👋 Hi {{first name | reader!}}

In this edition of ⚗️DistillED, we explore a term coined by Dylan Wiliam: feedback as detective work, the shift from simply giving students the answers to giving them a puzzle worth solving.

In this edition

📊 Do Now: Quick Poll!

🗳️ Honestly — how much time do your students spend acting on your feedback vs. how much time you spend writing it?

Login or Subscribe to participate

What is Feedback as Detective Work?

Feedback as detective work is a strategy in which the teacher withholds the location or solution to an error and instead gives the student a solvable challenge: find it, fix it, figure it out. Dylan Wiliam, who coined the framing, puts it directly in Embedded Formative Assessment (2011):

"If I had to reduce all of the research on feedback into one simple overarching idea, at least for academic subjects in school, it would be this: feedback should cause thinking."

Dylan Wiliam - Embedded Formative Assessment (2011)

The detective frame is a mechanism for making that happen. Instead of telling students what is wrong, you tell them that something is wrong, approximately where to look, and how many errors to find. Some examples of feedback as detective work include:

  • Maths: "Five of these equations are incorrect. Find them and fix them."

  • English Language: "Paragraph 3 — your argument about plastic waste — has three lines that could be stronger. Find them. Rewrite each using a rhetorical device."

  • French: "There are five places where you've used the incorrect verb ending. I've highlighted two. Find the other three."

  • Science: "Two of your conclusions don't follow from your results. Find them and rewrite them so they do."

  • History: "There are three claims in this essay that need stronger evidence. Find them and develop each one."

  • Geography: "Four of your case study facts are either missing units or used incorrectly. Find them and correct them."

  • English Literature: "There are two quotations in this response that aren't embedded or analysed. Find them and redraft those sentences."

Traditional feedback and feedback as detective work applied to the same Year 10 persuasive writing task. The intervention occurs at Step 2, what the teacher chooses to withhold determines whether the cognitive work stays with the teacher or transfers to the student.

In typical ⚗️DistillED fashion, here's what Feedback as Detective Work is — and what it isn't:

👎 Feedback as Detective Work IS NOT:

  • Vague → Saying "some of these are wrong" without any cues leaves students flailing, not thinking

  • Cryptic for its own Sake → The point is productive struggle, not confusion

  • A Replacement for all Feedback → It's one mode in a broader feedback repertoire

  • Only for Maths → The structure works across every subject and every age group

👍 Feedback as Detective Work IS:

  • A Challenge With a Solvable Answer → Students are given enough information to actually find the error

  • More Work for the Student Than the Teacher → That's intentional — and that's the point

  • Subject-Agnostic → Works in writing, languages, science, maths, humanities

  • Designed to Cause Thinking → The response is cognitive, not emotional

Why Does Actionable Feedback Matter?

The problem with most feedback is what happens once students get it back. Ruth Butler's 1988 study makes the problem concrete. Students received one of three feedback types after completing a task: comments only, grades only, or both. Those who received comments maintained high interest and kept improving. Those who received grades alone didn't, and those who received both grades and comments were almost identical to the grades-only group, not the comments group.

The recall data explains why this is the case. When asked what feedback they remembered, 89% of students who received both a grade and a comment recalled the grade. Only 45% could recall even part of the comment despite being specifically asked to remember both. In other words, the grade replaced the comment. Once students see a score, high scorers decide they don't need the comments and low scorers decide they don't want them.

Kluger and DeNisi's 1996 meta-analysis is the one that should change how every teacher thinks about feedback. They reviewed over 3,000 studies, applied strict quality criteria, and ended up with 131 that were rigorous enough to include. Across those studies, they found that in more than a third of cases, feedback actually lowered performance. The explanation isn't that the feedback was poorly designed. The problem lies in where feedback directs attention: when it draws attention toward the self rather than the task — as normative, ego-involving feedback tends to do — performance suffers. Instead of the recipient processing the information, they're managing a personal threat and asking themsevles “Am I good at this or not?”

Feedback as detective work removes that moment entirely, which means there's nothing to react to emotionally and nowhere to go but the work itself. Feedback as detective work is designed around that question. It structures the feedback so that the only available response is action not an emotional response. Dylan Wiliam is blunt about the cost of ineffective feedback:

"I once estimated that in England we spend about two and a half billion pounds a year on feedback and it has almost no effect on student achievement."

Dylan Wiliam as cited in What Does This Look Like in the Classroom? (Carl Hendrick & Robin Macpherson, 2017)

That's not a case against feedback. It's a case against feedback that produces no cognitive effort. So, how do we build action into feedback process?

Let’s explore.

How Do Implement Feedback as Detective Work?

