Embedded, constant time delay instruction by peers without disabilities in general education classrooms.

Jameson et al. (2008) asked: can regular classmates teach kids with intellectual disabilities new skills? They trained four middle-school peer tutors in one 30-minute session. The package was simple: a one-page script, a live demo, and two practice rounds with feedback.

Peers then ran constant time delay trials during normal class. They gave a direction, waited four seconds, then gave a full prompt if needed. Target skills were things like reading sight words or naming coins. Sessions lasted only five minutes and happened two to three times a day.

All four peer tutors hit a large share correct teaching steps after the first day. Students with intellectual disabilities learned every skill to mastery in 5–12 sessions. Teachers, parents, and the peers all rated the program as helpful and easy to use.

No one needed extra rewards. The kids kept working because the quick turns and praise were fun.

Weitz (1982) did the first peer-tutor study with students who had intellectual disabilities. That team used edible rewards and long training. Jameson et al. (2008) shows you can drop the candy and still win if you script and demo well.

Paul et al. (1987) proved classwide peer tutoring boosts spelling for typical first graders. Matt’s team narrows the same idea to short, embedded trials for kids with significant delays. Same setting, different focus.

Kay et al. (2020) found prompt type matters for kids with autism. Matt’s work says the delivery agent also matters: peers can run constant time delay just as well as adults.

You can clone this tomorrow. Pick a classmate, hand them the script, watch them demo once, and send them back to their seat with a timer. Five-minute bursts keep the noise down and the data up. Kids with ID get more trials without pulling a parapro away from the group. Everyone gains social status and skills at the same time.