By Matt Harrington, BCBA · Behaviorist Book Club · April 2026 · 12 min read

From Navy Dolphins to Applied Practice: Bob Bailey's Scientific Legacy in Behavior Analysis

Overview & Clinical Significance

Bob Bailey occupies a singular position in the history of behavior analysis. As a biologist and chemist by training who became one of the foremost animal trainers of the twentieth century, Bailey's career is a case study in what happens when genuine scientific rigor is applied to the practical challenges of behavior change. His work directing the US Navy's Marine Mammal Program, his collaboration with Marian and Keller Breland, and his decades of workshop-based training have shaped how thousands of practitioners understand the mechanics of learning.

For BCBAs and behavior analysts working in applied settings, the significance of Bailey's perspective lies not just in the technical contributions — the Keep Going Signal, systematic data collection in training — but in his insistence on treating training as a scientific enterprise rather than an intuitive art. In a field that is sometimes tempted by jargon fluency without procedural precision, Bailey's voice is a corrective. He argues from experimental logic: if you can't measure it, you don't know if it's working.

This course is framed around an interview format, which provides access to Bailey's thinking in a less filtered way than textbook summaries. Interviews capture the heuristics and principles that guide expert practitioners, including the reasoning behind methodological choices that may not appear in published research. For behavior analysts at any stage of their career, engaging with the intellectual history of the field through a figure like Bailey deepens the conceptual foundation that daily practice rests on.

The clinical significance for a BCBA is this: the principles Bailey applied to training dolphins, pigeons, and other animals are the same principles applied to human learners. The behavioral mechanisms do not change across species. What changes is the context, the communication modality, and the ethical framework. Understanding Bailey's work reinforces the universality of behavioral principles and the importance of applying them with the same precision and data-driven discipline he brought to animal training contexts.

Background & Context

Bob Bailey's entry into behavior analysis followed an unconventional path. Trained as a chemist and biologist, Bailey joined the US Navy in the 1960s to work on the Marine Mammal Program, which sought to harness the navigational and acoustic abilities of marine animals — particularly bottlenose dolphins — for military applications including mine detection and equipment recovery. The program required training complex behavioral repertoires under variable environmental conditions, which forced Bailey and his colleagues to develop systematic, data-driven training protocols far in advance of what was common in the broader animal training community at the time.

Bailey's collaboration with Marian and Keller Breland is a central thread in this history. The Brelands were students of B.F. Skinner who left academic settings to establish Animal Behavior Enterprises (ABE), a commercial animal training company that applied operant conditioning principles at scale. Their famous 1961 paper on instinctive drift — in which trained animals reverted to species-typical behaviors that competed with the reinforced operants — remains one of the most cited demonstrations that biological predispositions constrain operant training in ways that a purely arbitrary stimulus-response model does not predict. Bailey worked alongside and eventually led ABE, carrying forward both the Skinnerian foundation and the pragmatic lessons learned from years of applied work.

The Keep Going Signal (KGS) is one of Bailey's most recognized contributions. In essence, a KGS is a conditioned reinforcer delivered mid-behavior that signals to the learner that the behavior is on the right track and that completing the full response chain will produce the terminal reinforcer. The KGS reduces response latency, improves behavioral fluency, and provides real-time information to the learner without interrupting the behavioral chain. Applied behavior analysts have adapted this concept for use with human learners, particularly in building complex skill chains and in increasing behavioral persistence.

Bailey's emphasis on data collection — his insistence that trainers record every trial, track acquisition curves, and make programming decisions based on data rather than impression — mirrors the data systems that BCBAs use daily. His is the voice of someone who discovered empirically, over decades of practice, that intuition without data leads to inconsistent results and missed learning opportunities.

Clinical Implications

The clinical implications of Bailey's work for BCBAs center on precision, measurement, and the scientific mindset. The first implication is that behavioral procedures must be implemented with technical fidelity. Bailey's training programs for the Navy depended on exact timing of reinforcer delivery, precisely defined response criteria, and consistent stimulus control procedures. In clinical ABA, the same demands apply. A prompt delivered a half-second too late or a reinforcer presented inconsistently across instructors produces different learning outcomes than a precisely calibrated procedure — even when the topographical elements appear identical.

This precision argument has direct implications for treatment integrity assessment, which is a component of ethical service delivery under the BACB Ethics Code. BCBAs who supervise direct-line staff need systems for measuring whether procedures are being implemented as written. Bailey's career is a reminder that the gap between a training protocol on paper and the actual behavior of the trainer in the moment is where most learning failures occur.

The second clinical implication is the value of cross-species generalization as a conceptual tool. When BCBAs encounter a learner for whom a standard procedure is not producing the expected outcome, thinking through the behavioral principles at a species-general level — what would happen if this were an animal training context? — can illuminate variables that are being overlooked. This is not about reducing human learners to animal analogies; it is about using the clarity of an animal training frame to identify whether reinforcement timing, response definition, or stimulus control procedures are the source of the problem.

The third implication is for professional development. Bailey's admonition to adopt a scientific mindset and collaborate across disciplines applies directly to BCBAs navigating complex cases. A BCBA who treats their clinical work as a series of experiments — asking what hypothesis this intervention is testing, what the data would look like if the hypothesis is correct, and what they will do if it is not — is practicing in the spirit of the behavioral science tradition that Bailey represents.

