Gerald Ajam

“Engagement is evidence of learning”

Engagement is a condition for learning, not a measure of it.

Status Position

Published September 2026

All counterpositionsLearning

Usually said

“Engagement is evidence of learning”

The shift

From: Engagement as the measure

To: Engagement as a precondition

More precisely

Cognitive engagement with the material is a necessary condition for learning, and learning itself is shown only by what the learner can do later.

In short

Engagement matters. A learner who has checked out will not learn much. But a busy room shows that people are doing something, and says little about what they will be able to do next month. So treat the right kind of engagement as a precondition, and look for learning in later performance.

The argument

From the claim to a better one

  1. The strongest formulation

    Learners who are not attending, participating or persisting cannot learn much. Engagement is a necessary condition, and it is visible in real time when learning is not.

  2. Why it is attractive

    It can be seen and counted today. Learning shows up weeks later, if anyone checks, and a busy, cheerful room feels like success to everyone in it.

  3. Where it holds

    It holds where the activity itself demands the thinking that causes learning: retrieving, explaining, predicting, deciding. There, engagement with the task and engagement with the material are the same thing.

    Where it breaks

    It breaks when the engaging part is the wrapper. Clicks, streaks, time on task and enjoyment all rise for reasons unrelated to understanding, and the most effective strategies often feel worse to learners while they work. This is a disagreement about evidence, not values.

  4. What evidence matters

    Coe (2013) lists busy, engaged students among the poor proxies for learning. Fredricks, Blumenfeld and Paris (2004) separate behavioural, emotional and cognitive engagement, and only the last is close to the mechanism. Roediger and Karpicke (2006) and Deslauriers et al. (2019) show learners rating the less effective condition as the better one. What matters is delayed performance on a task the learner could not do before.

  5. So, more precisely

    Cognitive engagement with the material is a necessary condition for learning, and learning itself is shown only by what the learner can do later.

Reading it

The precise version, word by word

Cognitive engagement1 with the material is a necessary condition2 for learning, and learning itself is shown only by what the learner can do later3.

  1. Cognitive engagement

    Effortful thinking about the material itself: recalling it, explaining it, applying it. It is different from being busy, which is behavioural engagement, or enjoying it, which is emotional engagement.

  2. Necessary condition

    Learning rarely happens without it, and its presence guarantees nothing. Its absence tells you something. Its presence is not proof.

  3. What the learner can do later

    Performance after a delay, on a task they could not do before, without the prompts that were there during the lesson.

What it changes in practice

  • Ask what learners are thinking about during an activity, and design it so the engaging part is the thinking.
  • Report engagement figures as inputs. Report learning with a delayed check.
  • Expect effective practice to feel harder, and tell learners so before they judge it by how it feels.

A case, in general terms

Take two versions of the same revision session. In one, learners reread their notes and highlight, and they rate it highly. In the other, they close their notes, try to recall the key points, then check. The second feels worse and looks less lively. A week later, the second group remembers more. Judged by engagement, the first session wins. Judged by learning, the second does.

Singapore