Four ways of engaging, from listening to working it through together

In one line: people tend to learn more as they move from just listening, to doing something with the material, to making sense of it in their own words, to working it through with someone else.

What they did

This isn’t a single experiment. Michelene Chi and Ruth Wylie, learning scientists in the USA, looked back across many earlier studies. All of them compared different things learners might do with the same material: listen to it, take notes, underline, draw a concept map, explain it to themselves, or discuss it with a partner.

They sorted those activities by what the learner’s head is doing, not by what the activity is called. Then they asked whether the sorting predicted who learned most.

What they found

• Four modes of engaging. Passive: taking it in, such as listening or reading. Active: handling it, such as underlining, copying out steps or pausing a video. Constructive: producing something new, such as explaining in your own words, asking “why?”, or drawing a diagram. Interactive: two or more people building on each other’s ideas.
• A predicted order. Interactive beats constructive, which beats active, which beats passive. They called this the ICAP hypothesis.
• The earlier studies mostly fitted. Across the studies they reviewed, learners who did more with the material generally learned more.
• The label isn’t the mode. The same task can sit on different steps. Taking notes word for word is active; summarising in your own words is constructive. A “discussion” where one person talks and the other nods isn’t interactive.

How much should you trust it?

A useful, widely used map; the broad direction is well supported, the exact order less so.

• Most of the support comes from re-reading earlier studies through this lens, not from new experiments designed to test it.
• Later classroom tests are mixed. A university biology study found interactive classrooms did a little better than constructive ones (Wiggins et al., 2017). Other studies found no extra gain from the interactive step.
• Critics argue you can’t reliably tell someone’s mode from the outside, and that the strict four-step order goes further than the evidence (Thurn et al., 2023).
• The evidence is from schools and universities. It hasn’t been tested in healthcare workplaces.

If you need to convince someone

In our words (a paraphrase, not a quotation): being busy isn’t the same as learning. What counts is how much people do with an idea: explaining it and working it through with others usually beats listening or copying it down.

What it means for you

1. Look at what people’s heads are doing, not what their hands are doing. Ticking through a form is active. Explaining why a patient scores as they do is constructive.
2. One step up is often enough. “Tell me in your own words” turns listening into making sense, and takes thirty seconds.
3. Make pair work a real exchange. Give people something to decide or disagree about together, or it slides back to two people working side by side.

How it appears in Teaching That Lands

In Session 2, participants make ten minutes of their own session active. The question asked while they work, “What will learners be doing in minute 3?”, is a question about where on this scale the activity sits. The 60-second pitches that follow ask each person to explain their design, then hear one thing that works and one question from the room. Session 2’s recap calls this idea “Active is a scale, not a switch”.

The small print

• Earlier versions. Chi set out an earlier form of the framework in 2009 (“Active-constructive-interactive”, Topics in Cognitive Science). The 2014 paper refined it and gave it the name ICAP.
• What the studies measured. Most measured test scores soon after learning, often in short lab or classroom tasks, so we know less about long-term memory or real-world performance.
• Not the learning pyramid. The popular pyramid with retention figures (5% for lectures, 90% for teaching others) has no research behind its numbers (Letrud, 2012). ICAP makes no claims about percentages.
• Related evidence. A large review of university science courses found active learning beat lecturing alone on exam scores and failure rates (Freeman et al., 2014).