The study, led by Dan Zhou with colleagues Mengxiao Yu, Jiayi Chen and Yujun Hu, recruited 219 typically developing children from five inclusive kindergartens in Zhejiang Province. The researchers framed their work in ecological systems theory, the idea that a child’s development unfolds within nested layers of influence, from immediate classroom interactions to broader cultural attitudes. Peer acceptance, they argue, is not a fixed trait of individual children but a product of this whole system, which means that understanding how preschoolers respond to autistic peers requires looking at what children think, what they feel, and what they are actually willing to do.

Methodologically, the team combined numbers with narratives. Each child completed a peer acceptance questionnaire built around three dimensions: cognition, or what children believe about a peer with autism; affect, or how warmly they feel toward that peer; and behavioural intention, or whether they would actually play with, help, or sit beside them. Crucially, the questionnaire used a vignette describing observable characteristics of a hypothetical child, such as differences in communication and play, without ever attaching a diagnostic label. This design choice, made in line with the study’s ethics approval, meant that children responded to behaviour they could see rather than to a category they might not understand, which is itself a revealing feature of how autism is perceived at this age.

The headline numbers are arresting. Across the three dimensions, only about 40 to 50 percent of children expressed explicit acceptance. The remainder was dominated by a response the researchers had not set out to spotlight: the option to say they did not know. These uncertainty answers ranged from 28.3 percent to 37.9 percent across questionnaire items, with nearly every item sitting at or above the 30 percent mark. In other words, for a substantial minority of preschoolers, the honest answer to questions about an autistic peer was neither yes nor no, but a shrug. For a field that has often focused on the balance between acceptance and rejection, this third category may represent a large and previously underappreciated share of young children’s social attitudes.

The pattern was not random noise. Uncertainty responses were highly consistent within individual children, suggesting a systematic response style rather than momentary confusion on a single question. A child who did not know whether they liked playing with the vignette character tended also not to know whether they would help them or sit next to them. That internal coherence implies that some children genuinely lack the cognitive scaffolding needed to form an attitude at all, whether positive or negative. In developmental terms, these children may simply not yet have the concepts, the vocabulary, or the social experience required to evaluate a peer whose behaviour differs from the norm.

Statistical analysis of the children who did express clear views added further texture. The three dimensions of acceptance were all significantly and positively correlated, with correlation coefficients ranging from .59 to .73, all significant at p less than .001. Children who understood autistic peers more positively also tended to feel warmer toward them and to intend more inclusive behaviour. But the correlations were not uniform. A Meng–Rosenthal–Rubin test comparing dependent correlations showed that the link between cognition and behavioural intention was significantly stronger than the link between cognition and affect, with a test statistic of Z = 3.37 and p less than .001. In plain terms, what preschoolers know appears to translate into what they do more directly than into what they feel, hinting that knowledge-based interventions might shift behaviour even before emotional attitudes catch up.

The qualitative half of the study, based on semi-structured interviews with 15 children and 5 teachers and analysed using reflexive thematic analysis, gave that statistical skeleton a human face. The researchers distilled four themes. The first was cognitive misunderstanding and gaps: children held fragmented, sometimes inaccurate ideas about why an autistic peer might behave differently, often lacking any explanatory framework at all. The second theme concerned difficulties in interaction initiation and response, with children unsure how to start a conversation or game, or how to interpret a peer who did not respond in expected ways. The third theme captured the coexistence of affective closeness and estrangement, showing that warmth and distance could live side by side in the same child’s account of the same classmate.

The fourth theme highlighted the critical role, and the limits, of teacher support. Teachers described themselves as central brokers of inclusion, arranging seating, modelling language, and gently coaching interactions. But they also acknowledged that their scaffolding could only go so far, particularly when children’s uncertainty ran deep. One of the most practically useful findings came from the teachers themselves: they perceived structured activities as more conducive to peer engagement than free play. In unstructured time, children with autism and their typically developing classmates alike were left to navigate ambiguity on their own, and uncertainty tended to win. In structured activities, by contrast, adults could engineer shared goals, defined roles, and predictable scripts that gave uncertain children a reason and a method to engage.

