Helping a Child With Autism Report Past Events Using Partial Written Prompts

Many children with autism can answer questions about what’s happening right now but struggle to describe what happened earlier. This skill—reporting past events—matters for social conversations, school routines, and safety. A recent study explored whether partial written prompts during a simple, game-like conversation could help a young child accurately share what she did hours before. Here’s what clinicians can take from it.

What Is the Research Question Being Asked and Why Does It Matter?

This study asked whether “partial written words” during a short, game-like conversation could help a young child with autism describe what she did earlier in the day. The child was asked about three types of events: reading a book, doing a craft, and playing in a gross motor room. The key twist? The materials changed every time—a different book each day, for example—so she couldn’t just memorize one answer.

This matters because many learners can answer “What do you like?” but struggle with “What did you do?” after time has passed. Reporting past events supports social conversation, school routines, and safety questions (like explaining an injury). It also helps caregivers understand what a learner experienced and preferred, which supports choice-making and better planning.

What Did the Researchers Do to Answer That Question?

One 4-year-old girl with autism participated. She could already read short, unfamiliar sentences. The team measured whether she could answer questions about an activity roughly 1 hour later (Opportunity 1), then again 20–40 minutes after that (Opportunity 2). Answers had to match what actually happened to count as correct.

In baseline, the adult asked questions with no extra support. Whether the child answered correctly or not, the adult moved on with brief neutral responses—no prompts, no reinforcement for specific answers.

In intervention, the adult used stacking blocks with partial written text (sentence starters and parts of the questions). The adult and child took turns stacking blocks while moving through the conversation script. If the child didn’t answer or answered incorrectly, the adult used prompts (including echoic prompts) and praised correct answers. If she answered the whole set well, she earned about 5 minutes with a highly preferred item.

The researchers used a multiple baseline across the three activities, starting intervention at different times for book, craft, and gross motor. They also tracked “varied” correct answers during Opportunity 2—meaning the child gave a correct answer different from what she said in Opportunity 1.

How You Can Use This in Your Day-to-Day Clinical Practice

If your learner often says “I don’t know,” guesses, or repeats the same short answer when asked about earlier events, consider adding a simple prompt system that stays visible.

In this study, the written text wasn’t a full script the child read aloud. It was partial text paired with turn-taking and a clear routine, which may have helped her get started and then fill in details from the real event. This approach may fit best for learners who already read some words and can answer simple questions in the moment but struggle later.

You can build a short “talk routine” after a meaningful activity without the exact materials used here. The active ingredient to try: a consistent set of questions plus a visual or text cue that offers a sentence frame (for example, “I played with ___,” “I made ___,” “My favorite part was ___”). Keep the cue visible while the learner answers so they don’t have to hold the whole question in memory.

If the learner uses AAC, the same idea could work with partial text plus the device—but this study didn’t test that.

Plan for two check-ins, not just one. The researchers asked about the event at about 1 hour, then again 20–40 minutes later. In practice, you might try “later today” and “right before going home,” or “after recess” and “after lunch,” depending on the setting. The second check-in is a good time to encourage a different detail (“Tell me something else about it”)—but only after the learner is already successful with basic accurate reporting.

Use reinforcement carefully and respectfully. In the study, the child earned a preferred item after answering the set well, and the adult gave descriptive praise for correct answers. In practice, reinforce participation and accurate sharing without turning it into an interrogation. Offer choices: “Do you want to tell me with words, pictures, or show me?” or “Do you want to answer one question or two?” This protects assent and keeps the skill tied to communication, not compliance.

When you prompt, help the learner contact the right cues—not just copy you. The intervention included echoic prompts when the child was stuck or wrong. If you use echoic prompts, fade them as soon as possible so the learner isn’t just repeating. Shift from full echoic (“Say: I played with Sam”) to partial (“I played with…”) to a silent point to the text cue. The goal is that the question plus the cue helps the learner recall—not that the adult supplies the answer.

Program for novelty like the study did. They changed the materials every day (new books, new craft items, different play). This is a strong reminder: if you always ask about the same snack or game, you may teach a memorized response that doesn’t generalize. Rotate activities and rotate the details you ask about (who, what, where, favorite part) so the learner practices pulling information from real experiences.

Track two things in your data: accuracy and variety. Accuracy means the report matches the event. Variety means the learner can share more than one correct detail across time. Define variety simply—”a second correct detail that isn’t the exact same sentence as earlier.” This helps you avoid “robot answers” that are correct but not useful in real conversation.

Be clear about limits before adopting this widely. This was one child who could already read, and the blocks were never fully faded, so we don’t know if the skill would hold without the cues. We also don’t know which component mattered most: the text cues, the stacking game, the prompts, or the reinforcement. Treat this as a practical idea to try with careful measurement—not a guaranteed package for all learners. If your learner can’t read, or if you need reporting after longer delays (like days later), you may need different supports: pictures, event schedules, or a separate recall strategy.

Finally, check that the learner has the words needed to report. This study didn’t formally test whether the child could label every item used. In practice, if a learner can’t name “glue,” “slide,” or “kite,” they may fail a past-event question even if they remember. Preteach key vocabulary during the activity in a natural way. Accept multiple correct forms of the answer—spoken, AAC, pointing to a photo—so the learner can report with dignity and success.

Works Cited

McWilliams, M. S., & Hanson, R. J. (2025). The use of partial textual stimuli within an interactive task for increasing reports of past behavior with a child with autism. The Analysis of Verbal Behavior, 41, 57–67. https://doi.org/10.1007/s40616-025-00218-w