- Title: Why Peer Discussion Improves Student Performance on In-Class Concept Questions
- Authors: M. K. Smith et al.
- Access the original paper here
- Watch a video overview:
Paper summary
When students answer a conceptual question with a clicker, discuss it with neighbours, and revote, more get it right. But is that because they have understood the idea, or because they are copying confident classmates? In an undergraduate genetics course at the University of Colorado–Boulder, researchers tested this with 350 students. After discussing a question, students individually answered a second, differently worded question testing the same concept, with no answers revealed in between. Performance on this fresh question rose sharply. Crucially, even in groups where nobody initially knew the correct answer, students often reached the right answer on the new question, well above chance. This suggests peer discussion builds real understanding, not just social copying, even without an expert in the group.
If teachers remember one thing from this study, it should be…
Getting students to discuss a problem in small groups can build genuine understanding, even when no one in the group starts out knowing the answer. The value of discussion isn’t only that a knowledgeable student explains it; the act of reasoning together helps students work concepts out for themselves.
Paper Deep Dive
What are the key technical terms used in the paper?
- Clicker questions: multiple-choice questions students answer in class using handheld response devices.
- Isomorphic questions: two questions with different surface stories that test the same underlying concept.
- Naïve group: a discussion group in which no member initially knew the correct answer.
What are the characteristics of the participants in the study?
The study was in an undergraduate introductory genetics course for biology majors at the University of Colorado–Boulder. Analysis covered 350 students who answered all three questions across 16 isomorphic pairs assessed through the semester. Discussion groups averaged about three students.
What does this paper add to the current field of research?
Peer discussion was known to raise clicker scores, but many assumed this worked because informed students steered classmates to the right answer. This study is notable for showing that groups where nobody knew the answer still improved on a fresh question: evidence that discussion itself builds understanding, not just copying.
What are the key implications for teachers in the classroom?
- Let groups struggle before you step in. The headline finding is that students learn from discussion even when no one in the group knows the answer. So resist the urge to seed each group with a strong student or to jump in with the solution. Pose a real problem, let mixed-ability groups talk it through, and trust that the reasoning process itself does the work.
- Ask a second, differently worded question to check for real learning. A revote on the same question can rise simply because students copy a confident neighbour. To see whether understanding has actually transferred, follow discussion with a fresh question that tests the same concept in a new context, answered individually. If they get that right, the learning is real.
- Delay revealing the answer. In this study, correct answers and class response histograms were hidden until after the second question. That kept students reasoning rather than reading the room. Hold back the “right” answer until students have committed, so discussion stays about the concept, not about spotting the popular choice.
- Frame discussion as reasoning, not answer-swapping. The benefit came from students justifying their thinking and questioning each other, not from passing along a right answer. Prompt groups with “explain why” and “convince each other,” and treat disagreement as useful. The talk is the learning, so make talking-it-through the explicit goal.
Why might teachers exercise caution before applying these findings in their classroom?
This was one genetics course at a single university, with no control group who skipped discussion, so some improvement may reflect simply seeing the concept twice. The fresh question and repeated exposure may both help. Findings may differ by subject, age, and how discussion is run.
What is a single quote that summarises the key findings from the paper?
“Our results indicate that peer discussion enhances understanding, even when none of the students in a discussion group originally knows the correct answer.”