Why does this matter beyond the kindergarten walls? Early childhood is when social attitudes crystallise, and uncertainty at age five may harden into avoidance at age fifteen if nobody intervenes. The authors argue that the goal of inclusive education should not only be to convert rejection into acceptance, but to convert uncertainty into active acceptance, a two-step process that begins with giving children the concepts and language they currently lack. That could mean explicit, age-appropriate explanations of autistic behaviour, picture books and stories that normalise difference, and guided peer interactions that let children accumulate positive experiences they can draw on when forming attitudes. It could also mean refining teacher scaffolding so that support is targeted at the specific dimension where a child is stuck, whether that is understanding, feeling, or acting.

The study also carries a methodological lesson for the field. By including and analysing a don’t-know option rather than treating it as missing data, the researchers surfaced a phenomenon that conventional attitude surveys often bury. Their finding that uncertainty is coherent, prevalent, and dimension-spanning suggests that future studies of peer acceptance in young children should routinely measure and report it. For China, where inclusive preschool education is expanding rapidly and where the study was supported by Zhejiang and Hangzhou social science funding programs, the practical stakes are considerable. For the wider world, the message is universal: before children can accept a peer who is different, someone has to help them make sense of the difference. Roughly one in three preschoolers in this study was waiting for exactly that help, and the classroom, with its teachers and its carefully structured moments of shared play, is where that help is most likely to arrive.

Subject of Research: Peer attitudes and uncertainty toward children with autism spectrum disorder in Chinese inclusive preschools

Article Title: Uncertainty in peer attitudes toward children with autism spectrum disorder in Chinese inclusive preschools: a mixed-methods study

Article References: Zhou, D., Yu, M., Chen, J., & Hu, Y. (2026). Uncertainty in peer attitudes toward children with autism spectrum disorder in Chinese inclusive preschools: a mixed-methods study. BMC Psychology. https://doi.org/10.1186/s40359-026-05705-w

Image Credits: AI Generated

DOI: 10.1186/s40359-026-05705-w

Keywords: autism spectrum disorder, peer acceptance, inclusive education, preschool children, uncertainty, ecological systems theory, mixed-methods research, social cognition, China, teacher scaffolding, child development, BMC Psychology

The study’s methodological core is the combination of classic sociometry with modern mixture modeling. In traditional sociometric research, children are asked whom they like most and whom they like least among their classmates, and these nominations are converted into standardized scores for peer acceptance, peer rejection, social preference, and social impact. Kang and Shin fed these four dimensions simultaneously into a latent profile analysis, a technique that searches for hidden subgroups of individuals who share similar patterns across multiple continuous variables. Rather than asking researchers to impose arbitrary thresholds, the method lets the data themselves reveal how many meaningful groups exist and where their boundaries lie. The analysis converged on a three-profile solution that closely mirrors the classical categories of popular, rejected, and average, but derives them empirically rather than by rule.

What elevates the study beyond a descriptive exercise is its second question: what predicts membership in each profile? The researchers drew on three domains that developmental science has long flagged as influential in early social competence. The first is temperament, assessed through the preschool version of Cloninger’s Temperament and Character Inventory, which captures traits such as harm avoidance and reward dependence. The second is executive functioning, the family of self-regulatory abilities that includes inhibition, working memory, and attentional flexibility, measured with the Early Years Toolbox. The third is the quality of the teacher-child relationship, evaluated with the Student-Teacher Relationship Scale, which distinguishes warm, close bonds from relationships marked by conflict and dependency.

The predictive findings are striking in their nuance. Children with low temperamental harm avoidance, low reward dependence, and low behavioral inhibition were more likely to fall into the rejected profile. At first glance this seems counterintuitive; bold, uninhibited children might be expected to charm their peers. The authors’ interpretation is that in the preschool years, a lack of inhibition combined with low wariness can translate into behaviors that classmates find aversive: grabbing toys, interrupting, ignoring social cues, or dominating play. Harm avoidance, often framed negatively in adult personality research, may serve a protective function in early childhood, helping children pause before actions that alienate others.

