, and poor at separating tasks — still fighting the gearbox when they should be looking. Five faults on your sheet, one cause. Teach the gear-planning and four of them quietly dissolve: they were never independent problems, just secondaries hanging off one primary fault.
So before you say anything, ask which fault, if fixed, makes the most others disappear. Teach that one and let the rest wait — the same discipline as picking one focus at the kerbside debrief, applied live at the wheel. The faults you consciously ignore are not neglect. They are triage.
Feed it back without the commentary
You have found the cause and picked the pupil who needs it. Now, how do you get it into their head — and the enemy is your own instinct, which is to tell. Telling produces the running commentary: a verbal feed of every fault as it happens, flooding the pupil’s attention exactly when they need it on the road, and training them to drive by your voice rather than reading the road.
Client-centred learning is the fix, and fault feedback is where it most obviously pays. Instead of announcing the fault, ask a question that makes the pupil find it:
“How did that gear feel coming out of the junction?” ”…Bit laboured, actually. I think I should’ve been in second.” “So what’ll you try next one?”
The pupil who noticed their own high gear will change it next time; the pupil who was told will change it when you tell them, and not otherwise. Same information, opposite ownership. A fault the pupil identifies themselves is halfway remedied before you have said a word about the fix.
Two rules keep this from becoming its own kind of chatter. Save most of it for when the car is stopped. A moving car is a bad classroom; the real analysis belongs at the next safe pull-up or the kerbside debrief, where you can draw the exact moment the fault happened — the junction, their car, the gear — instead of describing it into a windscreen. (BitePoint’s fault-to-fix cards exist for exactly this.) And grade severity honestly, so the pupil knows what matters. Borrow the DL25’s fault categories as a shared vocabulary — the wide-but-safe turn is a driving fault you might not even stop for; the missed observation that made someone brake is a serious one you stop for now. Treat the trivial and the dangerous with the same grave tone and the pupil calibrates to the wrong things.
The exception overrides all of it: if it is a safety issue, intervene immediately. You do not run a Socratic dialogue while a pupil drifts toward a cyclist. Act — verbally or on the dual controls — make the car safe, then analyse afterwards. Let-it-run-and-discuss is for teaching; immediate intervention is for danger. Knowing which is which, in the half-second you get, is a large part of the job.
Common mistakes when identifying faults
Every one feels reasonable in the moment, which is exactly why they persist:
- Over-talking. Narrate faults as they happen and the pupil listens to you instead of the road, learning to drive by your voice. Silence, then a question at the next stop, beats a live feed every time.
- Listing every fault. The instinct to be thorough produces a pupil who feels shredded and remembers none of it. Pick the primary fault; the secondaries will keep.
- Fixing the symptom. “Brake earlier”, “check your mirrors”, “don’t stall” — aimed at the visible fault while the cause sits untouched. It comes straight back because you never went upstream.
- Correcting one-offs. Making a lesson of a single slip trains anxiety, not skill. Watch for the pattern before you spend a remedy.
- Never checking the remedy worked. You taught the gear-planning; did the late braking actually go? A remedy you never verify is a guess you filed as a fact. Contrive the situation again later and watch. Fault gone, analysis was right. Fault still there, your cause was wrong — back to the middle step.
The habit worth building
Strip all of this back and it is one small internal routine, run every time something goes wrong: what did I see, why did it happen, one-off or pattern, root or branch, now or at the next stop? Five questions, a second or two, before you open your mouth. The gap between an instructor who runs that loop and one who blurts “brake earlier” is the entire craft.
And there is a longer game in it. The pupil constantly told their faults learns to wait for the telling; the pupil asked why do you think that happened starts running the loop on themselves — catching the wrong gear before you mention it, feeling the late brake coming. Which is the whole point, because the day after they pass there is nobody in the passenger seat to identify anything for them. The faults will still happen. You are just teaching them to be the one who spots the cause.