Formative assessment strategies that change what your students submit next

Search for formative assessment strategies and you get techniques. Exit tickets. One-minute papers. Thumbs up, thumbs down. All of them useful. None of them a strategy.

A technique tells you something about the room. A strategy decides what you will do about it — and decides it before you look. Skip that second half and you spend a semester collecting evidence, then teach exactly the course you would have taught anyway.

That gap is where formative assessment usually dies. Not in the checking. In the acting.

Decide the branch before you run the check

Write the consequence into the plan, next to the check itself:

If more than a third of the room misreads the cash-flow question on Tuesday, Thursday becomes a worked-example session and the case study moves a week.

Decide that on Tuesday afternoon and you will decide against it. The syllabus is already behind, the room is tired, and the easy call is always to carry on. Decide it on Sunday, when the cost is abstract, and you will keep it.

It also makes the check honest. You are no longer asking "did they get it?" and hoping for yes. You are asking a question you have already agreed to act on.

The test every strategy has to pass

One question, and it is unforgiving:

If this came back bad, what would change?

• Nothing changes → it is decoration. Drop it and buy back the ten minutes.
• Only your notes change → it is research about your students, not teaching.
• Something in the next class, or in their next piece of work, moves → that is a strategy.

Most of what gets called formative assessment fails this test. Run it over your own course and it will tell you which checks to keep.

Four that work at course level

1. Chain the checks so each one opens the next

A single check tells you where the room is, once. A chain tells you whether your fix worked.

Check, teach to the gap, then check the same idea again in a different shape. Three light touches on one concept beat one heavy check across three. If the second touch comes back clean, you have evidence your intervention did something — which is more than most courses ever get.

2. Leave a reuse window

Feedback they cannot use is a report card. Before you set a check, look at the calendar and ask two things: do they have another piece of work to apply this to, and do they have time to apply it?

If the answer to either is no, move the check earlier. This is the same rule that separates feedback that changes the next piece of work from feedback that only explains a grade — and it costs nothing but planning.

3. Keep it out of the grade, but not out of consequence

Make a check count towards the mark and students optimise for the mark. They show you their best guess instead of their real confusion, which destroys the only thing the check was for.

Give it no consequence at all and they skip it.

Consequence is not the same as grade. A redo. Explaining their reasoning to the person next to them. A task on Thursday built directly on what they got wrong on Tuesday. All of those make the check matter without turning it into an exam.

4. Make the room do the reading

You cannot mark forty scripts a week, and the course that tries stops by week six.

Design checks that surface the answer without you taking anything home: peer explanation, where you listen rather than read; a show of reasoning on the board; a poll on the misconception you already know is coming. When you do need written work, sample it. Twelve scripts tell you what forty will tell you, in a quarter of the time.

What it actually costs

Be honest about this, because the usual advice is not. Each check costs you class minutes, planning time, and the discipline to follow through. The planning is where it lives or dies: a check without a prepared branch is just a pause in the lesson.

The cheapest version looks like this, per unit:

1. One check, ten minutes or less, inside a class you are already teaching. 2. One branch, written down in advance — what Thursday becomes if the result is bad. 3. One follow-up, on the same idea, in a different shape.

That is the whole strategy. Everything else is elaboration.

Start with one unit

Not the whole course. Pick the unit your students have struggled with for three years running — you already know which one it is. Put one check in it, write the branch next to it, and prepare the follow-up task now, while you still have time.

Then run it, and notice how different the next class feels when you know what is actually in the room.

Preparing that branch is the slow part: a worked example you did not plan to build, a follow-up task on the same outcome in a different shape, a rubric that holds up when you use the check for real. Quindaria builds both from your own course — your description, your outcomes, your topics — so the backup plan you decided on Sunday exists by Thursday.

Try it free: quindaria.com