Effects of prompting and reinforcement of one response pattern upon imitation of a different modeled pattern.
Prompts plus praise create a response that later modeling rarely changes—so teach or model in the order you want to keep.
01Research in Context
What this study did
The team worked with three preschool kids who had no diagnoses. Each child sat at a box with two holes. The adult first used hand-over-hand prompts to make the child drop marbles in the left hole only. Each drop earned loud praise.
Next the adult modeled a new pattern: drop marbles in the right hole only. The kids watched the model many times. The researchers counted which hole the child used after each phase.
What they found
All three kids kept using the left hole even after seeing the right-hole model dozens of times. The early prompted pattern stuck. When the order was flipped—model first, then prompts—the kids switched to the prompted side.
The sequence of training versus modeling decided the final response. Prompts plus praise created strong stimulus control that modeling alone could not shift.
How this fits with other research
Burack et al. (2004) showed modeling beats instructions for teaching kids to track reinforcement schedules. Both studies agree that the way you first build a response decides later control. The 1982 paper adds that once prompts lock in the pattern, modeling loses its power.
Vie et al. (2017) and Yuan et al. (2020) extended the same prompt-plus-praise tactic to children with autism. They found the method still works, whether the goal is picture recall or error correction. The core rule—prompt first, reinforce immediately—holds across populations.
Wilson et al. (2020) compared two kinds of video modeling. They saw modeling win only when no prior trained pattern blocked it. This supports the 1982 warning: sequence matters. Train first, model later, and the trained form will probably stay.
Why it matters
If you want a child to copy a new model, check what you already taught. A prompted and praised response will outrank a later model. Either teach the model first, or plan extra trials to overwrite the old pattern. This saves you from puzzling ‘why won’t they imitate?’ moments in session.
Get CEUs on This Topic — Free
The ABA Clubhouse has 60+ on-demand CEUs including ethics, supervision, and clinical topics like this one. Plus a new live CEU every Wednesday.
Run a quick probe: before you show a video model, check if the learner already has a prompted history with the task; if yes, insert extra unprompted trials or model first next time.
02At a glance
03Original abstract
Twelve preschool children participated in a study of the effects of explicit training on the imitation of modeled behavior. The responses trained involved a marble-dropping pattern that differed from the modeled pattern. Training consisted of physical prompts and verbal praise during a single session. No prompts or praise were used during test periods. After operant levels of the experimental responses were measured, training either preceded or was interposed within a series of exposures to modeled behavior that differed from the trained behavior. Children who were initially exposed to a modeling session immediately imitated, whereas those children who were initially trained immediately performed the appropriate response. Children initially trained on one pattern generally continued to exhibit that pattern even after many modeling sessions. Children who first viewed the modeled response and then were exposed to explicit training of a different response reversed their response pattern from the trained response to the modeled response within a few sessions. The results suggest that under certain conditions explicit training will exert greater control over responding than immediate modeling stimuli.
Journal of the experimental analysis of behavior, 1982 · doi:10.1901/jeab.1982.37-135