Acquisition, generalization, and maintenance of question-answering skills in autistic children.

The team taught three autistic children to answer wh-questions like "What is the girl doing?"

Each child sat at a table. The teacher showed a picture card, asked a question, modeled the answer, and gave praise or a tiny toy for a correct response.

Sessions ran in short discrete trials until the child answered new questions without help.

All children learned to answer the trained questions. They still got most answers right weeks later.

When new pictures had no clear cue, extra teaching rounds were needed. The authors warn: plan for that extra step.

Delamater et al. (1986) saw the same warning sign first. They compared picture cards with real objects while teaching object names. Kids who saw only pictures later struggled to name the real thing. The two studies agree: pictures work for first learning, but add real-world practice later.

Mattson et al. (2023) repeated the idea with play talk. They paired generic picture cues with social scripts and also saw gains plus partial generalization. The pattern holds across 34 years: picture prompts jump-start language, then you fade them or add new settings.

Alfuraih et al. (2024) stretched the approach to toddlers with severe delays. PECS picture training created new requesting, and the skills lasted. The target wh-question study and the newer PECS study form a bridge: once a child can request with pictures, you can move to answering questions with the same tool.

You can start wh-question programs tomorrow with simple picture scenes, praise, and tiny reinforcers. Just budget extra trials for questions that lack obvious visual hints, and later test the answers with real objects or new photos. This keeps the skill useful outside the therapy room.