Children back group claims over evidence, but privacy reduces bias, experiments reveal
As we move closer to Election Day 2026, voting preferences are moving back into focus—and with them, analyses of what drives partisanship at the polls. However, less frequently asked is when Americans ...
As we move closer to Election Day 2026, voting preferences are moving back into focus—and with them, analyses of what drives partisanship at the polls. However, less frequently asked is when Americans show evidence of partisan behavior: shortly or well after reaching the legal voting age? As teenagers? In elementary school?
A team of psychology researchers has found evidence of partisan behavior in children ages 5 to 9—they frequently endorsed their own group's claims even when evidence suggested otherwise, indicating group affiliation influenced their responses. However, the scientists also uncovered a potential remedy to such responses: When incentivized to tell the truth about what they had seen or when they could provide answers under the veil of privacy, the children were much less likely to adopt their own group's claims. The paper is published in the journal Cognition.
"Even young children will side with their group over the evidence of their own eyes, but mainly when they're responding publicly and when being accurate doesn't count for much," explains Andrei Cimpian, a psychology professor at New York University and the senior author of the paper. "However, if you allow them to respond in private or give them a reason to care about accuracy, the partisanship effect disappears."
The researchers, who included first author Bethany Lassetter, an NYU postdoctoral fellow at the time of the study; Natalie Hutchins, a doctoral student at the University of Virginia; and Lucas Butler, an associate professor at the University of Maryland, College Park, note the potential significance of the latter finding in offering insight into the developmental origins of political partisanship.
"Partisanship may start not as a conviction about what's true, but as a way of showing you belong or you're loyal to your group," observes Lassetter. "But there's an encouraging implication here, too: Conditions that reward accuracy or that lower the social stakes of an answer can pull people back toward the evidence."
Testing loyalty against evidence
The study included different groups of children, ages 5 to 9, in three experiments.
In the first, children were introduced to two groups, the green group and the orange group, differentiated only by the color of their T-shirts. The children then chose whether they wanted to be in the green group or the orange group, thereby creating an ingroup based on a shared T-shirt color. A series of manipulation checks confirmed that they viewed their group more positively than they did the other group. Children in a control condition were not introduced to groups—however, all participants saw the same experimental stimuli.
Next, children saw several pairs of fictional characters, one from their group and one from the other group, who made conflicting claims about objects with ambiguous identities (for example, an animal that looked equally like a horse and a cow). Characters in the control condition wore green and orange T-shirts, but no groups were ever mentioned. Children were asked what they thought the pictured objects were (a horse or a cow?).
Overall, compared with children in the control condition, children in a group (green or orange) were significantly more likely to endorse the claims made by their group's character about what the pictures showed—regardless of what the images actually depicted.
Privacy and prizes changed answers
The results raised a key question: Were the children siding with their group because they thought the group was accurate, or—instead—because they wanted to be loyal and go along with what their group said?
To examine this, the authors asked another group of children, ages 6 to 9, to undertake the same experimental task. However, unlike in the first experiment, the children were told their answers would be communicated privately.
In a third experiment, with another sample of children ages 6 to 9, children were put into two groups: One group received the experimental treatment identical to the first experiment, while another group was told that the more answers they got correct, the bigger the prize they would receive at the end. This was a way of testing whether incentives to tell the truth would outweigh group identification.
The impact of privacy and "truth incentives" was clear: Children who answered privately were more likely to accurately report what they saw than those who answered publicly. Similarly, those in the truth-incentive group were more likely to accurately report what they saw than those who received no such incentive.
Belonging, not belief, drove bias
Taken together, the experiments indicated that children's partisanship appears to be less about a search for truth and more about a desire for social connection, the authors conclude—and point to potential remedies for diminishing responses not supported by evidence.
"Understanding these early tendencies, and the conditions under which they intensify, diminish, or solidify into genuine belief, may ultimately help explain how group-based distortions in belief take hold and how they might be mitigated," the authors write.
Publication details
Bethany Lassetter et al, Investigating the origins of partisanship: What motivates children to preferentially endorse their ingroups' claims?, Cognition (2026). DOI: 10.1016/j.cognition.2026.106629
Who's behind this story?
Master's in TESOL from The New School. Passionate about language learning and editing science news on biology and space exploration. Full profile →
Master's in physics with research experience. Long-time science news enthusiast. Plays key role in Science X's editorial success. Full profile →
Citation: Children back group claims over evidence, but privacy reduces bias, experiments reveal (2026, July 10) retrieved 11 July 2026 from https://phys.org/news/2026-07-children-group-evidence-privacy-bias.html
This document is subject to copyright. Apart from any fair dealing for the purpose of private study or research, no part may be reproduced without the written permission. The content is provided for information purposes only.