Improving pretend play for children with autism through experiencing the stimulus properties of real objects

Audras-Torrent et al. (2021) worked with three preschoolers with autism.

The kids first explored real objects with their senses—smelled coffee, felt sandpaper, tasted lemon.

After each brief sensory round, the team asked the child to tact pretend features of toy versions.

They used a multiple-baseline design across behaviors to see if real experience made symbolic talk easier.

All three children quickly labeled pretend properties after touching, tasting, or smelling the real thing.

The new tacts lasted 7–10 weeks with no extra training.

Even a plain stick became ‘ice cream’ or ‘microphone’ once the child had sensed cold cream or heard loud music.

Welsh et al. (2019) also used natural-environment teaching to help kids tact senses, but focused on what other people see, hear, or smell.

Both studies show NET works for sensory language; Lee extends the idea into pretend play.

Akers et al. (2018) taught peer play with activity schedules and scripts.

Their kids learned hide-and-seek rules, while Lee’s kids learned flexible object labels—two different play pieces that fit together.

Koldas et al. (2025) used discrete trial training to teach reciprocal tacting.

Lee’s sensory-first NET approach gives you a quicker, less structured option for the same population.

You can add a 30-second sensory warm-up before pretend play. Let the child sniff real coffee beans, then offer a toy cup. The brief real experience unlocks richer, flexible language for toys and even random items like sticks. No extra drills, no fancy materials—just natural senses doing the teaching.