An analysis of the generalization of pretend play from real objects to toys

Three preschoolers with autism learned pretend actions with real objects.

They practiced pouring from a real cup, stirring with a real spoon, and talking on a real phone.

Later the team watched if the kids did the same actions when handed toy versions of those items.

All three children used the toys to pretend after training with real items.

Some toys transferred better than others; the toy phone worked best, the toy cup least.

Skills stayed high one month later.

Ptomey et al. (2021) and Brodhead et al. (2019) also got pretend play to generalize, but they used video clips and question-and-answer drills instead of real objects.

Agana et al. (2025) shows you can skip screens and extra talk; just let kids handle the real thing first.

Lee et al. (2020) taught flexible "this stands for that" play with words; Agana’s method is simpler—match real action to toy action.

Beaumont et al. (2008) used the same prompt-and-fade style for peer play, proving the tactic works across play types.

If a child won’t pretend with toys, start with the working kitchen item, real hairbrush, or actual flashlight.

Once the action is solid, swap in the mini toy and give a quick prompt.

You may see faster generalization with no extra programming.