Proposed codes: AC9EFLA05, AC9EFLA07, AC9EFLY04, AC9EFLY05, AC9EFLY06, AC9EFLY07.
Goal: Compose a caption with taught print, let a user test it against an image/description and state one thing the caption leaves unknown.
Prepare: two-view aid, blank caption line, controlled word tiles A | cup | bin | is | on | a | tray | mat, full-stop tile, one willing chosen reader. The word mat is earlier taught; no real objects are implied.
- Retrieve · 3 min. Read A cup is on a tray. with the image hidden. Ask for its end mark and where word gaps go.
- Purpose · 4 min. Show both invented views or exact alternatives. Say “A new reader should be able to use our short caption with either view. Which detail is in both?” Expect the cup on tray. Ask what only the wider view adds.
- Model a test · 5 min. Put A cup is on a mat. on a teacher strip. Sound through mat and compare with the drawing; the cup is on the tray, so this candidate fails. Cross the claim out on the teacher strip. Do not compose the learner's final caption for them.
- Child composes · 7 min. Learner chooses/writes/arranges a supported caption, ideally A cup is on a tray. They may make a different controlled caption only if every word has been taught and the chosen image supports it. Record first independent print choice, capital, spaces, stop and reread before assistance. A richer dictated caption is kept separately as adult-scribed print.
- Real reader check · 4 min. Chosen willing reader hears/sees the caption and finds the evidence in the drawing or alternative. Ask “What did the reader actually find? What is still outside our evidence?” If no user attends, mark user feedback not observed and do not invent it.
- Exit check · 2 min. Ask the learner to read the final caption and name one thing it does not establish, such as who moved the cup. Keep first/revised versions and the exact reader response.
Access routes: large/tactile tiles, raised two-view model or exact description, digital caption line, child-directed adult placement, quiet adult audience, AAC/sign/pointing/typing/drawing for explanation. A correct oral caption without a child print attempt is recorded as oral meaning, not independent decoding.