Equally intriguing is the finding about reward dependence, the temperament dimension reflecting sensitivity to social approval and reinforcement. Children with moderate levels of reward dependence were more likely to belong to the popular profile than children with very high levels. This suggests that an intense hunger for social reward, although it can motivate engagement, may also make a child appear clingy, overeager, or socially demanding to four- and five-year-old peers who are still learning the give-and-take of friendship. Moderate reward dependence, by contrast, appears to hit a developmental sweet spot: enough motivation to seek out and maintain interaction, without the intensity that overwhelms playmates.

The teacher-child relationship emerged as a second powerful predictor, and here the direction of effects is especially consequential for practice. Children in less close teacher relationships were more likely to be rejected by peers, while children with less conflictual teacher relationships were more likely to be popular. This pattern aligns with a growing body of longitudinal work showing that teachers function as social gatekeepers in early classrooms. A warm teacher-child bond can signal a child’s value to the group, while visible conflict between teacher and child may license peer disliking. The authors frame this within Bronfenbrenner’s bioecological model, in which proximal relationships between the child and the adults around them shape development nested within ever-wider systems of influence.

The theoretical stakes of the study extend to how educators understand risk. Peer rejection in early childhood is not a trivial matter of playground politics. Meta-analytic evidence links peer acceptance to academic achievement, and longitudinal studies show that rejection in the preschool years forecasts poorer school adjustment, internalizing symptoms, and later psychopathology. By identifying which children cluster into the rejected profile, and by pinpointing the temperamental, cognitive, and relational factors that push them there, the study offers a screening logic that is more precise than simply flagging children with low acceptance scores. A rejected child with high inhibition poses a different intervention problem from a rejected child with high harm avoidance, and the profile approach makes such distinctions visible.

Executive functioning, notably, played a less dominant role in the analysis than temperament and teacher relationships, though the study contributes to an ongoing debate about which self-regulatory skills matter most for peer standing. Prior longitudinal work has shown that executive function predicts the growth of peer relationships over time, sometimes even more strongly than emotion understanding. The present findings suggest that in a cross-sectional snapshot of four- and five-year-olds, the crystallized temperamental dispositions and the relational climate surrounding the child may be the more immediate differentiators of social group membership. Inhibition did appear in the rejected-profile prediction, tying impulsive action tendencies to peer aversion, but its effect was embedded within a broader configuration of traits rather than standing alone.

The study’s cultural context deserves attention. All participants were Korean children, and the researchers situate their work within a Korean early childhood landscape that has seen intense discussion of children’s social development, including research on classroom surveillance and its effects on teacher practice. Whether the three-profile structure and its predictors replicate in Western samples remains an open question, particularly given evidence that peer relationship dynamics and the salience of particular temperamental traits can vary across cultures. The data and study materials are publicly available through the Open Science Framework, and the study was approved by the Institutional Review Board at Chungbuk National University, transparency measures that should facilitate cross-cultural replication efforts.

For practitioners, the message is clear and actionable. Kang and Shin argue that support for young children’s peer relationships should be tailored to the child’s individual profile rather than delivered as a one-size-fits-all social skills curriculum. A rejected child may need help with behavioral inhibition and impulse control; a child on the edge of popularity may simply need a teacher relationship with fewer conflict-laden exchanges to consolidate peer standing. Teachers, in this framing, are not passive observers of classroom social hierarchies but active participants whose warmth or friction ripples outward into the peer ecology. As early childhood classrooms become increasingly recognized as the crucible where lifelong social trajectories begin, studies like this one provide the empirical map that educators have long needed to intervene early, precisely, and humanely.

Subject of Research: Latent peer relationship profiles and their predictors among young children

Article Title: Latent Profiles and Predictors of Young Children’s Peer Relationships

Article References: Kang, S., & Shin, N. (2026). Latent Profiles and Predictors of Young Children’s Peer Relationships. Early Childhood Education Journal. https://doi.org/10.1007/s10643-026-02352-z

Image Credits: AI Generated

DOI: 10.1007/s10643-026-02352-z

Keywords: peer relationships, latent profile analysis, temperament, executive functioning, teacher-child relationship, peer rejection, peer acceptance, early childhood, sociometry, preschool children, Latent, Profiles