Translated from the Portuguese original. Read the original (Português)
The Experimental Analysis of Behavior is the basic-research arm of the science of behavior. It doesn't try to solve anyone's problem right away: it tries to discover how behavior works, under laboratory conditions so controlled that alternative explanations are reduced to a minimum. Reinforcement, extinction, schedules, stimulus control — much of what shows up today in an intervention plan began as a bench question. What distinguishes this field from Applied Behavior Analysis is not the species studied or the setting: it is the question. One searches for principles; the other, for solutions to socially relevant problems.
The Experimental Analysis of Behavior (EAB) is the basic-research domain of the science of behavior: it investigates behavior in highly controlled environments, with the aim of discovering, expanding, and clarifying the fundamental principles that govern the relation between behavior and environment. Founded by B. F. Skinner starting from The Behavior of Organisms (1938), it consolidated a method of its own, marked by the study of the individual organism, by repeated measurement of responding, by within-subject comparisons, and by visual analysis of data. It is one of the field's four interlocking domains, alongside radical behaviorism (the philosophy), Applied Behavior Analysis (applied research), and professional practice.
Four domains, one science
People who come to behavior analysis through the door of the clinic usually know only one piece of it. The field, however, is organized into four domains that feed into one another. It helps to think of this as a chain: philosophy says what kind of explanation counts as an explanation; basic research discovers the principles; applied research checks whether those principles solve real-life problems; and professional practice delivers services informed by that science — in working with one person, but also in consulting with a school or training a team.
The top-down direction is the easiest one to tell as a story, but telling it only that way would be misleading. Traffic runs both ways: a clinical impasse — why did a behavior that had already stopped come back? Why did this reinforcer lose its effect? — becomes a laboratory question, and the answer returns, refined, to application. That back-and-forth is what the field calls translational research.
What happens in the laboratory
The popular image — a rat in a box — isn't wrong, just incomplete. The typical arrangement is an experimental chamber (the famous "Skinner box") in which the animal can operate a device, such as a lever or a disk, and receive a programmed consequence. The decisive detail isn't the box: it's the free-operant method.
In a discrete-trials procedure, the researcher presents one opportunity, records a hit or a miss, and moves to the next. In the free-operant method, there are no trials: the subject can respond whenever and however many times it wants, and what gets measured is the rate — how many responses per unit of time. This choice opened the way to seeing behavior as something that flows, speeding up and slowing down as consequences change.
The instrument that made this visible was the cumulative recorder: a strip of paper running at constant speed under a pen that steps up one notch with each response. The graph never goes down, so all the information is in the slope — a steep rise means a high rate; a gentle rise, a low rate; a flat stretch, a pause.
It was by reading records like these that Charles Ferster and Skinner wrote Schedules of Reinforcement (1957), the catalog that showed something decisive: changing only the rule for delivering reinforcement — and nothing else — already produces distinct, stable, and replicable response patterns.
The hallmarks of the method
The Experimental Analysis of Behavior developed a style of doing research quite different from what predominated in experimental psychology. Five choices sum it up:
- Response rate as the basic datum. Instead of asking "right or wrong?", the question is "how often?" — a continuous measure, sensitive to small changes. See dimensions of behavior measurement.
- Repeated, continuous measurement of clearly defined response classes, rather than one or two before-and-after readings.
- Within-subject comparisons. The subject is its own control: the same organism is compared under different conditions, instead of comparing groups of different individuals.
- Visual analysis of data represented graphically, instead of relying on tests of statistical significance — the effect has to show up in the graph itself.
- Describing functional relations between behavior and environmental variables, rather than testing formal theories.
This method was systematized in Murray Sidman's Tactics of Scientific Research (1960), and it is the methodological foundation that applied research built on. The field also got a home of its own: the Society for the Experimental Analysis of Behavior, founded in 1957, launched the Journal of the Experimental Analysis of Behavior (JEAB, 1958) the following year — and, in 1968, created the Journal of Applied Behavior Analysis (JABA), where applied research began to publish.
Experimental or applied: what really separates themIt's not the species, nor the lab coat, nor the setting. There is basic research done with people and applied work conducted in arrangements analogous to a laboratory's. The difference lies in the question: experimental research wants to know how behavior works; applied research wants to change a socially relevant behavior and demonstrate that it was responsible for the change.
Why rats and pigeons?
It's the question everyone asks — and it has two answers, plus one mandatory caveat:
Some processes cut across species
The field's bet is that certain basic principles — behavior followed by a consequence tends to change in frequency; removing the consequence weakens responding; environmental signals come to control when to respond — are broad biological processes, not quirks of a single species. When the same effect shows up again in different species, and in people too, confidence grows that there is a principle there, not an accident. This bet has limits, and each extension to a new species has to be demonstrated, never assumed.
