Atypical development of social and nonsocial working memory capacity among preschoolers with autism spectrum disorders.

Gong et al. (2023) watched autistic and typically developing preschoolers complete working-memory games. The games used both social pictures, like faces, and non-social pictures, like toys. The team tracked how many items each child could hold in mind across ages three to six.

They used a quasi-experimental design. This means they compared existing groups instead of randomly assigning kids. The goal was to see if the famous five-to-six-year memory jump happens in autism.

The jump did not happen. Autistic preschoolers stayed flat in both social and non-social working memory, while their typical peers surged ahead. The gap was largest for social items.

Kids who remembered fewer faces also had higher autism symptom scores. This link was weaker for toy pictures.

The finding lines up with McGonigle et al. (2014), who showed that autistic preschoolers look less at faces when toys are present. Both studies point to a social attention bottleneck that could starve the working-memory system of social data.

It also extends Carpenter et al. (2002). That early case series first mapped the social-cognitive profile of autistic preschoolers. Linlin et al. now add working memory as a key piece of that puzzle.

Finally, it contrasts with Leezenbaum et al. (2019), who saw slow but positive growth in daily living skills over the preschool years. Memory, unlike brushing teeth, shows almost no spontaneous gains, hinting that direct teaching may be needed.

If working memory is stuck, your instructions may vanish before the child can act. Break tasks into one-step chunks and use visual aids that stay in view. When you target social goals, start with brief, high-interest faces and reward looking and remembering. Track both accuracy and symptom load; the link tells you if memory gains are softening core autism features.