ABA Fundamentals

A Comparison of Medium Probability Versus High Probability Instructions to Increase Cooperation in the Context of the High Probability Instructional Sequence

Wilder et al. (2025) · Behavioral Interventions 2025
★ The Verdict

Medium-probability warm-ups can replace high-p ones for some kids with autism, so you can run the compliance sequence even when ‘easy’ tasks are scarce.

✓ Read this if BCBAs running discrete-trial or tabletop sessions with young children with autism.
✗ Skip if Practitioners working with older verbal teens or populations without compliance goals.

01Research in Context

01

What this study did

Wilder and team tested three boys with autism, during tabletop work at an ABA clinic. They compared two ways to start a demand chain: three quick high-probability instructions (like 'clap hands') versus three medium-probability ones (like 'touch cup').

An alternating-treatments design flipped the starting type every session. The goal was to see if the kids next agreed to a low-probability task (a hard academic demand) just as often after the medium-p warm-ups.

02

What they found

Two boys cooperated almost every time after either warm-up; the third boy showed only a small boost with medium-p cues. Overall, medium-probability instructions lifted compliance nearly as high as the classic high-p sequence for most kids.

Data paths for the two successful boys overlapped by 85-90 %, so the team called the medium-p option 'functionally equivalent' for them.

03

How this fits with other research

Chandler et al. (1992) also raised cooperation in kids with autism, but they used self-management instead of any prompting chain. Both papers show you can get more compliance without new reinforcers—just change how the child enters the task.

Chetcuti et al. (2023) remind us that early temperament stays stable; kids who start out wary may need the extra punch of true high-p cues. Wilder’s third boy fits that picture, so keep temperament in mind when you pick the warm-up level.

Kaya et al. (2025) took a parent-view approach and found pets help social spark. Wilder gives you a clinic-ready tactic for the same population: if you lack enough sure-fire high-p commands, medium-p ones can still spark that first yes.

04

Why it matters

You no longer need a long list of slam-dunk high-p instructions. Start with three easy-but-not-guaranteed requests you already have—‘sit down’, ‘feet still’, ‘look’. If the child obeys, jump straight to the hard task. Track for a week; if cooperation lags only for certain kids, reserve true high-p cues for them and save time for everyone else.

Free CEUs

Want CEUs on This Topic?

The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.

Join Free →
→ Action — try this Monday

Pick three already-mastered but not perfect instructions, use them as the warm-up chain, and measure next-demand compliance for one week.

02At a glance

Intervention
prompting and fading
Design
alternating treatments
Sample size
3
Population
autism spectrum disorder
Finding
positive
Magnitude
medium

03Original abstract

ABSTRACTThe high‐probability instructional sequence has been shown to be effective to increase cooperation with low‐probability requests. However, for some individuals, it may be difficult to identify high‐probability instructions, and some high‐probability instructions may become less likely to evoke cooperation over time. Thus, under some circumstances medium probability instructions, or instructions which may be less likely to evoke cooperation than high‐probability instructions, may be a useful temporary alternative to increase cooperation. In the current study, we compared medium probability instructions to high probability instructions to increase cooperation among three children with autism spectrum disorder. The results showed that for two participants, the medium probability instructions improved cooperation as much as the high‐probability instructions. For a third participant, the medium probability instructions improved cooperation over baseline, but not to the level observed with the high‐probability instructions. Results are discussed in terms of the mechanisms responsible for the effects of instructional sequences.

Behavioral Interventions, 2025 · doi:10.1002/bin.70001