A degree of control unattainable outside the laboratory
To claim that it was the reinforcer that changed the behavior, everything else has to be ruled out. In the laboratory, it's possible to standardize the organism's experimental history and keep lighting, time of day, level of deprivation, and physical environment constant, varying only one thing. With a child in a school that's impossible — and shouldn't be attempted. It is precisely because people's lives can't be controlled that science needs a place where control is possible.
Animal research is regulated by law
In Brazil, the scientific use of animals is governed by Law No. 11,794/2008, the Arouca Law. It created CONCEA (National Council for the Control of Animal Experimentation) and requires every institution that breeds or uses animals in teaching and research to be accredited and to have an Animal Use Ethics Committee (CEUA) — a body made up of veterinarians, biologists, faculty, researchers, and representatives of animal-protection societies. The CEUA reviews procedures in advance and must order the suspension of any procedure that doesn't comply with the law until the irregularity is corrected. Research with human participants goes through its own ethical review, with informed consent.
Discoveries that came off the bench
Four examples of findings that were born in basic research and today shape how the field thinks. Tap the cards:
Beware of the shortcutA principle demonstrated in the laboratory is not, by itself, a clinical prescription. Between "pigeons match responses to reinforcers" and "do this with this child" lies a mandatory path of applied research, replication, and social validity assessment. Be wary of anyone who skips that step — whether they dismiss basic research as "rat stuff" or use it to justify any procedure at all.
The bridge: translational research
Mace and Critchfield (2010) described well the risk that looms over both sides: purely basic research can become disconnected from the natural world it aims to explain, and purely applied research can become disconnected from the fundamental processes that sustain it. Translational research is the deliberate attempt to stitch the two together — importing questions from application into the laboratory and carrying methods and concepts from the laboratory into the clinic.
The field's best-known examples follow exactly that path. The functional analysis of challenging behaviors — whose reference procedure was described by Iwata and colleagues in 1982 — carried into the clinical setting the experimental logic of turning conditions on and off to identify a behavior's function. The token economy applies what basic research learned about conditioned reinforcers — stimuli with no reinforcing value of their own, which come to function as reinforcement through their association with other reinforcers, like the token that is later traded for something desired. And relapse research uses laboratory procedures to anticipate when a therapeutic gain is at risk of coming undone.
A Brazilian tradition too
The field's history in Brazil begins early. In 1961, Fred S. Keller arrived as a visiting professor at the University of São Paulo through the Fulbright program and, together with Carolina Bori, Rodolpho Azzi, and Myrthes Rodrigues do Prado, helped set up the country's first experimental-analysis-of-behavior laboratories. In August 1964, at the University of Brasília, the first course in the Experimental Analysis of Behavior in Brazil was taught — and Carolina Bori would go on to become the field's main driving force in the country.
That seed gave rise to an active community: ABPMC, the field's Brazilian association, founded on November 4, 1991, has held its annual meeting ever since — considered the largest Brazilian forum on behavior analysis and one of the largest in the world; and the Revista Brasileira de Análise do Comportamento (REBAC, Brazilian Journal of Behavior Analysis), published by the Behavior Theory and Research Center at UFPA, in Belém, has been circulating since January 2005.
Test your understanding
Quick quiz
Three questions to lock in the concept. You'll see the explanation right away.
Further reading
- Skinner, B. F. The Behavior of Organisms: An Experimental Analysis. Appleton-Century, 1938.
- Ferster, C. B. & Skinner, B. F. Schedules of Reinforcement. Appleton-Century-Crofts, 1957.
- Sidman, M. Tactics of Scientific Research: Evaluating Experimental Data in Psychology. Basic Books, 1960.
- Cooper, J. O., Heron, T. E. & Heward, W. L. Applied Behavior Analysis. 3rd ed., Pearson, 2020 — ch. 1 (domains of the science of behavior).
- Herrnstein, R. J. “Relative and absolute strength of response as a function of frequency of reinforcement”. Journal of the Experimental Analysis of Behavior, 4, 267–272, 1961.
- Iwata, B. A., Dorsey, M. F., Slifer, K. J., Bauman, K. E. & Richman, G. S. “Toward a functional analysis of self-injury”. Analysis and Intervention in Developmental Disabilities, 2, 3–20, 1982 (reprinted in Journal of Applied Behavior Analysis, 27, 197–209, 1994).
- Mace, F. C. & Critchfield, T. S. “Translational research in behavior analysis: Historical traditions and imperative for the future”. Journal of the Experimental Analysis of Behavior, 93, 293–312, 2010.
- Brazil. Lei nº 11.794, de 8 de outubro de 2008 (Arouca Law) — Brazilian law on procedures for the scientific use of animals; establishes CONCEA and the CEUAs.
This entry is educational and does not replace assessment, diagnosis or treatment by a qualified health professional. Every person is unique — for a specific case, seek individual guidance.