10 Ways to Increase Student Engagement in Lectures

How is this list ranked?

Each method carries an evidence grade:

• Strong, supported by meta-analysis or multi-year replicated study
• Reasonable, supported by specific studies, or a well-evidenced mechanism
• Weak, widely recommended on thin or contested support

And an ICAP mode, from Chi and Wylie's framework ranking cognitive engagement from Passive through Active and Constructive to Interactive.

Which techniques have strong evidence?

1. Peer instruction (Interactive, strong evidence). Pose a conceptual multiple-choice question, have everyone answer, have them discuss with a neighbour who disagreed, answer again, then reveal.

Crouch and Mazur (2001) reported ten years of Harvard introductory physics measured with the Force Concept Inventory: normalised learning gains roughly twice those of traditional lecture, reaching 0.74 by the sixth implementation.

Caveat: one institution, one subject, one instructor refining over a decade. Vickrey et al. (2015) found outcomes depend on implementation fidelity, so the re-vote is not optional.

2. A question at each transition (Active to Constructive, strong evidence). Szpunar, Khan and Schacter (2013, PNAS) inserted brief tests through a 21-minute lecture. Tested students were half as likely to report mind-wandering and took three times as many notes.

The active ingredient is the expectation of being asked, which changes how students attend throughout. Caveat: video lecture, laboratory setting.

3. Restructure a segment around student production (Constructive, strong evidence). Freeman et al. (2014), across 225 studies, found traditional lecturing produced 1.5× the failure rate of active learning, with exam scores about 6% higher.

This is the strongest evidence in the field, and it is about course design rather than any single trick. One segment per lecture where students produce something is the smallest version of it.

Which techniques have reasonable evidence?

4. Wait five seconds (Active, reasonable evidence). Most lecturers wait about a second after asking a question, which selects for students who already knew. Five seconds produces answers.

No meta-analysis, a very well-established mechanism, and it costs nothing, which is why it belongs near the top of any practical list.

5. Think-pair-share (Interactive, reasonable evidence). 30 seconds thinking, 60 seconds in pairs, then answers. It reaches the top ICAP mode with no technology, because the student reports the pair's conclusion, a much cheaper social act than speaking as themselves.

Fails when you skip the silent think.

6. Anonymous comprehension checks (Active, reasonable evidence). A show of hands measures confidence; a response from every seat measures comprehension. Over half of students never ask or answer a question all semester (Cooper et al., 2021), and the interview data identifies fear of peer judgement as the driver, so removing identification removes the barrier the research actually names.

7. Predict, then reveal (Constructive, reasonable evidence). Ask for a prediction before showing a result. Commitment changes how the answer lands. Cheap, and it converts a passive reveal into a constructive act.

Which popular advice is weaker than you think?

These are on every list. Their evidence is thinner than their popularity.

8. "Break every 10–15 minutes, attention collapses" (weak evidence). The 10–15 minute attention span is not supported by its own sources. Bradbury (2016) traced the claim and found the primary data does not establish it; the most-cited source "barely discusses student attention at all". The largest source of variation was differences between teachers, not format.

What to do instead: change the demand every 10 to 20 minutes rather than inserting a pause. The evidence supports changing what students must produce, not stopping the clock.

9. "Add videos and multimedia" (weak evidence). Under ICAP, watching a video is Passive, the same mode as listening to you. Swapping one passive input for another does not change cognitive engagement.

What to do instead: give students something to produce first, a prediction or a question to answer while watching, which moves the same video to Constructive.

10. "Use gamification to motivate" (weak evidence). The figures circulating for gamification and retention ("75% higher retention" and relatives) trace to content marketing, not research. No traceable study supports them.

There is a further problem specific to lectures: a public leaderboard rewards students who were already confident, which is precisely the group that was never the problem. The participation research points the other way, towards lowering the social cost of answering rather than raising the stakes.

What to do instead: for diagnosis, use anonymous checks. They reach the part of the room a scoreboard never does.

How do the ten compare?

What the evidence does not show

• That engagement equals learning. Hunsu et al. (2016), 53 studies and more than 26,000 participants, found response systems have a small effect on cognitive outcomes and a near-medium effect on engagement. The proxy moves more than the outcome.
• That gains persist. The 2025 systematic review of audience response systems in higher education notes that no included study used a delayed post-test.
• That these transfer beyond STEM. The strongest data is in undergraduate science. Plausible elsewhere, not demonstrated.

Frequently asked questions

What is the most effective way to increase student engagement?

Peer instruction has the strongest evidence of any single technique: pose a conceptual question, have everyone answer, let students discuss with a neighbour who answered differently, then answer again before revealing. Crouch and Mazur (2001) reported normalised learning gains roughly twice traditional lecture over ten years of introductory physics.

Do I have to redesign my course to increase engagement?

No. The highest-return changes are small and work with the lecture you already give: waiting five seconds instead of one after asking a question, replacing "any questions?" with a specific question, and putting a short question at each transition between topics. None requires preparation, technology or permission.

Does breaking a lecture every 10 to 15 minutes work?

Not for the reason usually given. The attention span claim behind it does not survive checking: Bradbury (2016) traced the citation chain and found the primary data does not support a 10 to 15 minute limit. Changing what the room has to produce every 10 to 20 minutes is the version of the advice the evidence supports.

Does gamification increase engagement?

The evidence is weaker than the marketing. Competitive scoring reliably raises the energy of a room, but a public leaderboard raises the cost of being visibly wrong for the students least likely to participate already, which is the opposite of what most lecturers are trying to achieve in a large hall.

Will more engagement raise exam scores?

Not directly, and this is where the category oversells itself. Hunsu, Adesope and Bayly (2016), across 53 studies and more than 26,000 participants, found a small effect on cognitive outcomes and a near-medium effect on engagement, attention and participation. Freeman's larger effect came from courses restructured around active learning, not from techniques added to an unchanged lecture.