GLOBAL PERSPECTIVE · INDEPENDENT THINKING
UNDERSTAND WHAT MOVES THE WORLD.
Human Behaviour · News analysis

When a Classroom Turns an Answer Into a Membership Test

In three experiments, preschoolers had more reason to keep a disproven answer when agreeing was the price of joining a group.

BY GABRIEL NICOLAEV – CYGURU · · 4 MIN READ
READING TOOLS
A child’s hands arranging colourful wooden shapes on a tabletop.
THE READING GUIDE

Inside this article

01What the two puppet teams had in common ↓02A response can protect a place in the group ↓03Give the group a way to stay together after a revision ↓

Event / source date: 9 November 2026 — PNAS publication. Duke reports on the study on 11 November. · Published 12 NOVEMBER 2026

What the two puppet teams had in common

In a set of experiments reported by Duke University on November 11, four- and five-year-olds faced a simple choice about a toy machine. Which block would make it light up? A team of puppets backed one answer. The children then saw that the team’s choice failed twice while the other block worked twice. The researchers, Zoe Renee Finiasz and Tamar Kushnir, changed one important part of the situation. In one condition, agreement about the block was a requirement for membership. In another, being nice was what qualified someone to join.

The teams held the same stated view about the blocks. This matters because a child who accepts a group’s guess may be responding to consensus, while a child who keeps the guess after seeing contrary evidence may also be protecting a relationship. By varying what membership required, the experiments tested one reason why the same evidence could produce different responses. The study covered 192 children across three experiments and was published in PNAS on November 9.

A response can protect a place in the group

In the second experiment, the researchers compared a group defined by shared belief with a setting in which the children and puppets initially made the same guess. The children revised more readily in the comparison condition. That helps narrow the explanation. Having previously agreed with the puppets was a feature that could be held apart from a group rule saying agreement determined who belonged.

The third experiment moved beyond an answer spoken in front of a team. With the puppets absent, children could act on a choice of block themselves. Those who had faced a membership rule tied to belief were more likely to act on the original choice. The result is interesting because behavior carried the effect into a situation where the puppets were away. It still leaves private conviction uncertain. A choice in a brief game cannot tell us exactly what the child thought the rule meant or how long it influenced them.

Give the group a way to stay together after a revision

Imagine a science activity where a table group chooses which material will keep ice from melting fastest. Each table names its prediction, and students are encouraged to present it as a team. Once the results arrive, a student at one table notices that their favorite material produced more meltwater than expected. If the group’s public role is to defend its named prediction, that student has to decide whether to challenge both the guess and the team’s standing at the same time.

A teacher could make a different rule explicit before testing. The group stays intact while it checks predictions and revises its account. Put the first guess and the new observation in separate columns so the class can see the work of changing an answer. Give a student the role of describing what would count as evidence for a revision. Use a second student to show how the group checked the result. Both students retain responsibility for the team’s work even when the original prediction falls away.

Watch the next turn after a correction

Classroom norms are easiest to read in what happens immediately after a student departs from an agreed view. A child says they now think the other material works better. Does the teacher ask them to show the measurement? Do the other children get a role in checking it? Or does the class react as though someone has spoiled the group’s presentation? A rule announced at the start matters less if the first real correction carries a social cost nobody addresses.

An educator trying a new discussion routine can keep the comparison modest. Ask each student to give an initial prediction privately. After the test, invite the group to agree on what happened and then let each student privately describe which option they would use next time and why. Review the gap between the public and private responses without treating it as a diagnosis of an individual. The gap may reflect confusion, deference, ambiguous evidence or a wish to avoid conflict. Follow up by improving how the evidence is shown and making revisions part of the group’s job.

Keep the classroom example at the right scale

The Duke and PNAS accounts describe an experiment with young children, puppet teams and a controlled demonstration. A real classroom involves existing friendships, a teacher’s authority and questions that often have ambiguous answers. The study did not test the classroom activities proposed here. Nor can a child’s block choice establish a mechanism for adult political polarization. Those are separate questions for researchers to study directly.

CYPAPER has explored changing your mind when a conclusion feels bound up with who you are. The study brings one small and manipulable condition into view. A group can put agreement into its membership rule or give its members a shared task that includes correcting each other. In a classroom, that choice becomes tangible when the first student shows the rest of the table something its original answer failed to explain.

Gabriel Nicolaev — CyGuru

Sources & further reading

Duke University research report ↗

PNAS abstract and significance statement ↗

A tabletop learning activity. Illustrative photograph by Magic Fan / Unsplash.

FOLLOW THE CONNECTION

Another angle on this subject

Continue through related ideas, decisions and consequences.