Formative assessment PD: how to plan a day that sticks.

By Yechiel · Founder, PD That Works. Yechiel builds PD That Works — the workshops, the daily AI coach, and the Day 30/60/90 reports — and wrote every page on this site.

Formative assessment is the most-requested PD topic in most schools and one of the least well-taught. The reason isn't the content — teachers generally accept the idea within ten minutes. It's that a day about checking for understanding usually ends with teachers holding a new technique and no plan for the Tuesday. Here's what the research actually supports, the moves worth a whole day, where these workshops reliably fail, and how to plan one that's still visible at Day 30.

What formative assessment actually is

The useful definition is narrower than the phrase suggests: an assessment is formative only if the information changes what happens next. A quiz that produces a grade and no adjustment is summative regardless of what it's called. An exit ticket read on the drive home and never acted on isn't formative either — it's just late.

That single test — did it change the next lesson? — is the whole idea, and it's the one thing a PD day has to make impossible to forget.

What the research supports — and the argument about it

The foundational work is Black and Wiliam's 1998 review, which surveyed some 250 studies and concluded that classroom formative assessment produced substantial learning gains (Assessment and Classroom Learning, Assessment in Education 5(1), 7–71; the companion piece is Inside the Black Box, hosted copy). Effect sizes of 0.4 to 0.7 entered the vocabulary of PD slide decks and never left.

Those numbers deserve a caveat that slide decks omit. A later meta-analysis by Kingston and Nash reviewed more than 300 K–12 studies and found only 13 reported enough information to compute an effect size — from which they calculated a weighted mean effect of about 0.20, with a median around 0.25 and notable variation by subject (roughly 0.32 in English language arts, 0.17 in mathematics, 0.09 in science). Their title says it plainly: a meta-analysis and a call for research.

The honest reading for a school leader: formative assessment is a real effect, worth teaching, and smaller and more variable than the number on the slide. That matters for how you plan the day — a workshop promising transformation will disappoint, and a workshop teaching four moves teachers can run next week will not.

The four moves worth a day

You can teach a great many techniques badly or four well. These four cover most of what a teacher needs, and each can be practiced in a room:

Rapid checks. Whole-class signals that take under a minute — hand signals, whiteboards, a four-corner question. The skill isn't the technique; it's designing a question whose wrong answers are informative, so a show of hands tells you which misconception is in the room.

Responsive exit tickets. The exit ticket is common; the responsive part is rare. The move is deciding, before the tickets are collected, what you'll do with each likely pattern — an opener, a warm-up, a short conference list.

Callbacks. Returning deliberately to something a check surfaced days earlier. This is the move that turns a one-off check into evidence of change over time, and it's the one teachers most often skip.

Participation protocols. Structures that make the check reach everyone rather than the four students who always answer — cold-call with prep time, two voices before mine, think-pair-share with an accountable share.

Notice what isn't on the list: any specific app. Tools are fine, but a PD day built around a platform teaches the platform, and platforms change.

Where formative assessment PD usually fails

Three patterns, all recognizable:

The technique becomes the point. Teachers leave with fourteen strategies and no principle. By November they're using none, because fourteen is not a change, it's a menu.

Data gets collected and never acted on. This is the most common failure and the most dispiriting: teachers dutifully gather exit tickets that change nothing. As our FA-101 kit puts it, a ticket you do not act on is just a goodbye wave with data on it.

The day ends with enthusiasm and no commitment. No specific goal, no one asking on Tuesday. This is the general case of why PD doesn't stick, and formative assessment is unusually vulnerable to it because the techniques are easy to admire and easy to drop.

Planning the day

A shape that works for a half-day with 25 teachers:

1. Open with the problem, not the solution. Teachers have to feel the void — teaching into a room without knowing what landed — before a technique means anything.
2. Study two or three cases together. Analysis before practice. Keep it screenless if you can; the room's attention is worth protecting.
3. Practice the moves on your feet. Stations or rotations, short enough that everyone runs each move rather than watching one demo. This is the part that gets cut for time and shouldn't be.
4. Build something team-sized. A one-page playbook of two or three moves the team will actually use, made in the room.
5. End with one commitment each. One move, one class, one date — plus who's going to ask about it. Not a theme. Something a teacher can tell on Tuesday afternoon whether it happened.

The last step is the one that determines whether any of the rest survives.

What happens the Monday after

A formative assessment day sets a coachable goal almost by construction — "two voices before mine, in every discussion, fifth period" is exactly the kind of thing daily follow-up can ask about. Five minutes a day, tied to the commitment each teacher made, is what turns four moves into practice.

And it gives you something to show. Engagement with the follow-up, movement on the goals teachers set, and the themes emerging across the faculty are the evidence a board asks for at Day 30 — how to measure whether the PD worked covers what that looks like and what it honestly isn't.

If you'd rather not build it

We sell the day already built: FA-101, Whose Data Is It Anyway? — a 3.5-hour workshop where improv comedy meets formative assessment, screenless by default, with a scripted facilitator guide, a 41-slide deck, ten handouts, and four station cards. It ends the way the plan above ends: each teacher names one Monday Move with a date and an accountability partner. $1,500 for a cohort of up to 25 teachers, with 30 days of platform follow-up included.

You do not need to buy anything to run a good formative assessment day. The outline above is the honest version of what we teach.

Common questions

What is formative assessment?

Assessment used during learning to adjust what happens next, rather than to grade what already happened. The practical test is whether the information changed the next lesson: a quiz that produces only a grade is summative regardless of what it's called, and an exit ticket nobody acts on isn't formative either.

Does formative assessment actually improve achievement?

The evidence supports a real but modest effect. Black and Wiliam's 1998 review of roughly 250 studies reported substantial gains and is the source of the widely quoted 0.4–0.7 effect sizes. A later meta-analysis by Kingston and Nash found only 13 of more than 300 K–12 studies reported enough detail to compute effect sizes, and calculated a weighted mean of about 0.20, varying by subject. Worth teaching; not a transformation.

What should a formative assessment PD day cover?

Four moves are enough for a day: rapid whole-class checks, responsive exit tickets, callbacks to earlier checks, and participation protocols that make checks reach everyone. Cover them with practice rather than demonstration, and end with each teacher naming one specific commitment for the following week.

How do you make formative assessment PD stick?

End the day with one specific commitment per teacher, then have something ask about that commitment inside the teaching week. Techniques presented without follow-up are forgotten within weeks; a specific goal plus a daily check-in is what turns a workshop into changed practice by Day 30.