Debrief a rough lesson into one concrete fix for tomorrow

Debrief a rough lesson into one concrete fix for tomorrow

Most teachers do not get coached. The lesson you knew was rough on the walk back from class becomes "I will do it differently next time", and next time looks suspiciously like last time. This prompt is the 10-minute self-debrief that gets you to one fix, not five.

The prompt

You are an experienced [SUBJECT] teacher and instructional coach in an Australian [PRIMARY / SECONDARY / TERTIARY] classroom. You debrief lessons by separating what the evidence actually shows from how the lesson felt in the room.

The lesson I just taught: [ONE OR TWO SENTENCES naming the concept and what you intended students to be able to do by the end, e.g. "Year 8 maths, solving two-step equations with the variable on both sides. Goal was for students to leave being able to solve at least three independently."]

How long it ran: [MINUTES]
Class context: [ONE SENTENCE on the class, e.g. "8C, mixed ability, last period Tuesday, four EAL/D students, week 3 of the unit"]

The signals that told me it did not land (be specific, no spin):
- Student work or talk: [WHAT YOU ACTUALLY SAW, e.g. "Half the room stalled on the first independent question; the three who finished did not show working."]
- Exit ticket or last check: [WHAT THE LAST CHECK FOR UNDERSTANDING SHOWED, e.g. "5 of 24 got the exit ticket right; 8 left it blank; the rest had the variable isolated incorrectly."]
- The bit I cannot read confidently: [ONE SENTENCE, e.g. "The strongest student was quiet the whole lesson, which is not normal for her."]

What I had planned to do next lesson: [ONE SENTENCE, e.g. "Move on to word problems involving variables on both sides."]

Do the following:

1. Diagnosis. From the signals above, name the ONE most likely reason this lesson did not land. Pick from:
   a. Cognitive load too high (too many new moves at once).
   b. Prerequisite gap (a skill assumed from earlier was not there).
   c. The task was the wrong vehicle (students could complete the task without doing the thinking).
   d. Students did not have a reason to think (no genuine problem in the room).
   e. Transitions wasted the cognitive energy (too long modelling, too short doing).
   f. Other (name it in one sentence).
   Justify your choice in one sentence using the specific signals I gave you. Do not pick more than one unless two are genuinely tied; if they are, say which goes in tomorrow's fix and which goes in a small-group follow-up later in the week.

2. Two micro-fixes. Two concrete changes I should make BEFORE I teach this concept again. Each must:
   a. Be specific enough that another teacher could implement it without asking me a follow-up.
   b. Take under 10 minutes to prepare.
   c. Target the diagnosis you named in step 1, not a generic teaching tip.
   For each fix, name the cost: what I am trading off (time, scope, depth, student autonomy) to make this change.

3. What NOT to change. One sentence on the part of the lesson that DID work, that I might be tempted to throw out in the rewrite. Justify against the signals, not a hunch.

4. The next-lesson call. Based on the diagnosis, do I move on to what I had planned, or do I need to circle back tomorrow? If circle back, give me the smallest re-teach (under 12 minutes) that addresses the diagnosis without re-running the whole lesson.

5. Sanity check. One sentence on what would make you, the model, wrong about the diagnosis. If the signals I gave you are too thin to call confidently, say so and name the one extra data point I should grab in tomorrow's first five minutes.

Australian English throughout. Do not use the phrases "deep dive", "unpack", or "scaffold their thinking".

What it does well

Forces the model to pick ONE diagnosis from a fixed list, which is the move that stops the output reading like a general PD blog post and turns it into a teaching decision. The "name the cost" line in step 2 quietly bins the kind of micro-fix that sounds clever in a debrief and dies in a real period 3, because the cost is "another 40 minutes of prep you do not have".

Where it falls over

The model is reading the signals you write up, not the room you stood in. If your signals are vague ("kids seemed lost"), the diagnosis will be a horoscope. Push back if the call does not match your gut: ask the model what signal would change its mind, then look at the student work again. The prompt also cannot replace a real coach who saw the lesson, and it will sometimes name "cognitive load" when the real problem was that the lesson was last period on a wet Tuesday.

Try it like this

For a Year 10 science lesson on balancing chemical equations: paste the exit ticket results and the most common wrong answer. The diagnosis often surfaces as "task was the wrong vehicle", because students balanced by trial and error without ever holding the conservation of mass idea. The micro-fix is usually a constraint that forces them to count atoms before they touch coefficients.

For a VCE English lesson on close analysis: paste two student paragraphs, one strong and one stalled. The diagnosis tends to come back as "prerequisite gap", typically the metalanguage the unit quietly assumed. The micro-fix is a 4-minute vocab activation at the front of tomorrow's lesson, not another modelling round.

If you only have five minutes today, run it on the lesson you mentally flagged as rough on the walk back from class, not the one that was obviously bad. The obviously bad one you have already half-fixed in your head. The rough one is where the diagnosis earns its keep.

The lesson you debrief honestly on Wednesday is the one you do not have to re-teach in Week 8.