Do clickers actually improve learning?
A little, on average. The largest meta-analysis of clicker studies, by Hunsu, Adesope and Bayly (2016), pooled 111 effect sizes from 53 studies with more than 26,000 participants. A 2025 systematic review in Humanities and Social Sciences Communications quotes their conclusion directly: "only a tiny effect of the clicker was observed on cognitive learning outcomes." The same review points out that only 20% of the studies in that meta-analysis used a pre-test and only 10% randomized their students.
The 2025 review itself screened 653 records down to 11 studies and concluded that "there is no consensus in the literature regarding how ARS contribute to academic performance."
Two well-controlled studies show the range:
- A positive result. Mayer and colleagues (2009) taught a large college educational psychology class three ways. Students who answered 2 to 4 clicker questions per lecture scored higher on course exams than an identical class that got the same questions without clickers (d = 0.38), and higher than a class with no in-class questions (d = 0.40). That was worth about a third of a grade point.
- A null result. In a randomized trial with 127 fifth-year medical students at the University of Adelaide, Duggan and colleagues (2007) found no difference in multiple-choice scores between voting-system lectures and traditional lectures (p = 0.785). Both lecturers also found the voting lectures harder to prepare.
An early review in CBE Life Sciences Education, by Caldwell (2007), put it this way: clickers "typically have either a benign or positive effect on student performance on exams, depending on the method and extent of their use."
What do clickers reliably change, if not test scores?
Participation and honesty. This is where the evidence is strongest.
Stowell and Nelson (2007) gave 140 introductory psychology students the same 30-minute lecture with different ways of answering questions:
| Method | Students who answered | Correct on in-class review questions | Correct on the quiz afterward |
|---|---|---|---|
| Hand-raising | 76% | 98% | 60% |
| Response cards | 97% | 92% | 52% |
| Clickers | 100% | 82% | 60% |
Look at the middle column. With hands up, the room looked almost perfect: 98% correct. With clickers, 82%. Yet both groups scored the same 60% on the quiz afterward. The hand-raising class was copying the confident students. The clicker answers were closer to what students actually knew, which is the information a teacher needs. The authors call this a more "honest (accurate) reflection of their actual learning." The quiz scores did not differ, which fits the meta-analysis above.
In practice: if your checks for understanding come back near-perfect, be suspicious. Anonymous, simultaneous answers give you a truer picture. Our guide to check for understanding techniques has more ways to get that picture without any hardware at all.
Why does peer discussion make such a difference?
Because students learn from each other, even when nobody in the group knew the answer to begin with.
Eric Mazur's Peer Instruction is the best-known version. Students vote on a concept question, discuss it with a neighbor, then vote again. Crouch and Mazur (2001) reported ten years of data from introductory physics and found increased mastery of both conceptual reasoning and quantitative problem solving.
The obvious objection is that the second vote just copies the smartest student at the table. Smith and colleagues (2009) tested exactly that in an undergraduate genetics course with 350 students. After the discussion and revote, students answered a second, similar question on the same concept, on their own. Across 16 question pairs:
- Scores on the new question were on average 21 percentage points higher than on the first individual vote, and 38 points higher on the difficult questions.
- Of the students who got the first question wrong and the revote right, 77% then got the new question right alone.
- The paper concludes that peer discussion improves understanding "even when none of the students in a discussion group originally knows the correct answer."
The wider evidence on active learning points the same way. A meta-analysis of 225 studies in undergraduate STEM by Freeman and colleagues (2014) found that exam scores improved by about 6% under active learning, and that students in traditional lectures were 1.5 times more likely to fail. Clickers are one of the simplest ways to run active learning in a big room, because they show you, in seconds, whether the discussion is needed.
Does answering questions in class help students remember?
Yes, and this is a reason to ask questions even when you already know the class understands. Roediger and Karpicke (2006) had students either restudy prose passages or take recall tests on them. After five minutes, restudying looked better. After two days and after a week, the students who had been tested remembered substantially more. Restudying also made students more confident, which is a trap: they felt they knew it better and did not.
That study used free recall tests, so read it as the reason the habit works. It did not test clickers. The practical point: a short question at the start of the next lesson, on last week's material, counts as learning time.
Is the UK research any different from the US research?
Not in its conclusions, though the vocabulary differs. In the UK, many universities call clickers "electronic voting systems" or EVS, and in schools you may hear "voting handsets" or "keypads." In the US, much of the work cited above comes from physics, biology and psychology courses. In the UK, the University of Glasgow's Steve Draper and Jim Boyle were early names, as a 2011 paper by Carol Withey at the University of Greenwich notes. Withey compared physical clickers and students' own phones with criminal law students, and argued that EVS increases engagement and supports learning.
What are the known downsides?
Kay and LeSage (2009) reviewed 67 peer-reviewed papers and listed the biggest challenges for teachers:
- Time needed to learn and set up the technology.
- Writing effective questions.
- Covering the material, because questions take lecture time.
- Responding on the spot when the chart shows the class has not understood.
They also noted student challenges: adjusting to a new way of learning, confusion when several views are discussed, and dislike of being monitored. In the Adelaide trial, both lecturers found the voting lectures harder to prepare. Caldwell's review cites surveys in which between 20% and 58% of students said they had seen a classmate cheat by using several clickers.
Setup time depends on the system. Some need nothing more than plugging a receiver into the laptop and adding question slides in PowerPoint, which is how OMBEA Response works. Others need student accounts, apps or a good Wi-Fi signal, which our guide to clickers or phones compares. The higher education and schools pages show how the clickers are used with university students and school pupils.
How should you use clickers, based on the evidence?
A plan for your next class, whether it is a 50-minute lesson in a high school (secondary school in the UK) or a lecture for 300 in a lecture hall (or lecture theatre):
- Ask 2 to 5 questions per 50 minutes. Caldwell's review found that this is the typical range, and Mayer's positive result used 2 to 4 per lecture. More is not better if it pushes out the discussion.
- Write concept questions. Use common misconceptions as the wrong answers. If nobody picks a wrong option, it was not a real distractor.
- Vote, discuss, vote again. Do not show the result of the first vote. Ask students to convince a neighbor, then revote. Caldwell suggests groups of no more than four to six.
- Decide what to do before you see the chart. Plan three moves in advance: most of the class wrong, teach it again; a split vote, run peer discussion; most of the class right, a brief explanation and move on.
- Go anonymous for honest checks. The Stowell and Nelson table shows what public answers do. Use named answers only when you need a record, such as an exit ticket.
- Grade lightly, if at all. Caldwell suggests giving partial credit for any answer and full credit for correct answers, so that students do not panic or cheat.
- Open the next lesson with last lesson's question. That is retrieval practice, and it costs you two minutes.
For more ideas for big rooms, see our guide to student engagement strategies for lecture halls.
