In one sentence
When novel words were embedded in shared picture-book reading, both children with developmental language disorder and peers with typical language recalled word forms and meanings better after repeated spaced retrieval than after repeated study alone.
What the researchers did
Souto, Leonard, Deevy, Christ, Karpicke, Schroeder and colleagues took a lab-proven retrieval-practice design and placed it inside everyday shared book reading. Thirteen children with developmental language disorder (mean age about 59 months) and fourteen children with typical language development (mean age about 57 months) learned eight novel words presented across two illustrated children’s books.
Half the novel words appeared in a repeated spaced retrieval condition; half in a repeated study condition. Children learned both word forms (for example, a made-up sound pattern) and arbitrarily assigned meanings across two learning sessions. Tests followed five minutes after the second session and again one week later, including recall and recognition.
What they found
- Both groups showed better recall of novel words in the spaced retrieval condition than in the repeated study condition.
- The retrieval advantage held for both word forms and their meanings.
- A recognition test did not show a clear difference between conditions.
- Children with typical language scored higher than the DLD group on word-form recall and recognition.
- Both groups showed only a slight decline in word-form recall after one week.
- There were no interactions reported that would overturn the main retrieval benefit pattern.
What this means for learners and educators
- Build ask-and-retrieve moments into shared reading — pause, prompt the new word or meaning, then continue — rather than only re-hearing the story.
- Spaced retrieval can travel out of the lab into a book-sharing routine for preschool-age learners, including children with DLD.
- Expect recognition checks to look friendlier than recall; design outcomes around saying the word when that is the goal.
- Keep the gap manageable: this design used two sessions plus a one-week check, not endless cramming in one sitting.
Limitations and what we don't know yet
Samples are small (13 and 14). Novel words and illustrated books were tightly controlled — real bedtime books vary more. Recognition did not mirror the recall benefit, so test format matters when you interpret “learning.” Future work still needs closer approximations to everyday shared reading (who prompts, how often, how noisy the home is). Benefits for DLD do not replace comprehensive language intervention.