Generative AI sharpens laryngeal lesion classification from radiomic
A new study used two methods to explore how pretend play differs among autistic children at three support levels described in the DSM-5-TR. The results showed a clear pattern. The researchers studied 42 autistic children aged 4 to 7 years. There were 14 children at each of the three support levels: Level 1, Level 2, and Level 3. They used a test called the Child-Initiated Pretend Play Assessment 2, or ChIPPA2. Pretend play is important for child development. It can help show how a child starts playing, keeps a game going, and shares play with others. Before this study, little was known about how pretend play differs among children at these three support levels. The study was published as an open-access paper in BMC Pediatrics. It found clear differences among the groups. Children at Level 1 showed the most complex pretend play, children at Level 3 showed the least complex play, and children at Level 2 fell between the two groups.
The researchers used scores to measure different parts of pretend play. One measure was the Percentage of Elaborate Play Actions, or PEPA. It showed how detailed and long-lasting a child's pretend actions were. Another measure was the Number of Object Substitutions, or NOS. This counted how often a child used one object as if it were something else, an important part of pretend thinking. The researchers also measured the Number of Imitated Actions, or NIA. They scored signs of supportive and non-supportive play and used the Enjoyment in Play Scale to measure how much children enjoyed playing. They compared the groups using Kruskal-Wallis tests. They also used Quade's rank analysis of covariance to account for differences in age and sex. This helped them check that the results were not simply due to some children being older or to differences in the numbers of boys and girls. The researchers found significant differences among the three support levels in PEPA scores, symbolic and combined NOS scores, NIA scores, supportive and non-supportive signs of play, and enjoyment of play.
The researchers also studied the children's play by watching video recordings of their sessions. They wrote detailed descriptions of what they observed and examined these descriptions using a method called qualitative content analysis. This led them to identify four main themes: how detailed and imaginative the play was, how children interacted with others during play, reduced involvement and breaks in play, and concrete, repetitive, and sensory-based ways of exploring objects. The study found that children at Level 1 engaged in more detailed pretend play and kept it going for longer. At Level 2, pretend play was less connected and developed. At Level 3, pretend play was brief and limited. Children who needed more support were more likely to focus on the physical and sensory features of objects. Their play often centered on what objects actually were and how they felt, rather than on what the objects could represent in an imaginary game.
When the researchers compared the test scores with what they saw in the videos, both methods pointed to the same pattern. Lower scores were linked to play that lasted for less time, involved less sharing with others, and included less use of imagination. This agreement is important for clinical practice because the numbers and the video observations showed the same underlying differences in two ways. For example, a child with a low PEPA score, meaning they used few detailed pretend actions, might also be seen stopping play often, not sharing play with the examiner, and using objects only for their real-world purpose instead of imagining other uses for them. Differences in the Enjoyment in Play Scale also showed that how much children enjoyed pretend play varied across support levels. The differences were not only about how well children could perform pretend actions.
The authors concluded that pretend play differed across the three support levels in autistic children and that the ChIPPA2 provided useful clinical information about how children started, continued, enjoyed, and shared their play. Because the assessment measures how children begin pretend play, develop their ideas, copy actions, and enjoy playing, the researchers suggested that it could help professionals assess each child and plan more individual support. This could help match care to a child's actual play abilities instead of relying only on their support level. The study used equal-sized groups and accounted for age and sex, which strengthens the evidence for the pattern found. The video observations also helped connect the scores to what children actually did during play. For clinicians, the main lesson is that assessing pretend play can reveal more than whether a child has learned particular play skills. It can show how an autistic child develops imaginative play, keeps it going, and shares it with others, helping identify where support may be most useful.
Subject of Research: Pretend play performance across support levels in autistic children: a mixed methods study
Article Title: Pretend play performance across support levels in autistic children: a mixed methods study
Article References: Caran, F. H. S., Stagnitti, K., Lucisano, R. V., Pfeifer, L. I., & Sposito, A. M. P. (2026). Pretend play performance across support levels in autistic children: a mixed methods study. BMC Pediatrics. https://doi.org/10.1186/s12887-026-07721-6