Formative doesn’t mean “before the test.” It means during formation, while understanding is still being shaped. Most classrooms use formative feedback as a synonym for low-stakes. Research from two university classrooms shows what changes when you treat the timing literally.
The difference isn’t in how teachers give feedback. It’s in when it reaches students relative to the learning moment. Get that wrong, and even well-designed formative feedback strategies don’t move grades the way they should.
This post covers the strategies that hold up, what two real classroom studies found about why timing is the variable most teachers overlook, and how to redesign your feedback so it lands inside the work rather than after it.
What Is Formative Feedback?
Formative feedback is information given to students about their performance relative to learning goals, with the aim of improving understanding before a final assessment. Done well, the Education Endowment Foundation’s review of 155 studies puts its average impact at six months of additional learning progress.
The word itself is the clue. Formative: still being formed. The feedback is meant to shape learning while there’s still time to act on it.
How It Differs from Summative Assessment
Summative assessment tells you what happened: the end-of-unit test, the final grade, the exam. It’s a record.
Formative feedback is what you do before you get there. A mid-lesson check, a quiz that explains a wrong answer, a question that redirects a student before they move on. The goal is to change what students understand, not measure it.
The distinction matters because both can look similar on the surface. A quiz is a quiz. But a quiz that gives students immediate explanations and lets them move forward is doing something different from a quiz that produces a grade three days later. That gap is where most feedback loses its effect.
What Two University Classrooms Found
You know how we’re constantly trying to get students interested in a topic? Motivated, emotionally invested, genuinely curious? Turns out that doesn’t move the grade needle as much as we think.
The Students Who Improved Were the Ones Who Did the Work
A study at King Fahd University measured three types of student engagement: emotional, cognitive, and behavioral. Emotional engagement is how much students enjoy the class. Cognitive is how much they feel they understand. Behavioral is whether they actually complete the work.
Only behavioral engagement predicted grades. The students who felt engaged and understood the material still didn’t outperform the control group. The ones who did were the ones who answered the questions, completed the exercises, and submitted what was asked.
Full study: Research Journal in Advanced Humanities.
Want the full breakdown? Read: King Fahd University Study on Edcafe AI and Participation
So if doing the work drives outcomes, the next question is: what helps students actually get through it? That’s where the Almuslim study comes in.
The One Design Change That Reduced Math Anxiety
Researchers in Indonesia built a Grade 10 statistics module inside Edcafe AI where students got feedback during each practice activity, not after the unit test. The module was reviewed by six independent experts before any student touched it.
100% of students said it improved their numeracy skills. Math anxiety dropped across the group. The researchers traced both outcomes back to one design decision: feedback arrived while students could still act on it, not after the grade was already set.
Full study: Jurnal Penelitian Pendidikan IPA
See how the feedback was built into the module: How Almuslim University Used Edcafe to Reduce Math Anxiety
The two studies are saying the same thing in different ways. Doing the work predicts grades. Feedback during the work is what makes formative feedback meaningful. That’s the design principle the rest of this post is built on.
Formative Feedback Strategies Most Classrooms Already Use
Most teachers already have a formative feedback toolkit. These are the strategies you’ll find in most classrooms, and for good reason: they’re practical and they work in a variety of settings.
1. Exit Tickets
At the end of a lesson or a key segment, give students two minutes to answer one specific question in writing before they leave. Not “what did you learn today.” Something concrete, like “explain in one sentence why X causes Y” or “show me the first step you’d take to solve this.” Collect them and skim before your next class. You’ll know exactly what to address before you even start planning.
2. Mid-Lesson Checks
Pause after a key explanation and ask every student to respond at the same time, mini whiteboards, thumbs, a quick poll, or a hinge question with forced choices. The key word is every. Asking “any questions?” gives you data from whoever feels brave enough to speak.
A simultaneous check gives you data from the whole room. Some techniques work better at different points in a lesson than others, so it’s worth knowing which one fits the moment.
3. Peer Feedback
After a draft or practice task, give students a structured prompt to use with a partner: “Name one thing that works, one thing to improve, one question you have about their work.” Without that structure, peer feedback defaults to “good job” or uncomfortable silence. The prompt is what makes it useful.
4. Quizzes With Explanations
Build short quizzes where wrong answers come with an explanation, not just a red X. Students see why before they move on. That single design choice is what separates a quiz that teaches from a quiz that just tests.
These strategies work well together, especially in self-paced setups where students need ongoing checkpoints rather than one end-of-unit test.
Why Most Formative Feedback Works for You, Not for Students
The four strategies above are good at giving you information. Exit tickets tell you where to start tomorrow. Mid-lesson checks show you who’s lost. Peer feedback helps students articulate their thinking.
But there’s a second loop most classrooms miss: feedback reaching students while they’re still doing the work. That’s what the Almuslim module was designed around, and it’s what the King Fahd data pointed to. Quizzes with explanations come closest to that model. The rest close the loop for you.
The redesign isn’t about replacing these strategies. It’s about adding the loop that students need.
How to Redesign Your Formative Feedback Around These Findings
Three practical shifts to make your formative feedback work the way the research describes.
1. Put Feedback Inside the Activity, Not After It
Build work that responds to students before they finish. A chatbot that asks a guiding question instead of giving the answer. A task with a check step before the next section unlocks. A structured assignment where students see your comments before the grade is set.
If you’re already using Edcafe AI, the explanation field on every quiz answer is the quickest place to start. For Google Forms or Quizizz, the same option sits in the answer settings.
2. Design Work That Requires Completion
Assignments that can be skimmed or submitted blank don’t create behavioral engagement. Work that requires students to respond, show their thinking, and submit structured answers does.
Edcafe AI’s Assignment Grader structures the submission so there’s no skipping the parts that matter. You get suggested scores and feedback to review before anything reaches the student. Google Classroom assignments with required fields or rubric-aligned tasks in your LMS work on the same principle.
[IMAGE: Edcafe AI Assignment Grader showing student submission with suggested score and feedback fields, caption: “Review suggested scores and feedback before anything reaches the student”]
3. Keep Teacher Oversight in the Loop
In-practice feedback still needs a teacher watching the data. Set up visibility: quiz dashboards that show which question the class is failing, chatbot summaries that flag stuck students, alerts when someone hasn’t engaged.
For how to stay in the loop without it becoming another marking pile, personalizing student feedback with AI covers keeping your judgment central while AI handles the data layer.
Before You Go
You don’t need to overhaul how you teach to apply this. Pick the smallest version: add an explanation to every answer option on your next quiz. That’s it. One change, one activity, one step closer to feedback that reaches students while it still matters.