Nine Weeks, One Coach, Six Classrooms

Do the arithmetic first; the superintendent's ask is not fundable with the hours available.

• Capacity: 12 h/week × 9 weeks = 108 h. Reserve roughly 20% for communication, data entry, and reporting, leaving about 85 coaching hours.
• Unit cost: a coached cycle (observation, feedback, joint planning) runs about 2.5–3 h per teacher. Six teachers × three cycles is 45–55 h - feasible across nine weeks, but not in three.
• Weeks 1–2: baseline observation in all six classrooms, agree on two observable target behaviors with each teacher, build the fidelity checklist.
• Weeks 3–4: School A's two requesting teachers start weekly cycles (1.8 referrals/100/day justifies the intensity). School B one teacher biweekly. School C one teacher biweekly at lower intensity - its rate is a quarter of A's.
• Weeks 5–8: rotate the remaining teachers, protecting School A intensity.
• Week 9: maintenance checks, evaluation, written handover.

What I decline: "all six trained by week 3." A three-week sprint buys exposure, not skill. I would rather take four classrooms to fidelity than six to a certificate. I put that trade-off in writing with these numbers so the decision is the superintendent's, made with the cost visible.

Training is the smallest part of the plan. It specifies:

• Operational definitions of the two target student behaviors and the three target teacher behaviors, so observation is not opinion.
• Measurement: 20-minute partial-interval observations twice weekly, with interobserver checks on 20% of sessions and an ≥80% agreement criterion.
• Fidelity checklist with a scored criterion and a decision rule: below 70% on two consecutive observations → re-teach that component, not the whole package.
• Feedback schedule and a named deliverer for each phase, including a building-level peer who takes over after week 6, so the effect survives my contract ending.
• Generalization and maintenance: checks at 4 and 8 weeks post-coaching, plus a plan for staff hired mid-year.
• Roles, cadence, escalation: biweekly one-page update to each principal; a defined path if a teacher refuses coaching.
• Evaluation: pre/post referrals per 100 students per day, interpreted with the explicit caveat that referral rate also reflects administrative practice and can fall simply because a principal changes the form.

Treat the skeptic as a source of information and a test of the plan, not as an obstacle to route around.

• Meet privately before any group training. Ask what she has seen fail before and what she would need to see to be convinced. Her answer usually names a real design flaw - often that previous initiatives were training-only with no follow-up, which is exactly the risk here.
• Use her as a baseline first. Observe her classroom early and give her feedback before her colleagues get any. Early, specific, private feedback converts more skeptics than persuasion does.
• Offer a bounded agreement: three coached cycles, then a joint review of her data. If her data do not improve, that is information about my plan, and I will say so out loud.
• Bound the cost. Do not spend hours disproportionate to the other five classrooms, and do not let one teacher's stance become the project's weather. Set a week-4 checkpoint with the principal.
• Credit her publicly if her classroom improves. Peer influence is the cheapest implementation support available, and it is the one the district keeps after I leave.