The Effects of Artificial Intelligence on Implementors’ Fidelity of Instructional Strategies During Handwashing Acquisition in Children with Autism
A tablet AI coach hands staff mastery-level hand-washing prompting in record time, yet kid gains still vary.
01Research in Context
What this study did
Griffen et al. (2023) tested a tablet app called GAINS. The app talks staff through hand-washing lessons with kids who have autism.
It gives least-to-most prompts, total-task chaining, and timed delays. The study checked if the AI coach could lift staff fidelity fast.
What they found
Staff hit mastery-level fidelity almost right away. Child hand-washing scores, however, went up for some kids and stayed flat for others.
How this fits with other research
Shillingsburg et al. (2022) and Green et al. (2020) also got staff to high fidelity fast, but they used live coaches plus slides. Griffen shows a tablet AI can do the same job with less manpower.
Ruppel et al. (2023) used Zoom feedback and saw strong upkeep. Griffen did not track long-term upkeep, so we don’t yet know if the AI coach sticks.
Parsons et al. (2019) ran a big RCT on tablet add-ons for kids and found only tiny child gains. Griffen sees the same pattern: staff soar, child skill mixed. The two studies look opposite until you notice both measure child outcome, not staff outcome, as the bottom line.
Why it matters
If you train aides or parents, GAINS gives you a turnkey way to lock in perfect prompting right away. Keep measuring the learner’s performance too, so you know when to tweak the task itself instead of the teaching.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →Load the GAINS demo on one tablet, run one five-minute hand-washing trial with a tech, and score both fidelity and child steps.
02At a glance
03Original abstract
Handwashing is a vital skill for maintaining health and hygiene. For individuals with intellectual and developmental disabilities (IDD), such as autism spectrum disorder, evidence-based strategies, such as prompting and task analysis, may be effective in teaching these skills. Due to the shortage of experts who teach individuals with IDD skills such as handwashing, staff working with children need a means of ensuring these instructional strategies are implemented with fidelity. This study examined the effects of a tablet-based application that used artificial intelligence (GAINS®) on four behavior technicians’ implementation of least-to-most prompting, total task chaining, and time delay during an acquisition of handwashing program with young children with autism. All four technicians increased fidelity immediately upon using GAINS and all four technicians reached mastery criteria within the shortest number of sessions possible. One child participant met mastery criteria, two showed some gains, and one demonstrated a high degree of variability across sessions. Limitations of the least-to-most prompting procedure, user design, considerations and directions for future research and practice are discussed.
Journal of Developmental and Physical Disabilities, 2023 · doi:10.1007/s10882-023-09937-1