Do behavioural approaches teach children with autism to play or are they pretending?

Tim and colleagues looked at every paper that tried to teach play to autistic kids with ABA. They asked one blunt question: are the children really playing, or just going through the motions?

The team pulled apart the tricks adults use—tokens, praise, prompts—and checked whether kids kept the new play at home, in the park, or with new toys.

Programs that tapped what a child already liked—sounds, movement, funny faces—worked better than bags of tokens. Kids learned faster and stayed longer with the toys.

Yet almost no study could show the play moved beyond the clinic table. Generalization stayed a mystery box.

Najdowski et al. (2003) had already listed the same tools—DTT, PRT, scripts—so Luckett et al. (2007) is the sequel that adds a warning label: ‘use intrinsic sparks, not stickers.’

Carr et al. (1985) showed pure operant drills barely helped constructive play; the 2007 review echoes that empty performance is not true play.

Pisman et al. (2020) later proved parents can weave language targets into play without killing fun, giving a real-world fix for the generalization gap flagged in 2007.

Dolev et al. (2023) supplies classroom proof: when teachers stay non-intrusive, kids gain more skills over the year—backing the review’s worry that adult-heavy drills can smother real play.

Stop paying kids with candy for every Lego click. Watch what already lights them up—spinning wheels, music, water—and build your prompt sequence there. Plan from day one to fade yourself out and to swap in new toys, peers, and rooms so the play travels. If you run parent or teacher training, show them how to embed targets without taking the lead; the kid should steer, you should follow.