By Matt Harrington, BCBA · Behaviorist Book Club · Research-backed answers for behavior analysts
Several behavioral factors contribute to elevated drowning risk. Many individuals with autism are attracted to water as a highly reinforcing stimulus, may elope toward it without awareness of danger, and lack the motor skills, swim ability, or self-rescue responses that typically developing peers develop incidentally. Additionally, deficits in responding to caregiver safety signals and generalized 'stop' behaviors reduce the effectiveness of adult supervision as a sole protective strategy. Behavioral intervention directly targets these repertoire gaps through systematic skill building and generalization programming.
BST is a four-component training methodology: verbal instruction describing the skill and rationale, modeling of the correct behavior, rehearsal by the learner in relevant conditions, and immediate feedback (both positive reinforcement for correct responses and corrective feedback for errors). For safety skills, BST is applied across multiple training trials in varied settings until performance is consistent. In-situ probes in naturalistic environments verify that training has produced generalized performance. BST is the evidence-based delivery vehicle for the vast majority of behavioral safety programs.
Preventative safety skills are those that reduce the likelihood of entering a dangerous situation — for example, stopping at a boundary near water, responding to a stop signal, or staying within a designated safe area. Reactive safety skills address what the individual does once exposure to a hazard has occurred — for example, performing a back float to stay afloat, reaching for a pool edge, or calling for help. A comprehensive safety program includes both categories, creating a layered protective system rather than relying solely on prevention.
In-situ probes must be designed so that the individual is assessed for the safety skill without being genuinely exposed to the hazard. This requires a spotter or immediate physical barrier ready to intervene, clear criteria for when to prompt (i.e., if the individual is about to enter actual danger), and careful documentation of what happened and why. The goal is ecological validity — assessing whether the skill occurs in natural contexts — not to allow the individual to come to harm. Written protocols and team training before probes begin are non-negotiable.
Prioritization should be risk-driven. Identify the hazards most likely to cause serious harm given the individual's environment, elopement history, behavioral topographies, and level of supervision. Water access in the home or community, traffic exposure, and cooking or household chemical access are common high-priority areas. After identifying hazards, consider the individual's current skill level — prerequisite skills (responding to stop signals, orienting to name) may need to be established before more complex safety chains can be taught effectively.
Caregivers are both implementation agents and critical partners in generalization programming. Safety skills trained only in clinic settings will not reliably transfer unless caregivers conduct practice trials in the home and community. BCBAs should train caregivers using BST — model the training protocol, have them practice with feedback, then fade to monitoring. Caregivers also need to understand how to conduct in-situ probes appropriately, how to respond when the individual fails a probe (immediate re-instruction and praise for safe behavior), and how to maintain skills over time through periodic practice.
Generalization — the occurrence of a trained behavior across untrained settings, stimuli, and response variations — is the defining success criterion for safety skills. A child who responds correctly to a stop signal in the therapy room but not at the edge of a pool has not achieved the training goal. Programming for generalization requires varying the training environment (different pools, different water settings), the trainer (different instructors), and the stimulus conditions (different times of day, different clothing, different distractors). Multiple-exemplar training and training in natural environments are the most effective generalization strategies.
Several ethical considerations apply. Code 3.01 requires prioritizing client welfare — this supports proactive safety programming when risk is identified. In-situ probes require careful informed consent from caregivers (Code 3.04) and clear explanation that probes may resemble the real scenario. BCBAs must also be aware of scope of competence (Code 2.01) — if swimming instruction is required, collaboration with certified aquatic instructors is appropriate. Finally, the use of any aversive contingencies in safety training (e.g., blocking physical access to a hazardous area) must be addressed through the appropriate behavior plan approval processes per BACB standards.
The behavioral literature supports BST-based safety training across multiple hazard domains for individuals with autism and intellectual disabilities, including fire safety, traffic safety, abduction prevention, and water safety. Published studies consistently demonstrate acquisition to criterion with appropriate BST implementation. Generalization to natural environments is more variable and depends heavily on whether generalization was programmed explicitly. Maintenance over time without periodic practice also tends to decay, supporting the inclusion of scheduled maintenance probes in all long-term safety programs. Research published in JABA and related journals provides specific guidance on protocol features that improve generalization and maintenance.
Documentation should include a risk assessment identifying the target hazards, operational definitions of each safety skill target, the BST protocol used (including training setting, trainer, materials, and mastery criteria), baseline in-situ probe data, session-by-session training data, generalization probe data across settings and trainers, maintenance probe schedule and results, and notes on any modifications made to the protocol. This documentation serves multiple functions: it supports clinical decision-making, demonstrates that the BCBA acted responsibly regarding a known safety risk, and creates a record that can be transferred to other providers if the client transitions services.
The ABA Clubhouse has 60+ on-demand CEUs including ethics, supervision, and clinical topics like this one. Plus a new live CEU every Wednesday.
Ready to go deeper? This course covers this topic with structured learning objectives and CEU credit.
EABA2025 Summer School (No.2): Behavioural Training for Critical Safety Skills — Tia Martin · 1 BACB General CEUs · $0
Take This Course →1 BACB General CEUs · $0 · BehaviorLive
Research-backed educational guide with practice recommendations
Side-by-side comparison with clinical decision framework
All behavior-analytic intervention is individualized. The information on this page is for educational purposes and does not constitute clinical advice. Treatment decisions should be informed by the best available published research, individualized assessment, and obtained with the informed consent of the client or their legal guardian. Behavior analysts are responsible for practicing within the boundaries of their competence and adhering to the BACB Ethics Code for Behavior Analysts.