Ethical Considerations

The BACB Ethics Code's foundational commitment to evidence-based practice (Code 2.09) resonates strongly with Bailey's scientific worldview. His insistence that every training claim be backed by data — that trainers not rely on testimonials, traditions, or intuition as substitutes for empirical evidence — aligns directly with the ethics code's requirement that behavior analysts rely on scientific evidence rather than popular opinion or unverified methods.

Code 1.01 addresses the reliance on scientific knowledge, requiring that BCBAs only practice based on established scientific principles. Bailey's career provides a model for what this looks like in practice: continuous data collection, willingness to revise procedures based on outcomes, and skepticism toward claims that cannot be tested or measured. For BCBAs who encounter pressure to adopt non-behavioral approaches — whether from families, educators, or administrators — Bailey's example provides a principled basis for maintaining scientific standards.

There is also an ethical dimension in Bailey's emphasis on the trainer's role in learning failures. Bailey has consistently argued that when training fails, the first question should be: what did the trainer do wrong? This stance runs counter to deficit-based explanations that locate the problem in the learner. In applied ABA practice, the parallel question is: when a client is not making progress, is the program designed correctly? Is treatment integrity adequate? Are the reinforcers still effective? Are the prerequisite skills in place? The ethical obligation to provide effective treatment (Code 2.09) demands this kind of programmatic self-examination before concluding that the learner is the problem.

Code 4.05, which addresses feedback to supervisees, also connects to Bailey's perspective. Bailey's workshop-based training is known for its direct, data-referenced feedback. Supervisors in ABA should provide feedback grounded in observable data — session recordings, treatment integrity scores, acquisition curves — rather than general impressions. This approach models the scientific rigor that Bailey championed and produces more effective supervisees.

Assessment & Decision-Making

Applying Bailey's scientific framework to clinical decision-making means building assessment into every phase of practice. The starting point is operational definition. Bailey's training programs required precise, observable definitions of each target behavior before training began. In clinical ABA, this translates to the requirement that target behaviors and replacement behaviors be defined in objective, measurable terms before programming begins — a basic but frequently underperformed step.

The next element is baseline assessment. Bailey would no more begin a training program without baseline data than a scientist would run an experiment without a control condition. BCBAs should treat the pre-treatment assessment phase with the same rigor: multiple baseline observations, standardized probe conditions, and documentation of current performance levels that provide a genuine comparison point for post-treatment data.

Decision-making within training should be data-driven. Bailey's data sheets tracked every trial; his acquisition curves were reviewed daily. Clinical ABA programs that review data only weekly — or only when there is an obvious problem — miss the early signals that would allow course corrections before significant time is lost. Implementing more frequent data review cycles, particularly for new programs or struggling learners, is a direct application of Bailey's operational philosophy.

The KGS concept offers a specific assessment-informed tool. Before implementing a KGS in a clinical program, BCBAs should assess whether the target behavior involves a chain that benefits from mid-chain feedback, whether the learner has a history with conditioned reinforcers that would support a KGS protocol, and whether the terminal reinforcer is potent enough to sustain the chain. These are the kinds of prerequisite questions Bailey would ask before implementing any novel procedure.

Cross-disciplinary collaboration — another of Bailey's themes — translates clinically into consultation with speech-language pathologists, occupational therapists, educators, and physicians when relevant. BCBAs who operate as isolated experts miss information that other disciplines hold, and the resulting programs are less comprehensive than they could be.

What This Means for Your Practice

The most direct takeaway from Bailey's interview is an invitation to examine the scientific quality of your own practice. Not the scientific language you use, but the actual scientific process: Are you collecting data on every session? Are you reviewing it frequently enough to detect problems early? Are you making decisions based on objective records or on clinical impression? Are you testing your interventions rather than assuming they work?

For most BCBAs working in busy clinical settings, the honest answer is that the ideal of continuous, precise data collection is partially compromised by caseload demands, documentation requirements, and the practical constraints of working with children in real environments. Bailey's example is not an indictment of everyday practice — it is a standard to aspire toward and a framework for identifying where precision is being sacrificed and what it costs.

One concrete application is auditing treatment integrity. Pull a sample of session recordings or direct observation data and compare actual implementation against written protocols. Identify where the deviation is occurring — prompting timing, reinforcer delivery, response definition — and address it in supervision. This is the kind of programmatic self-examination that Bailey's scientific mindset demands.

Another application is in how BCBAs communicate with families and other professionals. Bailey's emphasis on data over anecdote is a model for family meetings and interdisciplinary consultations. When discussing a child's progress, leading with data — objective acquisition curves, frequency counts, percentage correct — rather than clinical impressions builds credibility and keeps the conversation anchored to what is actually happening.

Finally, the cross-disciplinary spirit Bailey advocates suggests that BCBAs should be intellectually curious beyond the boundaries of their certification. Reading ethology, comparative psychology, and the history of behavioral science — including its animal training applications — enriches the conceptual repertoire available for solving complex clinical problems. Bailey's career is itself an argument for the value of bringing a broad scientific education to the practical work of behavior change.