Session B · Rehearsing error handling (facilitator guide)
60 minutes · Week 2 · Prerequisites: Modules 3–4 + transcript practice (spot the move) complete · 6–14 participants · Materials: role-play cards (Round 2 deck), hint-ladder cards, observation rubric, timers
Objective
Every participant responds to at least three scripted student errors with a probing question that exposes the misconception — not a correction — while holding 3–5 second wait time, and climbs the hint ladder one rung per turn. Observers score rubric Domain 3 and call “seconds of silence” out loud.
Why errors get their own session
Error handling is the highest-leverage, most counter-instinctive skill in the repertoire: the tutor’s every reflex — explain, fix, rescue, reassure — is wrong, and the correct behavior (a question, then silence) feels passive and even rude. That combination — high leverage, anti-instinctive — is exactly what rehearsal is for. Reading Module 3 is not enough; T2 shows it; this session makes the mouth do it.
Room setup
- Groups of 3 again, roles rotating; this session’s student cards carry misconception profiles (M1–M5 from the misconception bank): the student-role-player has a script of how they got stuck and a “reveal” they give only if the tutor’s probing question actually lands on it.
- Observers get two jobs: score Domain 3 rows, and count silence (say “wait” silently, fingers up per second — visible to the tutor, brutal and effective).
- Hint-ladder cards (L0–L5) face up on the table; tutors must place their hand on the rung they are about to use before speaking. Physicalizing the ladder breaks the “skip to the answer” reflex faster than any reminder.
Agenda (60 minutes)
| Minutes | Block | What happens |
|---|---|---|
| 0–5 | Frame | Objective; read one exchange each from T2 (“what the ladder sounds like”) and T6 (“what the takeover sounds like”). Establish the round structure. |
| 5–15 | Round 1 · one error, full ladder | Tutor gets a student card with ONE misconception (e.g., M3: 2(x − 4) = 2x − 4). The student commits the error and responds only to questions the tutor asks. Observer calls rung numbers and seconds of silence. Goal: student finds the fix before rung L3. |
| 15–22 | Debrief round 1 | Did the probing question land on the misconception, or nearby? Play back the exact wording. Re-run ONE exchange with a sharper question, same roles. |
| 22–34 | Round 2 · pressure test | Harder cards: student says “just tell me,” “I’m bad at math,” or goes silent. Tutor must hold the frame: acknowledge, stay on the ladder, never hand the answer. Observer watches for the two failure modes: rescuing and lecturing. |
| 34–42 | Debrief round 2 + micro-teach | Facilitator models the two hardest lines live: the “just tell me” response (“Let’s test it — substitute your guess back in; what do you get?”) and the stuck-silence response (drop a rung, never the answer). |
| 42–52 | Round 3 · chain drill | Quick-fire reps, 90 seconds each, rotating tutors: one error → probe → student reveals → confirm and ask why → offer a similar problem. Three moves in one chain; observers tick the chain on the rubric. |
| 52–58 | Whole-group debrief | Questions below. Commitment: each tutor names the ONE response they will rehearse in their head during real sessions this week. |
| 58–60 | Close | Self-reflection; facilitator logs rubric sheets for the Week-3 baseline (measurement plan). |
Observer checklist (Domain 3 + Domain 4) — print per round
| Look for | R1 | R2 | R3 |
|---|---|---|---|
| Asked before explained (any question within first 10 seconds) | ☐ | ☐ | ☐ |
| Probing question aimed at the misconception’s spot, not the label | ☐ | ☐ | ☐ |
| Held 3–5 seconds of silence (observer counted) | ☐ | ☐ | ☐ |
| One rung per turn — hand visibly on the ladder card | ☐ | ☐ | ☐ |
| Never stated the corrected step; never said the misconception’s name | ☐ | ☐ | ☐ |
| After correct answer: confirmed + ONE why-question | ☐ | ☐ | ☐ |
| Offered a similar problem (small variation) | ☐ | ☐ | ☐ |
| Errors normalized warmly (“that’s the trap this one sets”) | ☐ | ☐ | ☐ |
Debrief questions (pick 3–4)
- What did it take to not say the answer? (Name the feeling — the itch — so tutors recognize it mid-session.)
- When your probing question missed, what did it aim at? (Usually the symptom, not the misconception. The card “reveals” exist to train aim.)
- Students: what did the silence feel like from your side? (Almost always: shorter than the tutor feared. This single exchange changes wait-time behavior faster than any rule.)
- Which misconception was hardest to probe, and what question would you try next time?
Online delivery variant
- Breakout rooms of 3; the ladder “cards” are a pinned message; observers type the rung number in chat as the tutor speaks (creates a rung-by-rung record for the debrief — better than verbal calls on mute).
- Silence reads differently online: tutors should mute their mic during wait time so they can’t accidentally fill it, and say “thinking time” before it — the phrase is a legitimate rehearsal of the real-classroom version.
- Captions on; misconception cards and the checklist distributed in advance; breakout timers posted by the facilitator.
- The “reveal” mechanic works on video: student role-players type their reveal only when the probing question lands on it.
Facilitator notes
- The counter-instinct is the whole game. Expect and welcome bad rounds in Round 1 — the debrief exists because the takeover reflex is universal, not because anyone is weak.
- If a tutor can’t find any probing question, give them the stem, not the question: “When you [moved / divided / distributed] …, what happened to …?” Stems are on the back of the ladder cards.
- Keep the math fixed (one planned skill per round). This session trains response, not content knowledge — Session A owns the routine, the coaching cycle (coaching-cycle.md) individualizes it later.