Teaching an Infant to Request Help.

The authors pulled every experiment that compared constant time delay (CTD) with progressive time delay (PTD) for kids with autism.

They screened hundreds of papers and kept 22 that met strict rules. Then they counted errors, trials to mastery, and how often teachers had to change the procedure.

Both methods taught the new skill, but CTD made kids produce more wrong answers.

Teachers using CTD also stopped the script more often and it took longer for the cue to move from prompt to natural signal.

In short, PTD cut errors and kept the lesson on track.

Green et al. (1987) already said delayed prompting works, yet they warned we rarely check if the skill lasts. The new review echoes that gap—most CTD/PTD studies still skip maintenance probes.

Markham et al. (2020) found physical prompts beat gesture prompts for speed, but they did not test timing. McQuaid et al. (2024) now shows that even with the same prompt type, the timing pattern (fixed vs stretching) changes the error rate.

Frank‐Crawford et al. (2024) scanned 82 DTT papers and flagged the same weak spots: few teams report generalization or social validity. The CTD/PTD review fits right into their map—same success story, same missing pieces.

If you run discrete trials, choose progressive time delay to spare your learner extra errors and yourself extra corrections. Track maintenance and ask caregivers if the skill helps at home—those steps are still missing in most published programs.