Peer versus adult models and autistic children's learning: acquisition, generalization, and maintenance.
Peers teach new words just as well as adults—so let classmates do the talking.
01Research in Context
What this study did
Four boys with autism watched a peer or an adult say a new phrase. Then the boys tried to say it too.
The team switched who modeled each week. They tracked how fast the boys learned, used the phrase with new toys, and kept it 13 weeks later.
What they found
Peer models worked just as well as adult models. All boys learned the phrases in about the same number of trials.
Both groups used the new words with different toys and still had them three months later.
How this fits with other research
Landry et al. (1989) got the same result and added matrix training. Just four peer examples created a large share of new word combos.
Farmer-Dougan (1994) moved the idea to adults. House-mates taught each other to ask for things during daily routines.
Castañe et al. (1993) looked at active versus passive modeling. Kids learned sight words faster when they had to speak, not just listen.
Why it matters
You do not need an adult for every model. Use classmates, siblings, or peers in circle time. Pair them with brief matrix sets and ask the learner to repeat out loud. This cuts staff demand and builds natural social voices.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →Pick one target phrase, have a peer model it three times, then prompt your learner to echo.
02At a glance
03Original abstract
This study compared the effectiveness of a peer model and an adult model in teaching an expressive language task to four autistic boys. A BCBC design, counterbalanced across subjects, was used. After training criterion was reached, generalization of responding to an extratherapy school setting and to the home was measured. Thirteen weekly maintenance probes were conducted after training in each condition. Results indicated that all children learned through observing the peer and adult models and that few consistent differences occurred across the two conditions. The degree of generalization and maintenance of responding was consistently high in both conditions. The relation of these data to the modeling literature on autistic children and implications for developing educational programs for autistic children are discussed.
Journal of autism and developmental disorders, 1988 · doi:10.1007/BF02211819