Feedback as detective work requires a shift in what you withhold rather than what you provide. One of the most elegant adaptations comes from Charlotte Kerrigan, a language arts teacher whose approach Dylan Wiliam describes in Embedded Formative Assessment. Rather than writing comments in student exercise books, she wrote them on strips of paper. Each group of four students received their four essays plus four strips of feedback and had to match each comment to the correct essay. The matching task is the detective work. Students have to read their own writing closely enough to recognise which comment fits.

Here are two further examples from Wiliam’s Embedded Formative Assessment:

The maths version is the cleanest proof of principle. Ticking fifteen equations and crossing five is a grade in disguise. Telling students “five of these are wrong — you find them, you fix them" turns the same information into a task. As Wiliam notes: "it is often easier to check whether a solution is correct (for example, by substituting for a solved variable back into the original equation)." The detective structure works here because checking is accessible. Students can verify their own answers.

For writing, the dot-in-the-margin technique handles the same problem without full written feedback. A dot per line that needs attention. For students who need more structure, a coded dot: g for grammar, p for punctuation, s for spelling. They know something needs fixing. They have to find what.

Here's how to build the detective structure into your existing practice:

Step

Explanation

Example

1. Mark, But Don't Annotate


→ Purpose: Record what needs fixing without doing the student's thinking for them.

Mark the work as you normally would but instead of writing explanatory comments next to each error, simply note the total count at the top. The marking process stays the same; what changes is what you put on the page.

A History teacher marks a Year 10 essay on the causes of World War One. She identifies three claims that lack supporting evidence. She writes "3 claims need stronger evidence" at the top nothing else.

2. Give the Number, Not the Location


→ Purpose: Provide enough information to make the task solvable, but not so much that the thinking is done for the student.

Tell students how many errors or weak points there are without indicating where they are. This is the core of the detective structure the student has to do the searching.

"There are three claims in this essay that need stronger evidence. Find them and develop each one."

3. Add a Cue if Needed


→ Purpose: Lower the barrier for students who need support without removing the cognitive challenge entirely.

Narrow the search without giving the answer. Point to a section, a paragraph, or a type of error, not the specific line. The student still has to find and fix it.

"Look carefully at your second and third paragraphs. Two of the three are in there."

4. Build in Class Time to Act


→ Purpose: Feedback that isn't acted on in class rarely gets acted on at all.

The detective work happens in class not as homework. Start the next lesson with dedicated time to find and fix before anything else happens.

The teacher returns essays at the start of the next lesson. Students spend the first ten minutes working silently before any discussion takes place.

5. Close the Loop


→ Purpose: Confirm that students understood the problem rather than guessing their way to a correction.

Once students have found and fixed the errors, ask them to explain the change they made and why. A brief pair discussion or exit ticket serves the same purpose.

"You've rewritten that sentence — tell me what was wrong with the original and what you changed." Or: an exit ticket asking students to name one thing they found and explain how they fixed it.

Wiliam’s most famous principle underpins all of it: “feedback should be more work for the recipient than the donor.” (Wiliam, 2011). This strategy is about redirecting the cognitive load toward the student not the teacher. Have a go!

— Jamie

If you want more:

👉 Read The Feedback Gap — Poor Proxy #3: why visible feedback activity is one of the most common false signals of learning, and what's actually happening instead

👉 Read Circulation & Actionable Feedback — how to use live circulation to give feedback in the moment, before the work is even collected

👉 Read Check for Student Understanding — the techniques that tell you whether students are ready to move on, so feedback is targeted rather than blanket

⚗️DistillED Products:

👉 Get the One-Page Guides Bundle — printable A3 guides for evidence-informed CPD, including feedback and formative assessment

👉 Explore the DistillED Playbooks — practical implementation guides for evidence-informed teaching PD across your school or department

📥 Free Planning Tool

Feedback as Detective Work
To help you put these ideas into practice straight away, download the Feedback as Detective Work Planning Tool — a simple one-page template for designing detective feedback tasks across any subject. Use it to plan your challenge prompt, decide on your cues, and structure the class time students need to act on it.Print it, keep it next to your marking, or share it with your department as a starting point for changing how feedback works in your school.

⚗️DistillED+ Exclusive Content

This week, ⚗️DistillED+ members get access to a Feedback as Detective Work CPD PowerPoint, a Printable Detective Feedback Checklist, and the Feedback as Detective Work Planning Tool, breaking down how to design detective feedback tasks across subjects, construct specific challenge prompts, and build class time structures that ensure students actually act on the feedback they receive — so marking produces thinking, not just annotation.

🎟️ Exit Ticket

🎁 Referral Programme

Reply

Avatar

or to participate

Recommended for you