Children’s Gender Bias Emerges at Age Three, Study of Fairness Judgments Finds

A hidden distributor hands out apples to two animal puppets in a simple animated scene, and within seconds a toddler is asked a deceptively difficult question: who did this? When the distributor split the treats evenly, most three-and-a-half-year-olds reached for the drawing of a man. When the distributor kept both apples for one puppet, five-year-olds reached for that same male figure. A new study published in the International Journal of Early Childhood shows that these apparently contradictory choices are two stages of a single developmental story, one in which gender schemas, in-group loyalty, and emerging stereotypes gradually reshape how young children interpret the fairest and most unfair of everyday actions.

The research, conducted by Valentina Lucia La Rosa, Maria Luisa Indiana, Emanuela Borzì, Daniela Pirone, Elisabetta Sagone, Elena Commodari, and Alessandra Geraci at the University of Catania, builds on a well-established tradition in developmental psychology but adds a crucial cognitive twist. Earlier work had shown that children attribute fair and unfair distributions differently depending on the perceived gender of the distributor, yet comparatively little attention had been paid to the mental structures, known as schemas, that make such attributions possible. Gender schemas are organized knowledge structures that guide how children process, interpret, and remember gender-relevant information, and understanding when children first deploy them to encode social behavior is central to tracing the developmental origins of bias.

The team tested 160 typically developing children divided into four equal age groups: toddlers aged 2.5 years, preschoolers aged 3.8 years, kindergartners aged 5, and school-age children aged 8. Each child watched brief familiarization events in which a distributor, hidden behind a gray box, either gave one apple to each of two receivers, a cow and a pig, or handed both apples to a single receiver. Immediately afterward, the experimenter presented two schematic drawings, one male and one female, and asked the child to identify the distributor by picking up the correct picture. The position of the pictures was counterbalanced, the order of fair and unfair events was alternated across participants, and a manipulation check with 50 university students confirmed that the schematic figures were unambiguously recognized as male and female with 100 percent accuracy.

The results traced a non-linear developmental trajectory that surprised no one more than the researchers’ own carefully calibrated hypotheses. At 2.5 years, children chose the male and female figures at chance levels for both fair and unfair distributions, exactly as predicted by the literature on pictorial competence. Infants younger than three often treat pictures as objects to be touched and grasped rather than symbols that stand for something else, and the ability to hold both the physical properties of an image and its abstract referential meaning in mind simultaneously, what psychologists call dual representation, emerges only gradually. Without that symbolic foundation, the youngest participants simply could not map a schematic figure onto a real-world gender category, and their choices therefore revealed no gender-based pattern at all.

By 3.8 years, however, the picture changed dramatically. Seventy percent of children in this age group selected the male figure after watching a fair distribution, a statistically significant deviation from chance with a medium effect size, while their choices remained at chance after unfair events. This is the signature of what the researchers define as gender bias: a systematic tendency to attribute positive social behaviors preferentially to one’s own gender group, reflecting in-group favoritism without necessarily implying a negative evaluation of the out-group. The pattern was driven almost entirely by boys, 81.8 percent of whom attributed fair distributions to the male figure, while girls of the same age showed no significant tendency in either direction. According to Social Identity Development Theory, this asymmetry makes sense: in-group favoritism emerges as children consolidate a positive sense of group membership, and its strength depends on how salient and self-defining that group has become. For boys approaching their fourth birthday, gender may already function as a highly salient social category.

At age 5, the pattern flipped in a theoretically meaningful way. Children no longer associated the male figure with fair behavior; instead, 67.5 percent of them attributed unfair distributions to the male figure, again a significant departure from chance. This qualitative shift coincides with what decades of research identify as the developmental peak of gender stereotype rigidity, spanning roughly ages 5 to 7. During this window, children behave as what Martin and Ruble famously called gender detectives, diligently searching their environment for stereotype-consistent information and consolidating it into increasingly rigid expectations. Developmental Intergroup Theory predicts precisely this outcome: as gender becomes an ever more salient dimension of social life, children encode stereotype-consistent information with growing intensity, and here that intensity manifested as the association of men with dominant, inequitable behavior.

The most articulated pattern emerged among 8-year-old girls, who simultaneously attributed fair distributions to the female figure, at 71.4 percent, and unfair distributions to the male figure, at 78.6 percent. This combination of in-group favoritism and out-group negativity meets the study’s operational definition of gender prejudice, a qualitatively different construct from mere bias. Male 8-year-olds, by contrast, showed no significant gender-based pattern in either condition. The researchers suggest that girls at this age may have developed heightened sensitivity to gender-based social hierarchies, motivating a reactive intergroup stance in which protecting the in-group and devaluing the out-group become two sides of the same coin. Taken together, the findings sketch a trajectory in which boys display earlier but less differentiated in-group favoritism, while girls develop a later but more elaborated pattern that integrates group protection with out-group criticism.

The authors are careful to frame these results within a biosocial model rather than a simple nature-versus-nurture debate. An evolved sensitivity to group membership may provide the initial scaffold, but the direction of children’s attributions is unmistakably shaped by experience. From their earliest years, children observe that caregiving and daily resource allocation are unequally distributed across genders, and these accumulated observations likely feed the early association between the male figure and prosocial or dominant behavior. Crucially, the researchers emphasize that the attribution patterns observed here are not fixed developmental outcomes. Because they emerge from socialization, they remain modifiable: intervention research consistently shows that exposure to counter-stereotypical role models, such as female leaders and male caregivers in children’s stories, can attenuate gender bias even at preschool age.

The study does carry important boundary conditions. The sample consisted exclusively of White children from a single Italian region with limited socioeconomic diversity, so the observed patterns cannot be assumed to generalize to populations where gender norms and expectations about distributive fairness differ substantially. The use of highly schematic figures, while useful for isolating gender as the sole variable, reduces ecological validity, since real-world gender perception draws on a rich constellation of facial, vocal, and behavioral cues. Moreover, the forced-choice paradigm measures attribution behavior without directly observing the underlying schemas, and the authors call for future work pairing the task with independent assessments of gender schema knowledge, culturally diverse samples, longitudinal designs, and richer stimuli such as videos of dynamic social interactions.

Even with these caveats, the practical implications are difficult to overstate. The window in which gender schemas first begin to shape social reasoning opens before children turn four, long before formal schooling typically addresses equity in any systematic way. The findings argue for early childhood settings, nurseries, kindergartens, and caregiving environments, to actively promote counter-stereotypical representations of competence and fairness, particularly around resource sharing and leadership, rather than waiting for rigid prejudice to consolidate in middle childhood. Gender bias, the authors conclude, is not an inevitable consequence of development; it is a learned pattern, and what is learned early can be unlearned before it hardens. In a world where gender inequality continues to fuel conflict and injustice, the humble apples of a nursery-school experiment may mark one of the earliest and most promising points of intervention.

Subject of Research: The developmental emergence of gender schemas and gender bias in children's attribution of fair and unfair distributive actions.

Article Title: The Perception of Schemas and Gender Bias in the Attribution of Distributive Actions

Article References: La Rosa, V. L., Indiana, M. L., Borzì, E., Pirone, D., Sagone, E., Commodari, E., & Geraci, A. (2026). The Perception of Schemas and Gender Bias in the Attribution of Distributive Actions. International Journal of Early Childhood. https://doi.org/10.1007/s13158-026-00539-4

Image Credits: AI Generated

DOI: 10.1007/s13158-026-00539-4

Keywords: gender bias, gender schemas, distributive justice, child development, fairness, social cognition, gender stereotypes, developmental psychology, in-group favoritism, early childhood education, prejudice, resource distribution