A randomized clinical trial of three prompting systems to teach tact relations

Cihon et al. (2020) randomly assigned 27 children with autism to one of three prompting systems. The goal was to teach new tacts — naming common objects like cup or car.

Each child got one-on-one discrete-trial lessons. Therapists used either constant time delay, most-to-least prompts, or flexible prompt fading. Sessions ran until the child hit mastery.

All three groups learned the new words. Speed, errors, and total trials were almost identical. No single prompting style came out on top.

In short, the choice of prompt hierarchy did not change the outcome for these learners.

Belisle et al. (2020) extends this work. They also used most-to-least prompting with autistic kids, but taught tacts of feelings (hurt, tired). Their positive results line up with Cihon’s — the same prompt style still works when the content changes.

Pritchard et al. (1987) seems to disagree. Their single-case study found progressive time delay faster than least-to-most prompts for object labels. The gap disappears in Cihon’s larger RCT. The old study used learners with intellectual disability and measured minutes to mastery, while Cihon counted trials with autistic children. Different yardsticks, different answers.

Taber-Doughty (2005) adds a twist. When students picked their own prompt style, they learned faster. Cihon held the method constant per group, so learner choice may hide any small edge one prompt could have.

You can stop sweating the prompt hierarchy. Pick the one you can deliver with high fidelity and that your learner tolerates. Spend your energy on clear materials, quick reinforcement, and plenty of trials. If a child shows frustration, switch prompts — the data say any of the big three will work.