A Longitudinal Study of Joint Attention, Motor Imitation and Language Development in Young Children with Autism Spectrum Disorder in Taiwan.

Researchers in Taiwan watched 74 toddlers with autism every six months over the study period. They scored how well each child responded to joint attention and copied hand motions. They also tracked receptive and expressive language growth.

No treatment was given. The team simply wanted to see which early skills forecast later language.

Kids who looked when an adult pointed or showed a toy at age two had bigger vocabularies at age three and a half. The same was true for kids who could copy simple hand movements like clapping or waving.

The link stayed strong even after the team controlled for starting language level.

Chen et al. (2001) extends these results. They trained typical peers to run play groups for autistic early-elementary kids. Joint attention and words jumped up and stayed high after the peers took over, showing you can speed up the same cascade Lai-Sang observed.

Myers et al. (2018) seems to disagree at first glance. In toddlers with Down syndrome, responding to joint attention predicted language, but in typically developing toddlers mom's happy talk did the job instead. The Taiwan ASD toddlers follow the Down-syndrome pattern, not the typical one, so the pathway is diagnosis-specific, not universal.

Geurts et al. (2008) backs the measurement choice. They showed that watching kids play naturally gives the same joint-attention scores as the long gold-standard test, so you can replicate Lai-Sang's design without extra equipment.

If you work with toddlers who have autism, build programs that teach kids to follow your point and copy your hand moves. These two cheap, easy targets are not just developmental markers—they are springboards for later receptive and expressive language. Add peer-play elements from Chen et al. (2001) to turn the predictor into an intervention and you may see faster gains.