A review of research on procedures for teaching safety skills to persons with developmental disabilities.

McConkey et al. (2010) read every paper they could find on teaching safety skills to people with developmental disabilities. They did not run new experiments. They simply pulled together what others had already tried.

The team looked for studies on street crossing, fire drills, and stranger danger. They wanted to know which teaching tricks worked best.

Three tools showed up again and again: prompts, reinforcement, and role-play. When teachers used all three, most clients learned the safety skill.

The bad news was generalization. Many clients could show the skill at the table but failed during real street or fire drills.

Baruni et al. (2022) is the 2022 update of this same story. They kept the three tools but added two must-dos: test the skill in the real place and keep practicing there. The 2022 paper is the sequel that fixes the generalization gap.

Garcia et al. (2016) gives a live demo. They used the 2010 package—model, rehearse, praise—to teach fire evacuation to kids with autism. All kids learned and still passed a five-week surprise drill. The single-case study is the proof of the 2010 recipe.

Mountjoy et al. (1984) did almost the same thing decades earlier. They taught blind teens fire safety with group rehearsal and token reinforcement. Six of eight teens still passed an unannounced night drill. The old data foreshadows the 2010 advice.

Stop running only table-top safety drills. Pair the 2010 trio—prompt, reinforce, role-play—with real-world tests like Baruni et al. (2022) urge. Start in the classroom, then practice at the actual corner or hallway. Take data on the first real try; you will see at once if generalization is missing and can add more in-situ trials before the live fire alarm rings.